Journal AES 2003 May Vol 51 Num 5

sustainingmemberorganizations AESAES

VO

LU

ME

51,NO

.5JO

UR

NA

L O

F T

HE

AU

DIO

EN

GIN

EE

RIN

G S

OC

IET

Y2003 M

AY

JOURNAL OF THE AUDIO ENGINEERING SOCIETYAUDIO / ACOUSTICS / APPLICATIONSVolume 51 Number 5 2003 May

The Audio Engineering Society recognizes with gratitude the financialsupport given by its sustaining members, which enables the work ofthe Society to be extended. Addresses and brief descriptions of thebusiness activities of the sustaining members appear in the Octoberissue of the Journal.

The Society invites applications for sustaining membership. Informa-tion may be obtained from the Chair, Sustaining Memberships Committee, Audio Engineering Society, 60 East 42nd St., Room2520, New York, New York 10165-2520, USA, tel: 212-661-8528.Fax: 212-682-0477.

ACO Pacific, Inc.Air Studios Ltd.AKG Acoustics GmbHAKM Semiconductor, Inc.Amber Technology LimitedAMS Neve plcATC Loudspeaker Technology Ltd.Audio LimitedAudiomatica S.r.l.Audio Media/IMAS Publishing Ltd.Audio PartnershipAudio Precision, Inc.AudioScience, Inc.Audio-Technica U.S., Inc.AudioTrack CorporationAutograph Sound Recording Ltd.B & W Loudspeakers LimitedBMP RecordingBritish Broadcasting CorporationBSS Audio Cadac Electronics PLCCalrec AudioCanford Audio plcCEDAR Audio Ltd.Celestion International LimitedCerwin-Vega, IncorporatedClearOne Communications Corp.Community Professional Loudspeakers, Inc.Crystal Audio Products/Cirrus Logic Inc.D.A.S. Audio, S.A.D.A.T. Ltd.dCS Ltd.Deltron Emcon LimitedDigidesignDigigramDigital Audio Disc CorporationDolby Laboratories, Inc.DRA LaboratoriesDTS, Inc.DYNACORD, EVI Audio GmbHEastern Acoustic Works, Inc.Eminence Speaker LLC

Event Electronics, LLCFerrotec (USA) CorporationFocusrite Audio Engineering Ltd.Fostex America, a division of Foster Electric

U.S.A., Inc.Fraunhofer IIS-AFreeSystems Private LimitedFTG Sandar TeleCast ASHarman BeckerHHB Communications Ltd.Innova SONInnovative Electronic Designs (IED), Inc.International Federation of the Phonographic

IndustryJBL ProfessionalJensen Transformers Inc.Kawamura Electrical LaboratoryKEF Audio (UK) LimitedKenwood U.S.A. CorporationKlark Teknik Group (UK) PlcKlipsch L.L.C.Laboratories for InformationL-Acoustics USLeitch Technology CorporationLindos ElectronicsMagnetic Reference Laboratory (MRL) Inc.Martin Audio Ltd.Meridian Audio LimitedMetropolis GroupMiddle Atlantic Products Inc.Mosses & MitchellM2 Gauss Corp.Music Plaza Pte. Ltd.Georg Neumann GmbH Neutrik AGNVisionNXT (New Transducers Ltd.)1 LimitedOntario Institute of Audio Recording

TechnologyOutline sncPacific Audio-VisualPRIMEDIA Business Magazines & Media Inc.Prism Sound

Pro-Bel LimitedPro-Sound NewsPsychotechnology, Inc.Radio Free AsiaRane CorporationRecording ConnectionRocket NetworkRoyal National Institute for the BlindRTI Tech Pte. Ltd.Rycote Microphone Windshields Ltd.SADiESanctuary Studios Ltd.Sekaku Electron Ind. Co., Ltd.Sennheiser Electronic CorporationShure Inc.Snell & Wilcox Ltd.Solid State Logic, Ltd.Sony Broadcast & Professional EuropeSound Devices LLCSound On Sound Ltd.Soundcraft Electronics Ltd.Sowter Audio TransformersSRS Labs, Inc.Stage AccompanySterling Sound, Inc.Studer North America Inc.Studer Professional Audio AGTannoy LimitedTASCAMTHAT CorporationTOA Electronics, Inc.TommexTouchtunes Music Corp.TurbosoundUnited Entertainment Media, Inc.Uniton AGUniversity of DerbyUniversity of SalfordUniversity of Surrey, Dept. of Sound

RecordingVidiPaxWenger CorporationJ. M. Woodgate and AssociatesYamaha Research and Development

In this issue…

Inconsistent Loudspeaker Cone Displacement

Low-Frequency SpatialEqualization

Vector Sound Intensity Probe

Digital Audio BroadcastingEvaluation

Standards:Radio Traffic Data in BroadcastWave Files

Technical Council:Technical Committee Reports

Features…

114th Convention Report,Amsterdam

AUDIO ENGINEERING SOCIETY, INC.INTERNATIONAL HEADQUARTERS

60 East 42nd Street, Room 2520, New York, NY 10165-2520, USATel: +1 212 661 8528 . Fax: +1 212 682 0477E-mail: [email protected] . Internet: http://www.aes.org

Roger K. Furness Executive DirectorSandra J. Requa Executive Assistant to the Executive Director

ADMINISTRATION

STANDARDS COMMITTEE

GOVERNORS

OFFICERS 2002/2003

Karl-Otto BäderCurtis HoytRoy Pritts

Don PuluseDavid Robinson

Annemarie StaepelaereRoland Tan

Kunimaro Tanaka

Ted Sheldon Chair Dietrich Schüller Vice Chair

Mendel Kleiner Chair Mark Ureda Vice Chair

SC-04-01 Acoustics and Sound Source Modeling Richard H. Campbell, Wolfgang Ahnert

SC-04-02 Characterization of Acoustical MaterialsPeter D’Antonio, Trevor J. Cox

SC-04-03 Loudspeaker Modeling and Measurement David Prince, Neil Harris, Steve Hutt

SC-04-04 Microphone Measurement and CharacterizationDavid Josephson, Jackie Green

SC-04-07 Listening Tests: David Clark, T. Nousaine

SC-06-01 Audio-File Transfer and Exchange Mark Yonge, Brooks Harris

SC-06-02 Audio Applications Using the High Performance SerialBus (IEEE: 1394): John Strawn, Bob Moses

SC-06-04 Internet Audio Delivery SystemKarlheinz Brandenburg

SC-06-06 Audio MetadataC. Chambers

Kees A. Immink President

Ronald Streicher President-Elect

Garry Margolis Past President

Jim Anderson Vice President Eastern Region, USA/Canada

James A. Kaiser Vice PresidentCentral Region, USA/Canada

Bob Moses Vice President,Western Region, USA/Canada

Søren Bech Vice PresidentNorthern Region, Europe

Markus ErneVice President, Central Region, Europe

Daniel Zalay Vice President, Southern Region, Europe

Mercedes Onorato Vice President,Latin American Region

Neville ThieleVice President, International Region

Han Tendeloo Secretary

Marshall Buck Treasurer

TECHNICAL COUNCIL

Wieslaw V. Woszczyk ChairJürgen Herre and

Robert Schulein Vice Chairs

COMMITTEES

SC-02-01 Digital Audio Measurement Techniques Richard C. Cabot, I. Dennis, M. Keyhl

SC-02-02 Digital Input-Output Interfacing: Robert A. Finger, John Grant

SC-02- 05 Synchronization: Robin Caine

John P. Nunn Chair Robert A. Finger Vice Chair

Robin Caine Chair Steve Harris Vice Chair

John P. NunnChair

John WoodgateVice Chair

Bruce OlsonVice Chair, Western Hemisphere

Mark YongeSecretary, Standards Manager

Yoshizo Sohma Vice Chair, International

SC-02 SUBCOMMITTEE ON DIGITAL AUDIO

Working Groups

SC-03 SUBCOMMITTEE ON THE PRESERVATION AND RESTORATIONOF AUDIO RECORDING

Working Groups

SC-04 SUBCOMMITTEE ON ACOUSTICS

Working Groups

SC-06 SUBCOMMITTEE ON NETWORK AND FILE TRANSFER OF AUDIO

Working Groups

TECHNICAL COMMITTEES

SC-03-01 Analog Recording: J. G. McKnight

SC-03-02 Transfer Technologies: Lars Gaustad, Greg Faris

SC-03-04 Storage and Handling of Media: Ted Sheldon, Gerd Cyrener

SC-03-06 Digital Library and Archives Systems: William Storm Joe Bean, Werner Deutsch

SC-03-12 Forensic Audio: Tom Owen, M. McDermottEddy Bogh Brixen

TELLERSChristopher V. Freitag Chair

Correspondence to AES officers and committee chairs should be addressed to them at the society’s international headquarters.

Ray Rayburn Chair John Woodgate Vice Chair

SC-05-02 Audio ConnectorsRay Rayburn, Werner Bachmann

SC-05-03 Audio Connector DocumentationDave Tosti-Lane, J. Chester

SC-05-05 Grounding and EMC Practices Bruce Olson, Jim Brown

SC-05 SUBCOMMITTEE ON INTERCONNECTIONS

Working Groups

ACOUSTICS & SOUNDREINFORCEMENT

Mendel Kleiner ChairKurt Graffy Vice Chair

ARCHIVING, RESTORATION ANDDIGITAL LIBRARIES

David Ackerman Chair

AUDIO FOR GAMESMartin Wilde Chair

AUDIO FORTELECOMMUNICATIONS

Bob Zurek ChairAndrew Bright Vice Chair

CODING OF AUDIO SIGNALSJames Johnston and

Jürgen Herre Cochairs

AUTOMOTIVE AUDIORichard S. Stroud Chair

Tim Nind Vice Chair

HIGH-RESOLUTION AUDIOMalcolm Hawksford Chair

Vicki R. Melchior andTakeo Yamamoto Vice Chairs

LOUDSPEAKERS & HEADPHONESDavid Clark Chair

Juha Backman Vice Chair

MICROPHONES & APPLICATIONSDavid Josephson Chair

Wolfgang Niehoff Vice Chair

MULTICHANNEL & BINAURALAUDIO TECHNOLOGIESFrancis Rumsey Chair

Gunther Theile Vice Chair

NETWORK AUDIO SYSTEMSJeremy Cooperstock ChairRobert Rowe and Thomas

Sporer Vice Chairs

AUDIO RECORDING & STORAGESYSTEMS

Derk Reefman ChairKunimaro Tanaka Vice Chair

PERCEPTION & SUBJECTIVEEVALUATION OF AUDIO SIGNALS

Durand Begault ChairSøren Bech and Eiichi Miyasaka

Vice Chairs

SIGNAL PROCESSINGRonald Aarts Chair

James Johnston and Christoph M.Musialik Vice Chairs

STUDIO PRACTICES & PRODUCTIONGeorge Massenburg Chair

Alan Parsons, David Smith andMick Sawaguchi Vice Chairs

TRANSMISSION & BROADCASTINGStephen Lyman Chair

Neville Thiele Vice Chair

AWARDSRoy Pritts Chair

CONFERENCE POLICYSøren Bech Chair

CONVENTION POLICY & FINANCEMarshall Buck Chair

EDUCATIONDon Puluse Chair

FUTURE DIRECTIONSKees A. Immink Chair

HISTORICALJ. G. (Jay) McKnight Chair

Irving Joel Vice ChairDonald J. Plunkett Chair Emeritus

LAWS & RESOLUTIONSRon Streicher Chair

MEMBERSHIP/ADMISSIONSFrancis Rumsey Chair

NOMINATIONSGarry Margolis Chair

PUBLICATIONS POLICYRichard H. Small Chair

REGIONS AND SECTIONSSubir Pramanik Chair

STANDARDSJohn P. Nunn Chair

AES Journal of the Audio Engineering Society(ISSN 0004-7554), Volume 51, Number 5, 2003 MayPublished monthly, except January/February and July/August when published bi-monthly, by the Audio Engineering Society, 60 East 42nd Street, New York, NewYork 10165-2520, USA, Telephone: +1 212 661 8528. Fax: +1 212 682 0477. E-mail: [email protected]. Periodical postage paid at New York, New York, and at anadditional mailing office. Postmaster: Send address corrections to Audio Engineer-ing Society, 60 East 42nd Street, New York, New York 10165-2520.

The Audio Engineering Society is not responsible for statements made by itscontributors.

COPYRIGHTCopyright © 2003 by the Audio Engi-neering Society, Inc. It is permitted toquote from this Journal with custom-ary credit to the source.

COPIESIndividual readers are permitted tophotocopy isolated ar ticles forresearch or other noncommercial use.Permission to photocopy for internalor personal use of specific clients isgranted by the Audio EngineeringSociety to libraries and other usersregistered with the Copyright Clear-ance Center (CCC), provided that thebase fee of $1 per copy plus $.50 perpage is paid directly to CCC, 222Rosewood Dr., Danvers, MA 01923,USA. 0004-7554/95. Photocopies ofindividual articles may be orderedfrom the AES Headquarters office at$5 per article.

REPRINTS AND REPUBLICATIONMultiple reproduction or republica-tion of any material in this Journal requires the permission of the AudioEngineering Society. Permissionmay also be required from the author(s). Send inquiries to AES Edi-torial office.

SUBSCRIPTIONSThe Journal is available by subscrip-tion. Annual rates are $180 surfacemail, $225 air mail. For information,contact AES Headquarters.

BACK ISSUESSelected back issues are available:From Vol. 1 (1953) through Vol. 12(1964), $10 per issue (members), $15(nonmembers); Vol. 13 (1965) to pre-sent, $6 per issue (members), $11(nonmembers). For information, con-tact AES Headquarters office.

MICROFILMCopies of Vol. 19, No. 1 (1971 Jan-uary) to the present edition are avail-able on microfilm from University Microfilms International, 300 NorthZeeb Rd., Ann Arbor, MI 48106, USA.

ADVERTISINGCall the AES Editorial office or send e-mail to: [email protected].

MANUSCRIPTSFor information on the presentationand processing of manuscripts, seeInformation for Authors.

Patricia M. Macdonald Executive EditorWilliam T. McQuaide Managing EditorGerri M. Calamusa Senior EditorAbbie J. Cohen Senior EditorMary Ellen Ilich Associate EditorPatricia L. Sarch Art Director

EDITORIAL STAFF

Europe ConventionsZevenbunderslaan 142/9, BE-1190 Brussels, Belgium, Tel: +32 2 3457971, Fax: +32 2 345 3419, E-mail for convention information:[email protected] ServicesB.P. 50, FR-94364 Bry Sur Marne Cedex, France, Tel: +33 1 4881 4632,Fax: +33 1 4706 0648, E-mail for membership and publication sales:[email protected] KingdomBritish Section, Audio Engineering Society Ltd., P. O. Box 645, Slough,SL1 8BJ UK, Tel: +441628 663725, Fax: +44 1628 667002,E-mail: [email protected] Japan Section, 1-38-2 Yoyogi, Room 703, Shibuyaku-ku, Tokyo 151-0053, Japan, Tel: +81 3 5358 7320, Fax: +81 3 5358 7328, E-mail: [email protected].

PURPOSE: The Audio Engineering Society is organized for the purposeof: uniting persons performing professional services in the audio engi-neering field and its allied arts; collecting, collating, and disseminatingscientific knowledge in the field of audio engineering and its allied arts;advancing such science in both theoretical and practical applications;and preparing, publishing, and distributing literature and periodicals rela-tive to the foregoing purposes and policies.MEMBERSHIP: Individuals who are interested in audio engineering maybecome members of the AES. Applications are considered by theAdmissions Committee. Grades and annual dues are: Full members andassociate members, $90 for both the printed and online Journal; $60 for on-line Journal only. Student members: $50 for printed and online Journal; $20for online Journal only. A subscription to the Journal is included with all mem-berships. Sustaining memberships are available to persons, corporations, ororganizations who wish to support the Society.

Ronald M. AartsJames A. S. AngusGeorge L. AugspurgerJeffrey BarishJerry BauckJames W. BeauchampSøren BechDurand BegaultBarry A. BlesserJohn S. BradleyRobert Bristow-JohnsonJohn J. BubbersMarshall BuckMahlon D. BurkhardRichard C. CabotEdward M. CherryRobert R. CordellAndrew DuncanJohn M. EargleLouis D. FielderEdward J. Foster

Mark R. GanderEarl R. GeddesDavid GriesingerMalcolm O. J. HawksfordJürgen HerreTomlinson HolmanAndrew HornerJyri HuopaniemiJames D. JohnstonArie J. M. KaizerJames M. KatesD. B. Keele, Jr.Mendel KleinerDavid L. KlepperW. Marshall Leach, Jr.Stanley P. LipshitzRobert C. MaherDan Mapes-RiordanJ. G. (Jay) McKnightGuy W. McNallyD. J. MearesRobert A. MoogBrian C. J. Moore

James A. Moorer

Dick PierceMartin PolonD. PreisFrancis RumseyKees A. Schouhamer

ImminkManfred R. SchroederRobert B. SchuleinRichard H. SmallJulius O. Smith IIIGilbert SoulodreHerman J. M. SteenekenJohn StrawnG. R. (Bob) ThurmondJiri TichyFloyd E. TooleEmil L. TorickJohn VanderkooyAlexander VoishvilloDaniel R. von

RecklinghausenRhonda WilsonJohn M. WoodgateWieslaw V. Woszczyk

REVIEW BOARD

Flávia Elzinga AdvertisingIngeborg M. StochmalCopy Editor

Barry A. BlesserConsulting Technical Editor

Stephanie Paynes Writer

Daniel R. von Recklinghausen Editor

Eastern Region, USA/CanadaSections: Atlanta, Boston, District of Columbia, New York, Philadelphia, TorontoStudent Sections: American University, Berklee College of Music, CarnegieMellon University, Duquesne University, Fredonia, Full Sail Real WorldEducation, Hampton University, Institute of Audio Research, McGillUniversity, Peabody Institute of Johns Hopkins University, Pennsylvania StateUniversity, University of Hartford, University of Massachusetts-Lowell,University of Miami, University of North Carolina at Asheville, WilliamPatterson University, Worcester Polytechnic UniversityCentral Region, USA/CanadaSections: Central Indiana, Chicago, Detroit, Kansas City, Nashville, NewOrleans, St. Louis, Upper Midwest, West MichiganStudent Sections: Ball State University, Belmont University, ColumbiaCollege, Michigan Technological University, Middle Tennessee StateUniversity, Music Tech College, SAE Nashville, Northeast CommunityCollege, Ohio University, Ridgewater College, Hutchinson Campus,Southwest Texas State University, University of Arkansas-Pine Bluff,University of Cincinnati, University of Illinois-Urbana-ChampaignWestern Region, USA/CanadaSections: Alberta, Colorado, Los Angeles, Pacific Northwest, Portland, San Diego, San Francisco, Utah, VancouverStudent Sections: American River College, Brigham Young University,California State University–Chico, Citrus College, Cogswell PolytechnicalCollege, Conservatory of Recording Arts and Sciences, Denver, ExpressionCenter for New Media, Long Beach City College, San Diego State University,San Francisco State University, Cal Poly San Luis Obispo, Stanford University,The Art Institute of Seattle, University of Southern California, VancouverNorthern Region, Europe Sections: Belgian, British, Danish, Finnish, Moscow, Netherlands, Norwegian, St. Petersburg, SwedishStudent Sections: All-Russian State Institute of Cinematography, Danish,Netherlands, Russian Academy of Music, St. Petersburg, University of Lulea-PiteaCentral Region, EuropeSections: Austrian, Belarus, Czech, Central German, North German, South German, Hungarian, Lithuanian, Polish, Slovakian Republic, Swiss,UkrainianStudent Sections: Aachen, Berlin, Czech Republic, Darmstadt, Detmold,Düsseldorf, Graz, Ilmenau, Technical University of Gdansk (Poland), Vienna,Wroclaw University of TechnologySouthern Region, EuropeSections: Bosnia-Herzegovina, Bulgarian, Croatian, French, Greek, Israel,Italian, Portugal, Romanian, Slovenian, Spanish, Serbia and Montenegro,Turkish Student Sections: Croatian, Conservatoire de Paris, Italian, Louis-Lumière SchoolLatin American Region Sections: Argentina, Brazil, Chile, Colombia (Medellin), Mexico, Uruguay,VenezuelaStudent Sections: Taller de Arte Sonoro (Caracas)International RegionSections: Adelaide, Brisbane, Hong Kong, India, Japan, Korea, Malaysia,Melbourne, Philippines, Singapore, Sydney

AES REGIONAL OFFICES

AES REGIONS AND SECTIONS

AES JOURNAL OF THE

AUDIO ENGINEERING SOCIETY

AUDIO/ACOUSTICS/APPLICATIONS

VOLUME 51 NUMBER 5 2003 MAY

CONTENT

PAPERSAssessment of Voice-Coil Peak Displacement Xmax...................................................Wolfgang Klippel 307Peak voice-coil displacement is an important parameter for specifying the maximum acoustic output at low frequencies. However, the absence of a single definition for distortion produces ambiguous results that can not be easily compared. For example, choices include distortion in the voice-coil current or cone displacement. Alternatively, a parametric method that provides more detailed information about the cause of the distortion is proposed. A comparison between performance-based and parameter-based techniques illustrates the advantages and disadvantages.

Modal Equalization of Loudspeaker–Room Responses at Low Frequencies.................................................................Aki Mäkivirta, Poju Antsalo, Matti Karjalainen, and Vesa Välimäki 324Compensation for the dominant low-frequency modes in small rooms traditionally uses equalization filters in cascade with the main sound source. In an alternative implementation, multiple sources produce better modal cancellation when traditional methods fail. Case studies show that the extra degrees of spatial freedom afforded by the additional sources make the system more robust. Modal equalization is a design option when the modal density is not high and when the modes are low frequency. It can also be combined with cascade equalization.

A Low-Cost Intensity Probe ......................................R. Raangs, W. F. Druyvesteyn, and H.-E. de Bree 344Unlike ordinary microphones, a sound intensity probe measures the energy flow as a vector direction. It can be computed as the product of scalar pressure and vector velocity. In a conventional probe, velocity is computed as the difference in pressure at a small fixed distance. The authors propose a novel means of directly measuring velocity using the temperature difference between two heated wires mounted in a microminiaturized substrate. When combined with a standard pressure sensor, the probe measures sound intensity over the full spectrum at a single point in space. The paper provides examples of several methods for calibration of the particle velocity sensor used, such as in a standing-wave tube, reverberant room,anechoic space, and reverberation room. Two examples of sound-intensity measurements are provided and are compared with a conventional sound intensity probe.

ENGINEERING REPORTSIndustry Evaluation of In-Band On-Channel Digital Audio Broadcast Systems..............David Wilson 358The proposed techniques for terrestrial broadcasting of digital audio from iBiquity were evaluated using standard metrics: coverage, compatibility, interference, and quality. In order to maintain compatibility with existing AM and FM analog broadcasting, digital information was added as low-amplitude side bands around the main analog spectrum. While the proposed solution achieves the desired goal of preserving the existing competitive balance between radio broadcasters, there is an additional interference outside the protected geographic region. Subjective listening tests confirmed that digital audio improved the quality even at reduced bit rates.

STANDARDS AND INFORMATION DOCUMENTSAES46-2002 AES standard for network and file transfer of audio — Audio-file transfer and exchange — Radio traffic audio delivery extension to the broadcast WAVE file format ................. 369AES Standards Committee News........................................................................................................... 384Julian Dunn; resonance of loudspeaker cones

FEATURES114th Convention Report, Amsterdam .................................................................................................. 386

Exhibitors............................................................................................................................................... 402Program.................................................................................................................................................. 405

TECHNICAL COUNCIL REPORTSTechnical Committee Reports: Emerging Trends in Technology ........................................................ 442

DEPARTMENTS News of the Sections ........................................452Upcoming Meetings ..........................................458Advertiser Internet Directory............................459Sound Track........................................................460New Products and Developments....................460

Available Literature ...........................................461Membership Information...................................463In Memoriam ......................................................465Sections Contacts Directory ............................466AES Conventions and Conferences ................472

PAPERS

0 INTRODUCTION

Loudspeakers that have similar linear parameters maybehave quite differently at higher amplitudes. In the large-signal domain, the physical limits require a compromisebetween maximum amplitude, efficiency, signal distor-tion, cost, weight, size, and other factors.

Thus assessing the large-signal performance by a num-ber of meaningful parameters becomes more and moreimportant. There are a few “traditional” parametersdescribing the permissible load and the maximum outputof the driver. One is the maximum (linear) peak displace-ment Xmax, which limits the displaced air volume and the

maximum sound pressure output at low frequencies, asshown in Fig. 1.

The parameter peak displacement Xmax, listed on nearlyevery serious specification sheet, is the interface betweendriver and loudspeaker system design. However, manufac-turers use different ways to assess Xmax, and stated valuesare not comparable.

1) Small [1] and Gander [2] suggested a motor topology–based method, where Xmax has been derived from geomet-rical data such as gap depth and voice-coil height. Thisapproach neglects voice-coil offset, magnetic field asym-metries, suspension nonlinearity, and other driver defects.

2) Small [1] also suggested the first performance-basedmethod using an harmonic distortion measurement at theresonance frequency fs to find the peak displacement Xmaxthat will give 10% total harmonic distortion in the sound

J. Audio Eng. Soc., Vol. 51, No. 5, 2003 May 307

Assessment of Voice-Coil Peak Displacement Xmax*

WOLFGANG KLIPPEL, AES Fellow

Klippel GmbH, Dresden, Germany

The voice-coil peak displacement Xmax is an important driver parameter for assessing themaximum acoustic output at low frequencies. The existing standard AES2-1984 defines thepeak displacement Xmax by measuring harmonic distortion in either voice-coil current ordisplacement. This freedom of choice gives completely different and controversial results.After a critical review of this performance-based technique, an amendment of this method issuggested. Alternatively, a parameter-based method is developed giving more detailed infor-mation about the cause of the distortion, limitations, and defects. The relationship betweenperformance-based and parameter-based methods is discussed, and both techniques are testedwith real drivers.

* Manuscript received 2002 June 18; revised 2002 July 29,and 2003 January 17 and February 24.

Fig. 1. Fundamental component of radiated sound pressure frequency response at 1-m distance measured at input voltages u increasedin 5-dB increments.

50

60

70

80

90

100

100

Pfar

[dB]

frequency [Hz]100020

u

KLIPPEL PAPERS

pressure output. This approach assesses the overall transferbehavior of the driver and does not require any informationabout the motor topology or loudspeaker nonlinearities. Amodified method became part of the recommendation AES2-1984 [3] but has not been recognized as a common ref-erence for Xmax. Gander [2] also measured Xmax by search-ing for 3% third-order harmonic distortion in the soundpressure output. For drivers having a dominant Bl(x) non-linearity and a sufficiently linear suspension he found agood agreement between his performance-based andtopology-based methods. Clark [4] developed an alterna-tive performance-based method for Xmax, which uses twocriteria to consider the effects of the Bl(x) and suspensionnonlinearities. Exciting the driver by a two-tone signal hemeasured the amplitude modulation (AM) of the high-frequency tone (voice tone) due to high voice-coil dis-placement generated by a low-frequency tone (bass tone).He used a modulation factor of 29.29% (equivalent to areduction in the voice tone level of 3 dB) as a first crite-rion that limits Xmax. To consider nonlinearities of the sus-pension he measure the ratio between displacement andvoltage of the bass tone and suggested a reduction of theexcursion sensitivity down to 50% as a second criterionthat limits Xmax.

3) Clark [5] also suggested the first parameter-basedmethod for assessing Xmax using criteria obtained from acertain decrease of the nonlinear force factor Bl(x) andcompliance Cms(x) characteristic.

4) Finally, some manufacturers use undefined methodsresulting in impressive values of Xmax without a clear rela-tionship to the physics involved.

Since loudspeaker system design requires reliableobjective data to select an optimum driver, there is a needfor quantifying Xmax more accurately. This paper discussesexisting methods of Xmax assessment and suggest new,more clear and accurate definitions of this parameter. Thenumerical simulations and practical measurements areperformed by using the Klippel analyzer system [6].

1 GLOSSARY OF SYMBOLS

a Radius of circular radiatorBl(x) Force factor of motor depending on

voice-coil displacementBlmin Minimum force factor ratio used as

threshold for XBlCms(x) Mechanical compliance of driver sus-

pension, inverse of stiffness Kms(x)Cmin Minimum compliance ratio used as

threshold for XCXdc Dc component in voice-coil displacementd Threshold of acceptable distortion used

in performance-based methoddt Total harmonic distortion defined ac-

cording to IEC 60268dh2,f1 Second-order harmonic distortion con-

sidering sound pressure componentP(2 f1)

dh3,f1 Third-order harmonic distortion con-

sidering sound pressure componentP(3 f1)

d2 Second-order modulation distortionconsidering sound pressure compo-nent P( f2 f1)

d3 Third-order modulation distortion con-sidering sound pressure componentP( f2 2 f1)

d2,i Second-order modulation distortion con-sidering current component I( f2 f1)

d3,i Third-order modulation distortion con-sidering current component I( f2 f1)

dh2,f2 Second-order harmonic distortion con-sidering sound pressure componentP(2 f2)

dh3,f2 Third-order harmonic distortion con-sidering sound pressure componentP(3 f2)

f Frequencyf1 Frequency of bass tone in two-tone

signalf2 Frequency of voice tone in two-tone

signalfs Resonance frequency of loudspeakersF Bl(x)i Electrodynamic force driving the mec-

hanical systemFm(x, i) Reluctance forcei Electric input currentLe(x) Part of voice-coil inductance that is

independent of frequencyLx(x) dLe/dx Local derivative of Le(x)L2(x) Part of voice-coil inductance that is

dependent on frequencyMms Mechanical mass of driver diaphragm

assembly including voice-coil and airload

pn Peak sound pressure in near field of acoustic radiator

P( f ) FFT spectrum of sound pressure signalQts Total loss factor of driver at fs con-

sidering all system resistanceQms Loss factor of driver at fs considering

driver nonelectrical resistances onlyR2(x) Electrical resistance due to additional

losses caused by eddy currentsRe Dc resistance of voice coilRms Mechanical resistance of driver sus-

pension lossesu Driving voltage at loudspeaker terminalsv Velocity of voice coilx Instantaneous voice-coil displacementXmax Maximum peak displacementxpeak Peak sinusoidal displacementxrms Rms value of voice-coil displacementXC Displacement limit due to Cms(x)

nonlinearityXclip Displacement limit due to mechanical

clippingXBl Displacement limit due to Bl(x) nonlinearityXL Displacement limit due to Le(x), L2(x),

308 J. Audio Eng. Soc., Vol. 51, No. 5, 2003 May

PAPERS VOICE-COIL PEAK DISPLACEMENT

and R2(x) nonlinearitiesXD Displacement limit due to Doppler

effectZe( f ) Electric input impedance of loudspeakers∆Zmax Maximum variation of electric input

impedance used as threshold for XLρ0 Density of air

2 CRITICAL REVIEW OF AES2-1984

2.1 Definition of Xmax

The current AES recommended practice [3] defines:

… the voice-coil peak displacement at which the“linearity” of the motor deviates by 10%. Linearitymay be measured by percent distortion of the inputcurrent or by percent deviation of displacement ver-sus input current. Manufacturer shall state methodused. The measurement shall be made in free air at fs.

2.2 AmbiguitiesIn the existing definition the linearity of the motor

determines the peak displacement Xmax. However, the termlinearity and the techniques for its assessment are notclearly defined. It is not clear whether the motor linearityis only restricted to the variation of the force factor Bl(x)versus displacement x or considers other driver nonlinear-ities such as the voice-coil inductance Le(x) and the com-pliance Cms(x) of the mechanical suspension.

The first suggestion to monitor the percent distortion ofthe input current may be interpreted as a harmonic distor-tion measurement using an excitation tone at the resonancefs. However, it is not clear whether the percent distortionrefers to the total harmonic distortion as suggested by Small[1] and defined in the IEC standard [7] or to the third-order

distortion suggested by Gander [2] or to other relative dis-tortion measures used in loudspeaker measurements.

The second suggestion to monitor the deviation of thevoice-coil displacement versus current is an unusual tech-nique. Unfortunately the AES standard neither describesthis measurement in detail nor refers to any literature. Clark[4] interpreted the second suggestion as monitoring the pro-portionality of cone excursion to drive voltage, whichcomes close to his excursion sensitivity criterion. The devi-ation from proportionality can be assessed by measuring thetotal harmonic distortion in the voice-coil displacement.

2.3 AssumptionsThe Xmax definition in AES2 makes also the following

assumptions:

• The motor linearity is the only and most critical factorfor assessing Xmax.

• The measurement of harmonic distortion in current ordisplacement produces comparable values of Xmax.

• There is a simple relationship between distortion ampli-tude and peak displacement.

• The distortion increases monotonically with the ampli-tude of the input signal, and 10% distortion correspondsto only one unique value of Xmax.

• The measurement of harmonic distortion at the reso-nance fs reveals effects of motor nonlinearity adequately.

2.4 Fictitious DriverThe validity of the assumptions of AES2-1984 are

checked by applying the existing definition of Xmax to afictitious driver. The driver corresponds to the equivalentcircuit depicted in Fig. 2 and has the parameters listed inTable 1.

The force factor Bl(x), shown in Fig. 3, is not a constant


Fig. 2. Electromechanical equivalent circuit of driver considering dominant nonlinearities.

Mms

Cms

(x) Rms

Bl(x)

Le(x)R

e(T

V)

v

Fm

(x,i)

i

Bl(x)v Bl(x)i

L2(x)

R2(x)

U

i2

Table 1. Small-signal parameters.

Parameter Value Unit

Re 3.5 ΩLe(x) constant 1 mHR2 0.5 ΩL2 1 mHBl(0) 5 N/ACms(x) constant 0.7 mm/NQms 7Qes 0.68Qts 0.62fs 47 Hz

KLIPPEL PAPERS

parameter but a nonlinear function of the voice-coil dis-placement. Bl(x) has a symmetrical bell-shaped form,approaching zero at high positive and negative displace-ments. The early strong decay of the Bl curve is typical fora voice-coil height of 5 mm that equals the gap depth.Since the Bl(x) curve is perfectly symmetrical to x 0,the driver produces only third-order and other odd-orderdistortions as long as the equilibrium point is stable and nobifurcation occurs.

All other nonlinearities inherent in real drivers are neg-lected in the fictitious driver to keep the test case as sim-ple as possible. Thus the compliance Cms(x) constantand the voice-coil inductance Le(x) constant areassumed as independent of the displacement x. Those sim-plifications are acceptable for the validity check because itis sufficient to find at least one case where the standard-ized definition fails.

According to the measurement conditions defined inAES2-1984, the operation of the driver is modeled in free air.

2.5 Simulation of Large-Signal BehaviorThe fictitious driver is modeled precisely by the nonlin-

ear differential equation [Eq. (14)] given in the Appendix.For arbitrary input time-domain signals all state variables(current, displacement, and so on) can be calculated by thenumerical integration [8] of Eq. (14). For a sinusoidalstimulus the fast Fourier transform (FFT) analysis is usedto separate the fundamental from the distortion compo-nents and to calculate the total harmonic distortion,

d ft _ i

%

P f P f P f P Kf

P f P f P Kf

2 3

2 3100

2 2 2 2

2 2 2

g

g

_ _ _ _

_ _ _

i i i i

i i i

(1)

as defined by IEC 60268 [7].Fig. 4 shows the total harmonic distortion dt of the cur-

rent, sound pressure, and displacement versus the peak

displacement x predicted for the fictitious driver excitedwith a single tone at f fs of variable voltage u.

2.6 Applying Xmax DefinitionSearching for a total harmonic distortion of dt 10%,

the input current gives a peak displacement Xmax of about0.6 mm according to AES2-1984. That is a very smallvalue compared to the voice-coil height of about 5 mm.No manufacturer would agree to specify the workingrange of a loudspeaker to such a small signal domain. Inthe range of 0.6 mm < x < 0.6 mm Bl(x) varies by only5%, and the distortion in the radiated sound pressure is amere 2%. Most likely the manufacturer would considerthe alternative method. The total harmonic distortion indisplacement remains very small and does not reach 10%,even if the coil is entirely outside the gap (x 20 mm).Obviously it makes no sense to the peak displacementXmax to four times the voice-coil height.

2.7 Xmax from Sound Pressure DistortionSome users modified the current Xmax definition and

applied the threshold of 10% to the total harmonic distor-tion in the radiated sound pressure following the sugges-tion of Small [1]. Usually this provides more reasonableestimates of Xmax. However, this method may lead to mul-tiple values of voice-coil displacement corresponding tothe same value of distortion. For example, the fictitiousdriver provides three different values (1.5 mm, 8 mm, and13.5 mm) as candidates for Xmax. The question arises:which of these values of voice-coil displacement shouldbe considered as Xmax?

2.8 What Is Wrong with the Existing Definition?Apparently some of the assumptions made in the exist-

ing Xmax definition are not valid. First, the harmonic dis-tortion in the voice-coil current and displacement at fs arenot of the same order of magnitude. The reason is quitesimple. The amplitude of the fundamental component ofthe voice-coil current is minimal at the resonance fre-quency fs where the electrical impedance is maximal.


Fig. 3. Bl(x) product versus voice-coil displacement of fictitiousdriver used in simulation.

Fig. 4. Total harmonic distortion in current (g), sound pressure(– – –), and displacement (–––) for a single excitation tone at fsversus voice-coil peak displacement x.

0 ,0

0 ,5

1 ,0

1 ,5

2 ,0

2 ,5

3 ,0

3 ,5

4 ,5

5 ,0

5 ,5

-1 5 -1 0 -5 0 5 1 0 1 5

B l

[N /A ]

d isp la ce m en t x [m m ]

0 2 4 6 8 10 12 14 16 18 200

10

20

30

40

50

60

70

80

voice coil peak displacement [ mm]

dt

%


However, the harmonic components of the voice-coilcurrent see a much lower impedance at higher frequenciesand therefore get amplified to a higher amplitude. Thissuppression of the fundamental’s spectral component istypical only for the voice-coil current and is inherent nei-ther to the displacement nor to the sound pressure signal.Also, there is no simple relationship between harmonicdistortion and peak displacement. Instead of a monotoni-cally equal increase the distortion level stagnates at a rel-atively small value, giving multiple values for Xmax. Fig. 5shows the displacement versus the amplitude of the inputvoltage.

At 5-mm peak displacement, the instantaneous level ofthe force factor Bl(x) reduces to 0.20Bl(x 0) and at 12mm the force factor almost vanishes. Surprisingly themotor works properly and produces almost a linear rela-tionship between displacement and voltage. There are tworeasons for that:

• The electrical damping caused by Bl(x)2/Re decreasesand the mechanical Qms dominates the total dampingQts(x). The rising value of Qts(x) compensates for thereduced excitation force F Bl(x)i.

• The 90-deg phase shift between current i and displace-ment x at the resonance frequency fs still provides goodexcitation conditions, as shown in Fig. 6. When current

and voltage are maximal at time t1, then the instanta-neous value of Bl (x) is also maximal. This produces ahigh driving force F Bl (x)i and a low value ofQes(x) 2π fsMmsRe/Bl (x)2. The high value of backEMF produces the small dip in the voice-coil current. Atthe time t2, when the coil is completely out of the gap,the low value of the force factor Bl(x) coincides with alow value of the instantaneous current and voltage.

Above and below the resonance frequency, variations ofBl(x) have a significant effect on the output. Fig. 7 showsthe rms amplitude of the voice-coil displacement Xrms as afunction of frequency f1 for varying input voltages u1.Increasing the voltage u1 from 2 to 20 V in 2-V incre-ments, there is a slower increase of the displacement(amplitude compression) at frequencies below and abovethe resonance fs. This effect is mainly caused by the phaserelationship between current i and displacement x, wherethe maximum current coincides with the reduced Bl(x),giving less excitation force F Bl(x)i to the fundamentalcomponent.

Excitation tones one octave above resonance may causean unstable behavior at high voltages u1, which is typicalfor an electrodynamic motor. Even if the rest position ofthe coil is well centered in a symmetrical Bl(x) curve, thecoil has a tendency to slide down on either slope of Bl(x).


Fig. 6. Voice-coil current i and displacement x versus time for excitation tone at resonance frequency fs 47 Hz.

-20

-15

-10

-5

0

5

10

15

20

0,28 0,29 0,30 0,31 0,32 0,33 0,34 0,35 0,36 0,37

X [m

m]

Time [s]

x(t) i(t)

[mm]

[A]

x(t)

i(t)

t1 t2 time [s]

Fig. 5. Displacement amplitude for a tone at fs 47 Hz versus input voltage u1.

0,0

2,5

5,0

7,5

12,5

0,0 2,5 5,0 7,5 10,0 12,5 15,0 17,5 20,0

X

[mm] (rms)

voltage u1 [V]

KLIPPEL PAPERS

Fig. 8 shows the total harmonic distortion dt in the radi-ated sound pressure versus the frequency f1 of the excita-tion tone for varying amplitudes of u1. The harmonic dis-tortion is maximal at excitation frequencies belowresonance. This is caused by the low excitation force F Bl(x)i due to the coincidence of the current maximum andthe Bl(x) minimum and the high-pass property of thesound pressure response.

There is a pronounced minimum at the resonance fre-quency where the effects of the nonlinear excitation forceand nonlinear damping partly compensate each other.However, there is a second maximum approximately oneoctave above resonance. Here the phase relationshipbetween current and displacement leads to a nonlinearexcitation. At higher frequencies, f > 10 fs, where theamplitude of displacement gets small, the harmonic dis-tortion becomes negligible. This is typical for any driverwith Bl(x) nonlinearity.

3 NEW PERFORMANCE-BASED METHOD

Although the current method for assessing Xmax basedon the measurement of harmonic distortion fails, the gen-

eral ideas of this approach are worth considering:

• Derive Xmax from the driver performance.• Dispense with a physical driver model.• Use standard measurement equipment.• Keep the procedure simple and fast.

3.1 Critical Distortion MeasurementsA single tone is a very popular stimulus in distortion

measurements because it can be generated easily, and themeasured harmonic distortion can be presented in relationto the excitation frequency. These results represent thetotal distortion produced by more complex audio signalsquite well as long as the transfer system comprises onlystatic nonlinearities imbedded in two linear systems withconstant amplitude response. For example, the limiting ofa power amplifier can be modeled by a memoryless sys-tem. In this case there is a simple relationship betweenharmonic and other intermodulation distortion, and themeasurement with a single tone is also meaningful for amusic signal of the same amplitude.

The dominant nonlinearities in electrodynamic trans-ducers are the parameters that vary with the voice-coil dis-


Fig. 8. Total harmonic distortion in radiated sound pressure of fictitious driver excited by a single tone f1 for varying voltage u1 (2-Vincrements).

0

10

20

30

40

50

60

70

80

90

100

dt

%

frequency f1 [Hz]

100 1000

Voltage u1

20

20 V

2 V

fs

Fig. 7. Voice-coil displacement amplitude of fictitious driver excited by a single tone f1 for varying voltage u1 (2-V increments).

0,0

2,5

5,0

7,5

10,0

12,5

Xrms

[mm]

frequency f1 [Hz]

100 1000

u1

20

2 V

fs

20 V


placement. The displacement x is essentially a low-pass-filtered signal. Also, the other state variables such as cur-rent i and velocity v have different spectral characteristics.In the nonlinear terms of the differential Eq. (14), such asthe electrodynamic driving force F Bl(x)i, the time-domain signals [x(t), i(t), …] are multiplied with eachother. The instantaneous spectrum of current, displace-ment, and velocity determines the spectral characteristicsof the intermodulation distortion in the output signal. Theresults of harmonic distortion measurements based on asingle-tone stimulus are not sufficient to predict otherkinds of distortion generated by a more complex excita-tion signal.

3.2 Two-Tone Excitation SignalMeasurements of intermodulation components are there-

fore required to get more meaningful results. There aremany ways of performing such measurements. Usually amultitone stimulus is used consisting of two or more com-ponents. An extensive number of excitation tones mightrepresent an audio signal quite well, but it will also pro-duce a lot of data, which have to be interpreted [9]. Theexisting IEC 60268, however, provides a more practicalapproach. A two-tone signal will provide the most impor-tant information if the frequencies f1 and f2 of the first andsecond excitation tones are selected carefully. Since thedominant nonlinearities of most common transducers arerelated to the voice-coil displacement, the first tone f1 gen-erates sufficient voice-coil displacement and is called basstone. The second tone f2 represents any higher frequencycomponent in the passband of the transducer and is calledvoice tone. Fig. 9 shows the sound pressure spectrum ofthe fictitious driver excited with a two-tone stimulus, rep-resented by bold lines.

The thin lines in the sound pressure spectrum in Fig. 9are the harmonic components at multiple frequencies of f1

and the difference and summed-tone intermodulation atf2 (k 1) f1 and f2 (k 1) f1, respectively, centeredaround the voice tone f2. Usually all higher order compo-nents decrease rapidly with the increasing order of k. ThusIEC 60268 considers only the low-order components sum-marized as second-order modulation distortion,

%dP f

P f f P f f100

2

2

2 1 2 1

_

_ _

i

i i(2)

and third-order intermodulation distortion,

%dP f

P f f P f f2 2100

3

2

2 1 2 1

_

_ _

i

i i(3)

referred to the amplitude of the voice tone f2.Although the amplitudes of both excitation tones are

equal, the sound pressure level of fundamental f2 in Fig. 9 ismore than 20 dB lower than that at f1 fs. The amplitudecompression of the voice tone f2 is shown more clearly inFig. 10, where the sound pressure level of both fundamen-tals is displayed versus the terminal voltage u1 u2.

For input voltages below 2 V there is a linear relation-ship between input and output amplitudes because thepeak displacement is below 1.5 mm. At higher voltagesthe sound pressure level of the voice tone stagnatesbecause the coil remains outside the gap for most of thetime and the effective excitation of f2 does not increase. Ifthe voice tone f2 is measured without the bass tone f1, thenthere will be almost no amplitude compression.

Fig. 11 shows the third-order modulation distortionaccording to IEC 60268 versus frequency f2 while the basstone is fixed to the resonance frequency f1 fs. The volt-age u1 u2 is increased by 2-V increments.

Neglecting some interference between harmonic andintermodulation components at multiples of f1, the inter-modulation components d2 and d3 are almost constant for


Fig. 9. Spectrum of radiated sound pressure signal of fictitiousdriver excited by two tones f1 fs 47 Hz and f2 980 Hz atu1 u2 20 Vrms.

Fig. 10. Amplitude of fundamental sound pressure componentfor two-tone excitation signal at f1 fs 47 Hz and f2 780 Hzversus input voltage u1 u2.

50

55

60

65

70

75

80

85

90

100

0 250 500 750 1000 1250 1500 1750 2000 2250 2500

p ( )

dB

Frequency [Hz]

Distortion Fundamental

f2

f1[dB]

50

55

60

65

70

75

80

85

95

10

[dB]

voltage u1 = u2 [V]

P(f2)

P(f1)

10.1

KLIPPEL PAPERS

f2 > 3fs. This is typical for drivers with dominant Bl(x)nonlinearity. The intermodulation distortions d2 and d3increase monotonically with the input voltage. For aninput voltage of u1 u2 1.3 Vrms the fictitious driverproduces already d3 10%. This corresponds to a peakdisplacement Xmax 1.2 mm.

The loudspeaker modeling and numerical simulationshow that the combination of harmonic and intermodula-tion distortion measurements provides essential informa-tion for defining Xmax more clearly.

3.3 Measurement SetupA two-tone signal with fixed frequencies is an optimum

stimulus, which can be produced simply by two sinusoidalgenerators. Performing a series of measurements withvarying frequencies f1 and f2 is not necessary, but varia-tions of the terminal voltage are required. This is a majordifference compared to measuring a linear system that hasthe same transfer function at low and high amplitudes.

Setting the bass tone f1 at the resonance frequency fsgives high voice-coil displacements, low voice-coil cur-rent, and sufficient sound pressure level output. The fre-quency of voice tone f2 is apparently not critical. It shouldbe much higher than that of f1 so that it generates not muchdisplacement but significant input current. To avoid inter-ference with harmonics of the fundamental frequency f1, afractional ratio f2/f1 5.5 should be used between bothtones. However, the IEC standard recommends f2 > 8f1, mak-ing the second-order modulation distortion d2 sensitive tothe Doppler effect. The standard also suggests an amplituderatio of u1 4u2. Using the same amplitude for both tones,u1 u2, would give similar values of the modulation distor-tion d2 and d3 but a much better signal-to-noise ratio for theintermodulation distortion, and allows a comparisonbetween the harmonics of the bass and voice. Although thesemodifications bring some advantages, it is recommended tostay close to the methodology defined in IEC 60268.

To assess the output distortion, monitoring of the soundpressure signal is required. It is recommended to set the

microphone in the near field of the driver, close to thediaphragm, in order to have sufficient signal-to-noise ratioand minimize the influence of reflections. If the driver ismounted in a flat baffle (half-space), then the sound pres-sure measurement in the near field [10] can be used to cal-culate the peak displacement xpeak according Gander’ssuggestion [2, eq. (23), p. 15],

.ρ π

xf a

p

4peak

n

02 2

(4)

Today the direct measurement of the displacement byusing an inexpensive laser displacement sensor based on thetriangulation principle is an efficient alternative. The inputcurrent is also monitored using a shunt or current sensor.

An FFT analysis of the sound pressure signal p(t) pro-vides the fundamentals P( f1) and P( f2) of bass tone andvoice tone, respectively, harmonics P(kf1) and P(kf2), and thesum and difference tone intermodulation P( f2 (k 1) f1)of order k. In addition to the measure distortions dt, d2, andd3, the second-order harmonic distortion

d fh2 _ i

%

P f P f P f P Kf

P f

2 3

2100

2 2 2 2

g_ _ _ _

_

i i i i

i

(5)

and the third-order harmonic distortion

d fh3 _ i

%

P f P f P f P Kf

P f

2 3

3100

2 2 2 2

g_ _ _ _

_

i i i i

i

(6)

are calculated to reveal the effects of symmetrical andasymmetrical parameter variations more clearly.


Fig. 11. Third-order intermodulation distortion in radiated sound pressure response for two-tone excitation comprising a variable tonef2 and a fixed tone f1 fs 47 Hz with varying voltage (2-V steps).

KLIPPEL

25

50

75

100

125

200 40001000

frequency f2 [Hz]

400

%

Voltage

20 V

2 V

voltage u1 and u2


3.4 Dominant Source of DistortionApplying the methods of IEC 60268 to the spectral

components of sound pressure, displacement, and currentleads to a set of distortion measures described in Table 2.

The distortion measurements listed in Table 2 give someclues about the physical causes that limit the peak dis-placement Xmax. The relationships are represented bycrosses in Table 3. The dominant nonlinearities caused bythe force factor Bl(x), inductance Le(x), and complianceCms(x) of the mechanical suspension and the Doppler effectmay produce substantial values of distortion (greater than5%). They are the limiting factors of Xmax in common

transducers and are emphasized by bold crosses. The vari-ations of the radiation conditions cause relatively small dis-tortions for frequencies below 1 kHz. The other nonlinear-ities, such as flux modulation and partial cone vibration,produce much less distortion in common transducers.

The use of Table 3 is quite simple. A driver having sig-nificant values of dh2,f1 and d2 suffers from Bl asymmetrycaused by a coil offset or field geometry. If a high value ofd2 coincides with significant d2,i in the input current, thenthe asymmetry of the inductance Le(x) should be reducedby using a shortcut ring or copper cap. The Doppler effectcan be easily identified by getting a high value of d2 cou-pled with a low value of d2,i. An asymmetrical suspension


Table 2. Distortion measures based on two-tone signal.

DistortionMeasures Interpretation

Xdc The dc part in the displacement is generated dynamically by signal rectification due to parameter asymmetries. Thedc part Xdc generated by the special two-tone signal is mainly caused by suspension asymmetries shifting the coilalways toward the minimum of the nonlinear stiffness curve Kms(x).

dh2,f1 The second-order harmonic distortion considering sound pressure component P(2 fs) in Eq. (5) is a good indicatorfor the asymmetrical stiffness Kms(x). It also reflects some effects of the asymmetrical force factor Bl(x). It is insensi-tive to the nonlinear inductance Le(x) because the amplitude of the current is low at the resonance.

dh3,f1 The third-order harmonic distortion considering sound pressure component P(3fs) in Eq. (6) is a good indicator of thesymmetrical variations of the stiffness Kms(x). It partly reflects the symmetrical variations of the force factor Bl(x). Itis insensitive to the nonlinear inductance Le(x) because the amplitude of the current is low at the resonance.

d2 The second-order intermodulation distortion considering sound pressure components P( f2 f1) in Eq. (2) is a goodindicator of the asymmetrical variations of inductance Le(x), the force factor Bl(x), and the Doppler effect. The effectof asymmetries in stiffness Kms(x) is negligible.

d3 The third-order harmonic distortion in sound pressure considering P( f2 2 f1) in Eq. (3) is a good indicator ofthe symmetrical variations of force factor Bl(x) due to the limited voice-coil height. The effects of the other non-linearities such as inductance Le(x), stiffness Kms(x), and Doppler effect are negligible.

d2,i The second-order intermodulation distortion considering current components I( f2 2 f1) in Eq. (2) is a good indica-tor of the asymmetrical variations of inductance Le(x). The effect of the other nonlinearities such as force factor Bl(x),stiffness Kms(x), and Doppler effect are negligible.

d3,i The third-order intermodulation distortion considering current components I( f2 2 f1) in Eq. (3) is a good indica-tor of the symmetrical variations of inductance Le(x). The effects of the other nonlinearities such as force factor Bl(x),stiffness Kms(x), Doppler effect, and radiation are negligible.

dh2,f2 The second-order harmonic distortion considering sound pressure component P(2 f2) in Eq. (5) is a good indicator ofthe reluctance force due to asymmetrical inductance Le(x). It also reveals flux modulation due to the asymmetricalvariation of Bl(i) versus voice-coil current i and other nonlinearities in the driver (such as partial vibration in thediaphragm). This measurement is insensitive to variations of force factor Bl(x) and stiffness Kms(x) versus displace-ment and Doppler effect.

dh3,f2 The third-order harmonic distortion considering sound pressure component P(3f2) in Eq. (6) is a good indicator of fluxmodulation due to the symmetrical variation of Bl(i) versus voice-coil current i. It also reflects some other minor non-linearities in the driver (such as partial vibration in the diaphragm). This measurement is insensitive to variations offorce factor Bl(x), stiffness Kms(x), and inductance Le(x) versus displacement and Doppler effect.

Table 3. Relationship between nonlinearities and distortion measures.

Physical Cause XDC dh2,f1 dh3,f1 d2 d3 d2,i d3,i dh2,f2 dh3,f2

Coil offset and asymmetry of Bl(x) x xCoil height x xAsymmetry in suspension x xSymmetrical limiting of suspension xAsymmetry in Le(x) x xSymmetrical variation in Le(x) x xReluctance force xFlux modulation x xDoppler effect xNonlinear radiation x xPartial cone vibration x x

Note: Bold symbols represent significant distortion.

KLIPPEL PAPERS

can be detected easily by high values of dh2,f1 and signifi-cant Xdc while the other second-order distortions are small.

3.5 New Xmax DefinitionSummarizing the considerations, a new Xmax definition

may be suggested: Xmax is the voice-coil peak displace-ment at which the maximum value of either the total har-monic distortion dt or the second-order modulation distor-tion d2 or the third-order modulation distortion d3 in theradiated sound pressure is equal to a defined threshold d.The driver is excited by the linear superposition of a firsttone at the resonance frequency f1 fs and a second tonef2 8.5fs with an amplitude ratio of 4:1. The total har-monic distortion dt assesses the harmonics of f1 and themodulation distortions d2 and d3 are measured accordingto IEC 60268. It is recommended to operate the driver ina baffle (half-space) to measure the sound pressure in thenear field and to use the threshold d 10%. The manu-facturer shall state Xmax, the dominant type of distortion(dt, d2, or d3), and the value of the threshold d used.

3.6 Practical Use1) Measure the resonance frequency fs of the driver.2) Excite the driver under voltage drive with a two-tone

signal at f1 fs and f2 8.5fs with an amplitude ratio of4:1.

3) Perform a series of measurements while increasingthe input amplitude and measuring the sound pressure inthe near field of the driver. If a displacement sensor isavailable, measure the voice-coil displacement.

4) Perform a spectral analysis of the sound pressuresignal and determine the total harmonic distortion andintermodulation distortion according IEC 60268. Thesound pressure measurement for assessing the distortiondoes not require a calibrated microphone. This measure-ment may be performed in the near field of the driveroperated in free air without any enclosure because the can-cellation from the rear radiation is negligible. However, ifthe voice-coil displacement is calculated from the soundpressure level of the bass tone f1 by using Eq. (4), a cali-brated microphone is required and the driver should beoperated in a baffle (half-space environment).

5) Search for the minimum value of the peak displace-ment where either dt, d2, or d3 is equal to the threshold d.For lack of better arguments and to preserve consistencywith the existing AES2-1984 it is recommended to used 10%. However, the manufacturer may use a differentvalue if stated with the measured Xmax.

6) State the peak displacement Xmax and the type of dis-tortion limiting the excursion.

For example, a statement

. @ % , < %X mm d d d3 8 10 10 max t2 3_ i

means that a driver provides a maximum peak displace-ment of Xmax 3.8 mm, where the second-order modula-tion distortion is dominant, and produces the threshold of10% distortion. This statement implies that the total har-monic distortion and the third-order distortion are lessthan 10%, which can be added in parentheses (optional).Thus the suspension and the voice-coil height are most

likely not the limiting factors for the excursion of thisdriver.

4 PARAMETER-BASED METHOD

Although the performance-based method gives someindication about the dominant source of distortion, thisapproach fails to assess the limiting factors of each non-linearity quantitatively. This information is required whenthe driver designer would like to improve the maximumoutput of the driver while keeping the cost and otherparameters constant. The system designer also needs thesedata to select a driver that produces distortions at Xmax thatare acceptable for a particular application (subwoofer,woofer, or full-band system). The parameter-basedmethod provides a separate value of maximum displace-ment for each driver nonlinearity, which is of practicalinterest. To avoid any confusion with the performance-based method, these values are called displacement limits.The nonlinearity with the smallest value will in the endlimit the peak displacement of the driver. The parameter-based method also uses a threshold, which should bedefined consistent with the thresholds in the current Xmaxdefinition to provide comparable results.

4.1 Displacement Limits Due to DriverNonlinearities

The maximum voice-coil displacement is limited by atleast three factors:

1) Excessive decrease in mechanical compliance of themechanical suspension (caused mainly by the natural lim-its of the spider)

2) Voice-coil excursion capability (limited mainly byhitting the backplate)

3) Excessive, subjectively unpleasant signal distortionin the sound pressure output (depending on loudspeakernonlinearities, nature of excitation signal, and audible acu-ity of the listener).

These limiting factors may be represented by separatedisplacement limits:

• XC represents mechanical loading imposed to suspen-sion and tolerable distortion due to Cms(x) nonlinearity

• Xclip represents the moving range without mechanicalclipping caused by hitting the backplate and otherdefects

• XBl represents tolerable distortion due to Bl (x)nonlinearity

• XL represents tolerable distortion due to Le(x), L2(x), andR2(x) nonlinearities

• XD represents tolerable distortion due to Dopplernonlinearity.

4.1.1 Displacement Limit XC

The maximum displacement related to the criticalmechanical strain of suspension may be obtained from thenonlinear stiffness characteristic Kms(x) or from its coun-terpart, the compliance characteristic Cms(x). Clark [5] sug-gested to evaluate the variations of the differential stiffnessdFR(x)/dx, where FR(x) is the elastic restoring force of the



suspension. Following his proposal but using the regularsuspension parameter, the minimum compliance ratio

%minC XC

C x

0100

< <min

ms

msC

X x X C C

J

L

KK_

^

^N

P

OOi

h

h(7)

is the ratio of the minimum value of the compliance withinthe working range XC and the value at the rest positionx 0. XC is implicit in the equation and can be found inthe nonlinear Cms(x) characteristic by using a predefinedthreshold Cmin. The large signal identification imple-mented in the Klippel distortion analyzer [6] determinesthe safe range of operation automatically by comparingthe Cmin value with a user-defined protection limit Clim.This parameter is easy to use, and it has proven to be areliable measure for determining the critical mechanicalstrain affecting the suspension.

4.1.2 Displacement Limit Xclip

The maximum displacement due to mechanical clippingmay be derived from the geometry of the moving-coil assem-bly, and may be verified by practical experiments. In a well-designed loudspeaker, Xclip should always be higher than XCto avoid a mechanical damage of the voice-coil former.

4.1.3 Displacement Limit XBL

The maximum displacement XBL limited by excessivemotor distortion may be obtained from the nonlinear forcefactor characteristic Bl(x). The minimum force factor ratio

%minBl XBl

Bl x

0100

< <min Bl

X x X Bl Bl

J

L

KK_

^

^N

P

OOi

h

h(8)

is the ratio of the minimum force factor Bl(x) in the workingrange XBl referred to the Bl value at the rest position x 0. XBl is implicit in the equation and can be found in the non-linear Bl(x) characteristic after defining the threshold Blmin.

4.1.4 Displacement Limit XL

The electrical impedance Ze( f, x) of the driver above theresonance frequency depends on the frequency and thedisplacement of the coil. Fig. 12 shows the magnitude of

the electrical impedance versus frequency f for the threevoice-coil positions x 7, 0, 7 mm. The typical reso-nance is not visible for x 7 mm due to the clampingof the coil. The increase of the high-frequency impedancefor a negative displacement and the decrease for positivedisplacement are typical for drivers having no shortcutring or copper cap on the pole piece.

The complicated frequency characteristic is caused bythe parainductance of the coil and additional losses dueto eddy currents. This can be modeled by a lumped-parameter model comprising the electrical dc resistanceRe, the voice-coil resistance Le(x), the parallel elementsL2(x) and R2(x), as shown in Fig. 2. Since the inductancesand the resistance R2(x) depend on the total magnetic flux,it is assumed that

.L

L x

L

L x

R

R x

0 0 0

e

e

e

e

2

2

^

^

^

^

^

^

h

h

h

h

h

h(9)

The parameters Le(0), L2(0), and R2(0) at the rest positionx 0 may be estimated from the electrical impedancemeasured at small amplitudes [11]–[13]. The variation ofthe impedance versus displacement x is directly related tothe magnitude of the intermodulation distortion generatedin the current and in the radiated sound pressure output.Thus the displacement limit XL is defined implicitly by

,

, ,%maxZ X

Z f

Z x f Z f

0

0100

< <max

e

e eL

X x X2

2 2

L L

_

_

_ _

i

i

i i

(10)

which is the ratio of the maximum variation of the electri-cal impedance at frequency f2 within the working rangeXL < x < XL and the impedance at the rest position x 0.

To keep the parameter-based method consistent withthe performance-based method, the frequency f2 8.5fs iscoupled to the resonance frequency fs, and the impedanceat high frequencies can be approximated by

,Z x f R L x sR x L x s

R x L x s

e e e2 2

2 2 2

2 2 2._ ^

^ ^

^ î h

h h

h h(11)


Fig. 12. Electrical impedance of free-moving coil corresponding to rest position and position of maximum positive and negative dis-placements with blocked movement.

KLIPPEL

0

5

10

15

20

25

30

35

45

101 102 103 104

Ohm

frequency [Hz]

+ 7 mm

X = 0 mm

- 7 mm

[Ohm]

KLIPPEL PAPERS

where s2 2π f2 j with j 1 .

4.1.5 Displacement Limit XD

The peak displacement XD considering the audibility ofthe Doppler effect can be calculated analytically using thesimple equation

Xf

d770peak

2

2 (12)

presented by Beers and Belar [14], using the peak dis-placement Xpeak in millimeters, the second-order modu-lation distortion d2 in percent according to IEC 60268,and the frequency f2 of the modulated voice tone. Usingf2 8.5fs consistent with the performance-basedmethod leads to a displacement limit due to the Dopplereffect,

.X

f

d90 5D

s(13)

where XD is in millimeters and fs in hertz.

4.2 Practical Use1) Measure the small-signal parameters such as reso-

nance frequency fs, dc voice-coil resistance Re, resistanceR2(0), and inductance L2(0) at x 0.

2) Measure the nonlinear characteristics, complianceCms(x), force factor Bl(x), and inductance Le(x), versusdisplacement x by using a static, quasi-static, or dyna-mic method [5], [6], [15]–[19]. Listen for excessivedistortion and assign Xclip Xpeak in case of mechanicalclipping.

3) Determine the peak displacements XC, XBl, XL, andXD by using the nonlinear characteristics and thresholdsfor Cmin, Blmin, Zmax, and d.

4) State the displacement limits XC, XBl, XL, Xclip, andXD together with the thresholds Cmin, Blmin, Xmax, and dused.

5 DEFINITION OF THRESHOLDS

Both the peak displacement Xmax from the performance-based method and the displacement limits XC, XBl, XL, andXclip from the parameter-based method depend on thethresholds. These thresholds should consider the audibil-ity of the distortion components and the maximummechanical load, and they should lead to comparableresults in both methods.

5.1 AudibilityThe new performance-based method uses the old dis-

tortion threshold d 10% for the maximum harmonic dis-tortion in the sound pressure suggested by Small [1]. Thisvalue is also applied for the second- and third-order inter-modulation distortion. At the current time there are no bet-ter arguments for using other values. The audibility of thenonlinear distortion generated by loudspeakers dependson the following factors:

• Linear driver parameters (resonance frequency fs andloss factor Qts)

• Driver nonlinearities [Bl(x), Le(x), Cms(x), and Dopplereffect]

• System application (crossover frequency, type of enclosure)• Excitation signal (nature, bandwidth, spectral and tem-

poral complexity)• Audible acuity of a listener.

For example, a suspension nonlinearity produces distor-tions confined to frequencies about the resonance. On theother hand Bl(x) and Le(x) nonlinearities produce substan-tial intermodulation throughout the audio band, whichmight be tolerable in subwoofer applications.

Thus the subjective evaluation of the nonlinear distor-tion is a complex issue. Digital transducer modeling givesnew possibilities for combining subjective and objectiveinvestigations by operating the loudspeaker under normalconditions and using ordinary music or any other signal asstimulus. Auralization techniques [20] are the basis forsystematic listening tests, providing more reliable data inthe near future.

5.2 Relationship between ThresholdsThe thresholds Cmin, Blmin, Zmax, and d used in the

parameter-based approach should be consistent with thedistortion thresholds of the performance-based approach.Numerical techniques based on the loudspeaker modelallow simulating the sound pressure output for some typi-cal shapes of driver nonlinearities and calculating theparameter ratios Blmin, Cmin, and Zmax that correspond tothe distortion threshold d 10%, as shown in Table 4. Ashort and a long voice-coil overhang is simulated by apower series expansion of Bl(x) using a quadratic andfourth-order term, respectively, while neglecting anyasymmetries. A nonlinear compliance Cms(x) having aquadratic term represents a progressive spider. The fourth-


Table 4. Minimum parameter variation generating 10% distortion in the radiated sound pressure.

Example Nonlinear Parameter Parameter Threshold Distortion Threshold

Motor with equal-length configuration Bl(x) b0 b2x2 Blmin 82% d3 10%Motor with largecoil overhang Bl(x) b0 b4x4 Blmin 82% d3 10%Progressive spider Cms(x) c0 c2x2 Cmin 74% dt 10%Linear spider withlimiting surround Cms(x) c0 c4x

4 Cmin 77% dt 10%Asymmetry in suspension Cms(x) c0 c1x Cmin 78% dt 10%Typical inductance characteristics Le(x) l0 l1x Zmax 10% d2 10%


order term describes the symmetrical limitation of the sur-round. A severe asymmetry such as caused by a cup spi-der can be modeled by a power series of Cms(x) truncatedafter the linear term. A typical inductance characteristiccan be approximated by a linear power series expansion.

A second-order and a fourth-order nonlinearity produces10% distortion at similar values of the parameter variation[about 82% for Bl(x) and about 75% for Cms(x)]. However,a higher order nonlinearity will produce much less distor-tion at a lower displacement, X < Xmax, than a parabola-shaped curve. Thus the nonlinear parameters themselvesare required to predict the distortion versus amplitude, andto explain the benefit of a motor with a larger voice-coiloverhang over an equal-length configuration.

5.3 Acceptable Mechanical LoadThe threshold Cmin 75% producing dt 10% distor-

tion seems to be relatively high compared to the mechan-ical load admissible to most drivers. In the Klippel distor-tion analyzer system, a minimum compliance ratio Cmin 50% is used as a default protection parameter. Most com-mon suspension systems will withstand variations down toCmin 20% for some period of time without suffering anydamage.

It is also possible that manufacturers define admissible

thresholds that seem proper for their particular productsand specify these values as measurement conditions alongwith the displacement limits.

6 PRACTICAL EXAMPLES

Both the performance-based and the parameter-basedmethods will be applied to two real drivers to illustrateboth techniques. Driver A has an extremely long voice coilcoupled with a limited suspension. By contrast, driver Buses a short coil with a very linear suspension. The non-linear parameters are measured dynamically by using theKlippel distortion analyzer.

6.1 Driver AThe force factor Bl(x) in Fig. 13 remains almost con-

stant over the measured range, producing low modulationdistortion. Considering a limit of Blmin 82%, the admis-sible peak displacement XB is beyond 4 mm. Apparentlythe magnetic field geometry is symmetrical and the coil isin the optimum rest position.

However, the compliance Cms(x) in Fig. 14 has a asym-metrical characteristic, which becomes obvious whencomparing the regular curve Cms(x) with the mirror curveCms(x) presented as a dashed line. Considering a limitvalue of Cmin 75%, the admissible peak displacementXC is 2 mm. Due to the asymmetry, the suspension limitsthe excursion only with negative displacement.

Fig. 15 shows the asymmetric characteristic of theinductance Le(x), which is typical for a motor withoutshortcut ring or copper cap. Considering the resonancefrequency fs 49 Hz, a dc resistance Re 6.8 Ω, and thelimit value Zmax 10%, the admissible peak displacementXL exceeds the measured range of 4 mm.

The admissible peak displacement XD producing 10%modulation distortion is about 18 mm. Searching for theminimum between the separate peak displacements XB,XC, XL, and XD, clearly the suspension limits the maximumdisplacement to approximately 2 mm.

Using the new performance-based method, the totalharmonic distortion dt, the second- and third-order modu-lations are measured versus the peak displacement andpresented in Fig. 16. Since suspension is the limiting fac-


Fig. 14. Compliance Cms(x) of mechanical suspension versusvoice-coil displacement x of driver A.

Fig. 15. Inductance Le(x) versus voice-coil displacement x ofdriver A.

0,00

0,25

0,50

0,75

1,00

-4 -3 -2 -1 0 1 2 3 4

[mm/N]

<< Coil In x [mm] coill Out

Cms(x) Cms (-x)

0,00

0,25

0,50

0,75

1,00

-4 -3 -2 -1 0 1 2 3 4

[mH]

<< Coil In x [mm] coil Out >>

Fig. 13. Force factor Bl(x) versus voice-coil displacement x ofdriver A.

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

4,0

4,5

-4 -3 -2 -1 0 1 2 3 4

[N/A]

<< Coil In x [mm] coil Out >>

KLIPPEL PAPERS

tor, the total harmonic distortion dominates and exceedsthe 10% limit at Xmax 2.4 mm first. The second-ordermodulation distortion d2 caused by asymmetrical induc-tance Le(x), Doppler effect, and Bl asymmetry reaches the10% distortion level at 4 mm. The third-order distortion,which is directly related to the voice-coil height and thesymmetrical Bl variation, is far below 10% up to a 6-mmdisplacement. Table 5 summarizes the other distortionmeasurements determined at u1 u2 u10% 3.4 Vrms,giving a peak displacement of Xmax 2.4 mm.

The second-order harmonic distortion dh2,f1 8.7% ofthe bass tone f1 dominates the third-order harmonic dh3,f1 5.4% due to the substantial asymmetry of the suspension.The positive dc displacement generated by rectification ofthe bass tone shifts the coil in the positive direction, wherethe compliance is maximal. Thus improving the symmetryof the curve will give more Xmax. The second-order distor-tion d2 in sound pressure and d2,i in current are on thesame order of magnitude, indicating that the inductanceasymmetry is the physical source while the contributionsof Bl(x) asymmetry and Doppler effect are much smaller.The harmonic distortion of the voice tone reveals theeffect of nonlinearities that are related to voice-coil cur-rent or mechanical stress in the diaphragm. However, the

distortion measures dh2,f2, dh3,f2, and d3,i are as usualbelow 1%, which can be neglected when compared to thedominant nonlinearities.

The manufacturer may state the following parameters:

. @ % , < %

@ %

> @ %

> @ %

@ % .

mm

mm

mm

mm

mm

X d d d

X C

X Bl

X Z

X d

2 4 10 10

2 75

4 82

4 10

18 10

max

min

min

max

t

D

C

B

L

2 3

2

_ i

6.2 Driver BFor loudspeaker B the following parameters are measured:

. @ % , < %

> @ %

. @ %

> @ %

. @ % .

mm

mm

mm

mm

mm

X d d d

X C

X Bl

X Z

X d

2 1 10 10

4 75

1 8 82

4 10

20 5 10

max

min

min

max

D

C

B

L

3 2 3

2

_ i

In contrast to loudspeaker A, the third-order intermodu-lation distortion limits the peak displacement Xmax. It cor-responds to the dominant force factor nonlinearity causingthe lowest displacement limit XB 1.8 mm. The suspen-sion, the inductance, and the Doppler effect give manymore excursion capabilities, which cannot be used.

The force factor Bl(x), as displayed in Fig. 17, reveals ashort voice coil with low overhang. Such loudspeakertypes are sensitive to an offset of the coil. Loudspeaker Bhas an offset of about 0.8 mm, causing a distinct asymme-try in the Bl(x) characteristic. Using Blmin 82%, themotor limits the peak displacement to XB 1.8 mm.

The compliance Cms(x) in Fig. 18 reveals a very linearsuspension. Cms(x) stays in the measured range aboveCmin 75% (XC > 4 mm).

Similar to driver A, the inductance Le(x) in Fig. 19 hasa typical shape with a maximum at the negative displace-ment. However, the absolute value of inductance is lessthan one-third of driver A. Considering the resonance fre-


Fig. 17. Force factor Bl(x) versus voice-coil displacement x ofdriver B.

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

-4 -3 -2 -1 -0 1 2 3 4

[N/A]

<< Coil in x [mm] coil out >>

Bl(x) Bl (-x)

Fig. 16. Total harmonic distortion dt (–––), second-order modu-lation d2 (– – –), and third-order modulation distortion d3 (g)versus peak displacement x of driver A.

0 1 2 3 4 5 60

5

10

15

20

25

30

35

40

Harmonic

3rd

modulation

2nd

modulation

Peak Displacement [mm]

%

Table 5. Results of performance-based method.

Loudspeaker A Loudspeaker B

f1 fs 50 Hz 44 Hzf2 8f1 400 Hz 360 Hzu @ d 10% 3.4 Vrms 1.87 VrmsXdc 0.31 mm 0.03 mmdh2,f1 8.7% 4.3%dh3,f1 5.4% 3.6%dt 10% 5.7%d2 7.9% 8.4%d3 1.5% 10%d2,i 7.7% 3.3%d3,i 0.5% 1.2%dh2,f 2 0.58% 0.4%dh3,f 2 0.34% 0.4%Xmax @ d 10% 2.4 mm 2.1 mm


quency fs 44 Hz, the dc resistance Re 3.8 Ω, and thelimit value of Zmax 10%, the peak displacement XLexceeds 4 mm. The Doppler effect limits the peak dis-placement to XD 20.5 mm.

Despite the asymmetry of the Bl(x) curve, which is notconsidered in the definition of the threshold Blmin, theresults of the parameter-based approach agree well withthe results of the performance-based method. Fig. 20shows the distortions dt, d2, and d3 versus the peak dis-placement Xpeak measured with a two-tone signal. Above2-mm displacement the third-order modulation distortiond3 becomes dominant and reaches the threshold of 10% atXmax 2.1 mm. Although the parameter-based methoddoes not consider the shape of the nonlinear characteristicbut only the crossing point at 82% variation, there is asimilar result for XBl.

According to Table 3 significant values of d3 show thatthe symmetrical Bl variation due to a short coil limitsXmax. The total harmonic distortion dt would allow muchhigher peak displacement but at Xmax 7 mm any voicetone will produce about d3 70% modulation distortion.The second-order modulation distortion d2 is twice that ofthe harmonics. An additional measurement of the second-order modulation d2,i in the input current reveals more infor-mation about the source of the distortion. Since d2 6.8%in sound pressure is significantly higher than d2,i in current,the force factor asymmetry makes a significant contribution.

7 SUMMARY

A critical review of the numerical simulation on a ficti-tious loudspeaker and practical measurements on realloudspeakers show that the method in AES2-1984 doesnot provide a clear and useful definition of Xmax. This ismainly caused by some ambiguities in the wording and,more importantly, by using assumptions that are not validin theory and practice. Clearly, the measurement of har-monic distortion is not sufficient for assessing all impor-tant aspects of large-signal performance. Nonlinearitiesinherent in transducers such as force factor Bl(x), induc-tance Le(x), and Doppler effect produce significant modu-lation distortion. The current IEC 60268 provides all ofthe methods required for assessing these kinds of distor-tion and for defining Xmax more clearly and reliably. Thenew definition is based on a two-tone measurement thatcan be accomplished with straightforward equipment(microphone, sinusoidal generator, FFT analyzer). Theresulting distortion measurements are also valuable fortransducer diagnostics to improve the driver design orselect the optimum driver for a particular application.

The second part of the paper addressed an alternativemethod for assessing separate displacement limits whichclosely related with the nonlinear driver parameters. Incontrast to the performance-based approach, which meas-ures some effects of the nonlinear systems, the parameter-based method refers to the physical causes. The nonlinearcurves and other linear parameters are summarized in afew numbers describing the limiting effect of each drivernonlinearity [Bl(x), Cms(x), Le(x), and Doppler]. This is asubstantial data reduction, where some particularities ofthe nonlinear curves are neglected. Despite the simplica-tions made in both methods, the minimum value of thedisplacement limits XBl, XL, XC, and XD is comparable tothe peak displacement Xmax derived from distortion meas-urements. Numerical tools are available for transformingand comparing the results of both methods.

It is a good idea to state Xmax based on distortion meas-urements because it can easily be verified by simple


Fig. 19. Inductance Le(x) versus voice-coil displacement x ofdriver B.

Fig. 20. Total harmonic distortion dt (–––), second-order modu-lation d2 (– – –), and third-order modulation distortion d3 (g)versus peak displacement x of driver B.

0,00

0,05

0,10

0,15

0,20

0,25

0,30

-4 -3 -2 -1 -0 1 2 3 4

[mH]


0 1 2 3 4 5 6 70

10

20

30

40

50

60

70

80

%

Peak Displacement [mm]

2nd

modulation

3rd

modulation

Harmonic

Fig. 18. Compliance Cms(x) of mechanical suspension versusvoice-coil displacement x of driver B.

0,0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

-4 -3 -2 -1 -0 1 2 3 4

[mm/N]


Cms(x) Cms (-x)

KLIPPEL PAPERS

equipment. The displacement limits XB, XC, XL, and XDgive additional information about the driver, which areimportant for the system design but require special meas-uring equipment. The main target of this paper was thedevelopment of a framework for assessing the peak dis-placement more reliably. Both methods presented are stillflexible by changing the thresholds used. This can beaccomplished easily by an agreement between driver andsystem manufacturer, considering special requirements.More long-term testing of loudspeakers with instanta-neous parameter monitoring and systematic listening testsusing a digital transducer model will provide more infor-mation about admissible load and impact on sound qual-ity. In the meantime the “traditional” threshold of 10%distortion is a good starting point.

8 ACKNOWLEDGMENT

This paper is the result of many fertile discussionsbetween the author and Alexander Voishvillo, DavidClark, and others.

9 REFERENCES

[1] R. H. Small, “Closed-Box Loudspeaker Systems,Part I: Analysis,” J. Audio Eng. Soc., vol. 20, pp. 798–808(1972 Dec.).

[2] M. Gander, “Dynamic Linearity and PowerCompression in Moving-Coil Loudspeakers,” J. AudioEng. Soc., vol. 34, pp. 627–646 (1986 Sept.).

[3] AES2-1984, “AES Recommended Practice Specifi-cation of Loudspeaker Components Used in ProfessionalAudio and Sound Reinforcement,” Audio EngineeringSociety, New York (r1997).

[4] D. Clark, “Amplitude Modulation Method forMeasuring Linear Excursion of Loudspeakers,” presentedat the 89th Convention of the Audio Engineering Society,J. Audio Eng. Soc. (Abstracts), vol. 38, p. 876 (1990 Nov.),preprint 2986.

[5] D. Clark, “Precision Measurement of LoudspeakerParameters,” J. Audio Eng. Soc., vol. 45, pp. 129–141(1997 Mar.).

[6] W. Klippel, “Distortion Analyzer––A New Tool forAssessing and Improving Electrodynamic Transducers,”presented at the 108th Convention of the Audio Engin-eering Society, J. Audio Eng. Soc. (Abstracts), vol. 48, p.353 (2000 Apr.), preprint 5109.

[7] IEC 60268-5, “Sound System Equipment. Part 5:Loudspeakers,” International Electrotechnical Commis-sion, Geneva, Switzerland (1989).

[8] W. Klippel, “Prediction of Speaker Performanceat High Amplitudes,” presented at the 111th Conven-tion of the Audio Engineering Society, J. Audio Eng.Soc. (Abstracts), vol. 49, p. 1216 (2001 Dec.), preprint5418.

[9] E. Czerwinski, A. Voishvillo, S. Alexandrov, and A.Terekhov, “Multitone Testing of Sound System Compon-ents––Some Results and Conclusions, Part 1: History andTheory,” J. Audio Eng. Soc., vol. 49, pp. 1011–1048 (2001Nov.).

[10] D. B. Keele, Jr., “Low-Frequency LoudspeakerAssessment by Nearfield Sound-Pressure Measurements,”J. Audio Eng. Soc., vol. 22, pp. 154–162 (1974 Apr.).

[11] W. M. Leach, Jr., “Loudspeaker Voice-Coil Induct-ance Losses: Circuit Models, Parameter Estimation, andEffect on Frequency Response,” J. Audio Eng. Soc., vol.50, pp. 442–450 (2002 June).

[12] J. R. Wright, “An Empirical Model for Loud-speaker Motor Impedance,” J. Audio Eng. Soc., vol. 38,pp. 749–754 (1990 Oct.).

[13] W. Klippel and U. Seidel, “Fast and Accurate Mea-surement of Linear Transducer Parameters,” presented atthe 110th Convention of the Audio Engineering Society, J.Audio Eng. Soc. (Abstracts), vol. 49, p. 526 (2001 June),preprint 5308.

[14] G. L. Beers and H. Belar, “Frequency-ModulationDistortion in Loudspeakers,” J. Audio Eng. Soc., vol. 29,pp. 320–326 (1981 May).

[15] K. Satoh, H. Takewa, and M. Iwasa, “TheMeasuring Method of Dynamic Force-to-DisplacementCharacteristics for Loudspeaker Suspension System andDriving Force,” presented at the 107th Convention ofthe Audio Engineering Society, J. Audio Eng.Soc. (Abstracts), vol. 47, p. 1000 (1999 Nov.), preprint5023.

[16] R. J. Mihelich, “Loudspeaker Nonlinear ParameterEstimation: An Optimization Method,” presented at the111th Convention of the Audio Engineering Society, J.Audio Eng. Soc. (Abstracts), vol. 49, p. 1216 (2001 Dec.),preprint 5419.

[17] M. Knudsen, “Loudspeaker Modelling and Para-meter Estimation,” presented at the 100th Convention ofthe Audio Engineering Society, J. Audio Eng. Soc.(Abstracts), vol. 44, p. 633 (1996 July/Aug.), preprint4285.

[18] E. S. Olsen, “Measurement of Mechanical Para-meter Nonlinearities of Electrodynamic Loudspeakers,”presented at the 98th Convention of the Audio Engin-eering Society, J. Audio Eng. Soc. (Abstracts), vol. 43, p.400 (1995 May), preprint 4000.

[19] J. Scott, J. Kelly, and G. Leembruggen, “NewMethod of Characterizing Drive Linearity,” J. Audio Eng.Soc., vol. 44, pp. 258–265 (1996 Apr.).

[20] W. Klippel, “Speaker Auralization––SubjectiveEvaluation of Nonlinear Distortion,” presented at the110th Convention of the Audio Engineering Society, J.Audio Eng. Soc. (Abstracts), vol. 49, pp. 526, 527 (2001June), preprint 5310.

APPENDIX

The differential equation describing the one-dimensionaloperation of a direct-radiating loudspeaker written in thegeneral state-space form is

z z z z ua b o ^ ^h h (14)

where z(t) [x, dx/dt, i, i2]T is the state vector of the sys-

tem and u is the voltage at the terminals of the voltage-driven loudspeaker. The matrix



and the vector

zL x

b 0 01

0e

T

^^

hh

R

T

SSS

V

X

WWW

(16) comprise the lumped parameters depending on the statevector z.


i

i

x

x

d

d

zC x M M

R

L x

Bl x L x

x

L x

L x

i

M

Bl x L x

L x

R

L x

R x

L x

R x i i

L x

R x

a

0

1

0

0

1 0

2

2

0

0

ms ms ms

ms

e

ms

e

e

e

2

2

2

2

2

2 2

2

2

l

l

^

^

^

^ ^

^

^

^ ^

^

^

^

^

^ _

^

^

h

h

h

h h

h

h

h h

h

h

h

h

h i

h

h

R

T

SSSSSSSSSSSS

V

X

WWWWWWWWWWWW

(15)

THE AUTHOR

Wolfgang Klippel was born in Halle, Germany, in 1957.He studied electrical engineering at the University ofTechnology of Dresden, Dresden, Germany, from whichhe received a Bachelor’s degree in communication engi-neering and electroacoustics in 1981 and a Master’s degreein the field of speech recognition in 1982.

After graduating, Dr. Klippel joined the loudspeakerresearch group of VEB Nachrichtenelektronik Leipzig. Hewas engaged in the research of transducer modeling,acoustic measurement, and psychoacoustics and developednew tools for loudspeaker design and quality assessment.In 1987 he received a Doctor-Engineer degree in technicalacoustics. His thesis was on the “MultidimensionalRelationship between Subjective Listening Impression andObjective Loudspeaker Parameters.”

He continued research on the audibility of the nonlinearloudspeaker distortion at the Institute of Technical Acoustics,Dresden, there he started modeling the nonlinear mecha-nisms in woofer and horn loudspeakers and developed novelcontrol systems dedicated to loudspeakers. In 1992 he sum-marized the results of his research in a thesis, “The NonlinearTransfer Characteristic of Electroacoustic Transducer,” sub-mitted for a Certificate of Habilitation.

After spending a postdoctoral year at the AudioResearch Group in Waterloo, Canada, and working atHarman International, Northridge, CA, he moved back toDresden in 1995, where he became a freelance consultantengineer. In 1997 he founded Klippel GmbH, whichdevelops novel kinds of control and measurements sys-tems dedicated to loudspeakers and other transducers.

PAPERS

0 INTRODUCTION

A loudspeaker installed in a room acts as a coupled sys-tem where the room properties typically dominate the rateof energy decay. At high frequencies, typically above afew hundred hertz, passive methods of controlling the rateand properties of this energy decay are straightforwardand well established. Individual strong reflections are bro-ken up by diffusing elements in the room or trapped inabsorbers. The resulting energy decay is controlled to adesired level by introducing the necessary amount ofabsorbance in the acoustical space. This is generally feasi-ble as long as the wavelength of sound is small comparedto the dimensions of the space.

As we move toward low frequencies, passive means ofcontrolling the speed of reverberant decay become moredifficult to use because the physical size of the necessaryabsorbers increases and may become prohibitively largecompared to the volume of the listening space, or

absorbers have to be made narrow-band. Consequently thecost of passive control of reverberant decay greatlyincreases at low frequencies. Methods for optimizing theresponse at a listening position by finding suitable loca-tions for loudspeakers have been proposed [1] but cannotfully solve the problem of controlling modal decay.Because of these reasons, and because active controlbecomes technically feasible when the wavelength ofsound becomes long relative to the room size, resulting ina less diffuse sound field in the room [2]–[6], there hasbeen an increasing interest in methods of active sound-field control at low frequencies.

Mode resonances in a room can be audible because theymodify the magnitude response of the primary sound or,when the primary sound ends, because they are no longermasked by the primary sound [7], [8]. The detection of amode resonance appears to be very dependent on the sig-nal content. Olive et al. report detection thresholds for res-onances for both continuous broad-band sound and tran-sient discontinuous sound, showing that low-Q resonancesare more readily audible with continuous signals contain-ing a broad frequency spectrum whereas high-Q reso-nances become more audible with transient discontinuous


Modal Equalization of Loudspeaker–RoomResponses at Low Frequencies*

AKI MÄKIVIRTA,1 AES Member, POJU ANTSALO,2 MATTI KARJALAINEN,2 AES Fellow,

AND VESA VÄLIMÄKI,2 AES Member

1Genelec Oy, FIN-74100 lisalmi, Finland2Helsinki University of Technology, Laboratory of Acoustics and Audio Signal Processing,

FIN 02015 HUT, Espoo, Finland

The control of excessively long decays in a listening room with strong low-frequency modesis problematic, expensive, and sometimes impossible with conventional passive means. Asystematic methodology is presented to design active modal equalization able to selectivelyreduce the mode decay rate of a loudspeaker–room system at low frequencies in the vicinityof a sound engineer’s listening location. Modal equalization is able to increase the rate ofinitial sound decay at mode frequencies, and can be used with conventional magnitude equal-ization to optimize the reproduced sound quality. Two methods of implementing active modalequalization are proposed. The first modifies the primary sound such that the mode decay ratesare controlled. The second uses separate secondary radiators and controls the mode decayswith additional sound fed into the secondary radiators. Case studies are presented of imple-menting active modal control according to the first method.

* Manuscript received 2002 July 15; revised 2003 February14. Parts of this paper were presented at the 111th Convention ofthe Audio Engineering Society, New York, 2001 November30–December 3.

PAPERS MODAL EQUALIZATION OF LOUDSPEAKER–ROOM RESPONSES

signals [8]. The antiresonances (notches) are as audible asthe resonances for low Q values. The audibility of antires-onances reduces dramatically for wide-band continuoussignals when the Q value becomes high [8]. The detect-ability of resonances reduces approximately 3 dB for eachdoubling of the Q value [7], [8] and low-Q resonances aremore readily heard with zero or minimal time delay rela-tive to the direct sound [7]. The duration of the reverber-ant decay in itself appears an unreliable indicator of theaudibility of the resonance [7], as audibility seems to bedetermined more by the frequency-domain characteristicsof the resonance.

Traditional magnitude response equalization cannotguarantee the control of the modal decay rate becausesuch equalizers typically have a system transfer functionof significantly lower order than the loudspeaker–roomsystem.

A system having a unity transfer function (no change inmagnitude or phase of the transmitted sound) does notmodify the transmitted signal. By using a DSP filter it isin theory possible to design a filter that equalizes a loud-speaker–room system to approximate a unity transferfunction at a single location in the listening space within alimited frequency range. Such filtering may shorten themodal decay times too aggressively and may also lead toa system overload by boosting notched frequency areas,and is therefore often impractical for equalizing realrooms. Also, at higher frequencies the unity transfer func-tion target can lead to a highly local correction in the roomspace with little practical value. High-order equalizershave often been found to actually decrease frequencyresponse flatness for points other than the design targetlocation.

Contrary to the unity transfer function equalization tar-get, the principle of modal equalization is rather to bal-ance the decay rate of low-frequency modes to correspondto the reverberation time at mid and high frequencies inorder to minimize the audibility of modal resonances atlow frequencies. We also assume that once modal equal-ization has been performed, a conventional magnituderesponse equalizer (for example, an equalizer designed onone-third-octave smoothed magnitude response data) isapplied to improve the subjective magnitude response bal-ance at low frequencies and between the low-frequencyarea and higher frequencies. Mode equalization thereforebecomes an additional tool to improve the audio qualityin rooms beyond what is attainable using conventionalmagnitude equalization alone.

In this paper we present methods to actively control low-frequency reverberation. We will first present the conceptand two principal methods of applying modal equalization.A target for mode decay time versus frequency will be dis-cussed based on existing recommendations for high-qual-ity audio monitoring rooms. Methods to identify and para-meterize modes in an impulse response are introduced. Themode equalizer design for an individual mode is discussedwith examples. Several case studies of both syntheticmodes and modes of real rooms are presented. Finally,aspects about the synthesis of infinite impulse response(IIR) mode equalizer filters are discussed.

1 THE CONCEPT OF MODAL EQUALIZATION

To be useful, modal equalization must affect a changein the room decay characteristics, at least within an areasufficiently large to allow binaural monitoring withoutradical changes in the perceived reverberation or col-oration of sound as a result of small movements of headlocation.

The present work is restricted to frequencies below 200Hz and environments where the sound wavelength relativeto the room dimensions is not very small, that is, smallrooms such as monitoring rooms instead of halls. We arenot aiming at global control of the sound field in a room,but we try to introduce a change at the primary listeningposition, typically in the vicinity of a sound engineer’s lis-tening location.

These limitations lead to a problem formulation wherethe modal behavior of the listening space can be modeledby a distinct number of modes such that they can be indi-vidually controlled. Each mode is modeled by an expo-nential decay function,

.sin ω φeh t A t τm m m m

t m^ _h i (1)

Here Am is the initial envelope amplitude of the decayingsinusoid, τm is a coefficient that denotes the decay rate, ωmis the angular frequency of the mode, and φm is the initialphase of oscillation.

We define modal equalization as a process that canmodify the rate of modal decay. It can be viewed as a spe-cial case of parametric equalization, operating individu-ally on selected modes in a room. Modal equalization canbe applied in a room by modifying the primary soundusing a filter or by introducing in the room one or moresecondary sound sources emitting a correcting sound.

A mode resonance is represented in the z-domain trans-fer function as a pole pair with pole radius r and poleangle θ,

.H zr r1 1

1

θ θm j j1 1 e ez z^

` `

hj j

(2)

The closer a pole pair is to the unit circle, the longer is thedecay time of a mode. To shorten the decay time, the res-onance Q value must be reduced by moving poles towardthe origin. Therefore modal equalization can also beviewed as a process of moving the locations of room trans-fer function poles.

Mode decay time modification can be implemented inseveral ways––either the sound going into a room throughthe primary radiator is modified or additional sound isintroduced in the room with one or more secondary radia-tors to interact with the primary sound. The first methodhas the advantage that the transfer function from a soundsource to a listening position does not affect mode equal-ization. In the second case differing locations of primaryand secondary radiators lead to different transfer functionsfrom the sound source to the listening position, and thismust be considered when calculating a corrective filter.We will now discuss these two cases in more detail, draw-


MÄKIVIRTA ET AL. PAPERS

ing some conclusions on necessary conditions for controlin both cases.

1.1 Type I Modal EqualizationType I implementation (Fig. 1) modifies the audio sig-

nal fed into the primary loudspeaker to compensate forroom modes. The total transfer function from the primaryradiator to the listening position represented in the zdomain is

H z G z H z m^ ^ ^h h h (3)

where G(z) is the transfer function of the primary radiatorfrom the electric input to the acoustic output and Hm(z) B(z)/A(z) is the transfer function of the path from the pri-mary radiator to the listening position. The primary radia-tor has essentially a flat magnitude response and a smalldelay in our frequency band of interest, and can thereforebe neglected in the following discussion,

.G z 1^ h (4)

We now design a pole–zero filter Hc(z) having a zero pairat each identified pole location of the mode resonances inHm(z). This cancels out the existing room response polepairs in A(z), replacing them with new pole pairs A(z) setto locations producing the desired decay time in the mod-ified transfer function Hm(z),

.H H mH z z zA z

A z

A z

B z

A z

B zc ml

l l^ ^ ^

^

^

^

^

^

^h h h

h

h

h

h

h

h(5)

This leads to a correcting filter,

.H zA z

A zc

l^

^

^h

h

h(6)

The new pole pair A(z) is chosen on the same resonantfrequency but closer to the origin, thereby effecting a res-onance with a decreased Q value. In this way the moderesonance poles appear to have been moved toward theorigin, and the Q value of the mode has been decreased.The coefficient sensitivity of this approach will be dis-cussed later with example designs.

1.2 Type II Modal EqualizationThe type II method uses a secondary loudspeaker at an

appropriate position in the room to radiate sound that inter-acts with the sound field produced by the primary loud-speakers. The loudspeakers are assumed to be identical in thetreatment that follows but this does not need to be the case in

practical implementations. Loudspeakers can be renderedsimilar by equalization for the purposes of modal equal-ization. The transfer function for the primary radiator isHm(z), and for the secondary radiator it is H1(z). Theacoustical summation in the room (Fig. 2) produces amodified frequency response Hm(z) with the desired decaycharacteristics,

H H H m .H zA z

B zz zm c 1l

l^

^

^^ ^h

h

hh h (7)

We can solve for a correcting filter Hc(z), where Hm(z) andHm(z) differ by modified pole radii,

H

m

.

HH z

z

H z H z

B z

A z

A z

B z

A z

A z A z

zA z

B z

cm

1

1

1

11

1

l

l

l

^^

^ ^

^

^

^

^

^

^ ^

^^

^

hh

h h

h

h

h

h

h

h h

hh

h

(8)

(9)

Note that if the primary and secondary radiators are thesame source, Eq. (8) reduces into a parallel formulation ofa cascaded correction filter, equivalent to the type I method,

H Hm .H z z z1 m cl ^ ^ ^h h h8 B (10)

A necessary but not sufficient condition for a solution toexist is that the secondary radiator can produce sound lev-els at the listening location in frequencies where the pri-mary radiator can, within the frequency band of interest,

! !, .forH f H f0 0m1_ _i i (11)

At low frequencies, where the size of a radiator becomessmall relative to the wavelength, it is possible for a radia-tor to be located such that at some frequency it does notcouple well into the room. At such frequencies the condi-tion of Eq. (11) may not be fulfilled, and a secondary radi-ator placed in such a location will not be able to affectmodal equalization. Because of this it may be advanta-geous to have multiple secondary radiators in the room. Inthe case of multiple secondary radiators, Eq. (7) is modi-fied to the form

H H HmH z z z z , ,m c n nn

N

11

!l ^ ^ ^ ^h h h h (12)


Fig. 1. Type I modal equalization using the primary sound source. Fig. 2. Type II modal equalization using a secondary radiator.

Hm

G

Hc

room

listeningposition

primarysource

primary source

listeningposition

Hm

Hc

roomsecondary sourceH1


where N is the number of secondary radiators.After the decay times of individual modes have been

equalized in this way, the magnitude response of theresulting system may be corrected to achieve a subjec-tively flat overall response. This correction can be imple-mented with several well-known magnitude response equa-lization methods, and is not addressed further in this study.

We will now discuss the identification and parameteri-zation of modes, and review some case examples of apply-ing modal equalization to various synthetic and realrooms, using the type I mode equalization method. Thetype II method using one or more secondary radiators willbe left for future study.

2 TARGET OF MODAL EQUALIZATION

After the in-situ impulse response at the primary listen-ing position has been measured using one of the standardtechniques for room response measurements capable ofproducing an impulse response estimate, the process ofmodal equalization starts by defining a target for equal-ization. This target describes the desired maximum decaytime at the frequencies of interest. There are severalrational ways to set this target.

Current recommendations for rooms intended for criticallistening [9]–[11] propose a requirement for the averagereverberation time Tm in seconds for midfrequencies (200Hz to 4 kHz) that depends on the volume V of the room,

. .TV

V0 25

/

m0

1 3J

L

KK

N

P

OO (13)

The reference room volume V0 of 100 m3 yields a rever-beration time of 0.25 s. An increase in reverberation timeat low frequencies is typical, particularly in rooms wherethe passive control of the reverberation time by absorptionis compromised. These rooms are likely to have isolatedmodes with long decay times, particularly at low frequen-cies. Recommendations [10] and [11] allow the reverber-ation time to increase linearly below 200 Hz by 0.3 s asthe frequency decreases to 63 Hz, with a maximum rela-tive change of 25% between adjacent one-third-octavebands. Below 63 Hz there is no requirement. Theserequirements are motivated by the goal to achieve a natu-ral sounding environment for monitoring [11].

Along these lines of reasoning, one way to define thedecay-time target is to define it relative to the midfre-quency decay time (Fig. 3). To comply with [10], [11] wecan define the target decay time relative to the mean T60 inmidfrequencies (500 Hz to 2 kHz), increasing linearly (ona log frequency scale) by 0.2 s as the frequency decreasesfrom 300 Hz down to 50 Hz.

There are also other approaches to setting the decay-time target. Because the aim of modal equalization is tominimize the audibility of mode resonances, it is rationalto aim at reducing the mode decay time down to the gen-eral decay time within the frequency range of interest. Oneway of doing so is to look at the statistical distribution ofthe decay time in the frequency band of interest. A per-centile value of the decay-time distribution can be chosen

as the target decay time. This may be a suitable choice toreduce the audibility of individual mode frequencies inrooms where it is difficult to achieve the low decay timesproposed in recommendations.

3 MODE IDENTIFICATION AND PARAMETERESTIMATION

The next task is to find and identify the modes that needto be equalized. They can be identified in the frequencyresponse by assuming that a mode produces a levelincrease in the magnitude response at a mode resonance,or by inspecting directly the measured short-term Fouriertransform as frequencies of prolonged decay time. We willdiscuss both methods of identifying modes.

One identification method is to search for magnituderesponse level increases produced by modes. A method ofdoing this is to note within the frequency range of interest(in our case f < 200 Hz) where the magnitude responseexceeds an average midfrequency magnitude responselevel (500 Hz to 2 kHz). Then local maxima above thismidfrequency reference level are noted and considered aspotential mode frequencies.

AR and ARMA modeling methods can be used to iden-tify resonance frequencies in order to find the resonanceswith the largest radii instead of directly inspecting themagnitude response data [12]. Because the selection of aproper model order for this task can be problematic, adetection function has been proposed based on the factthat at a modal resonance there is concurrently a highvalue of gain H(Ω) and a rapid change in the phase [12],

,max argΩ Ω ΩG H D H0^ ^ ^`h h hj9 C' 1 (14)

where H(Ω) is the Fourier transform of the measuredresponse, Ω is the normalized angular frequency, and D isa frequency-domain differentiation operator. A positivepeak in this detection function may indicate a mode thatneeds equalization.

One identification method is to inspect short-termFourier transform data decay profiles directly. A decayprofile PL(ω) describes at frequency ω the time t0 from theimpulse response maximum level after which the short-term Fourier transform bin level has decreased perma-nently below a profile detection level L,

: , < , > , .ω ω ω ωP t H t L t t L 0 0 0^ ^h h (15)


Fig. 3. Principle of setting reverberation time target relative tomeasured midfrequency octave-band reverberation time.

Frequency

target

measured

Rev

erbe

ratio

n tim

e


A mode detection function C(ω) is constructed as aweighted sum of k decay profiles in the Fourier transformdata detected at several profile levels L(k),

minω ω ωC a P P 1 ( ) ( )kk

N

L k L k

p

1

!^ ^ âh h h k8 B* 4

(16)

where the possible weight for each decay profile isdenoted by ak. Decay-time differences can be enhanced bysetting the exponent p > 1. This mode detection functionaccumulates value at frequencies with long decay times,typically having the maximum at the longest decay time.A sample decay profile and mode detection function willbe depicted later in the case examples.

3.1 Mode ParametersAfter finding the frequency of a possible mode with any

of the methods described, we must parameterize the modein order to design a mode equalizer filter. The parametersthat describe a mode are the decay rate coefficient τm,angular frequency ωm, and initial phase φm. Assuming thewavelength of sound within the frequency range of inter-est to be large relative to the room size, the initial phase istypically close to zero, and is assumed to have the valuezero henceforth.

Mode parameters are estimated using short-termFourier transforms of the transfer function impulseresponse. When certain time windowing is used, this isalso called the cumulative spectral decay [13]. A plot ofthe short-term Fourier transform data is frequentlyreferred to as a waterfall plot.

To determine the exact center frequency of a mode, weexploit a special property of windowed Fourier trans-forms. It can be shown that the spectral peak of aGaussian-windowed stationary sinusoid calculated using aFourier transform has the form of a parabolic function[14]. Therefore the precise center frequency of a mode canbe calculated by fitting second-order parabolic function,

G f af bf c 2_ i (17)

into the three Fourier transform bin values around a localmagnitude maximum indicated by the mode detectionmethods described earlier. The frequency where the second-order function derivative assumes the value zero is taken asthe center frequency of the mode,

.f

G ff

a

b0

2 &

2

2 _ i(18)

When time windowing is used, this fitting techniqueallows us to determine the mode frequency more preciselythan the frequency bin spacing of a Fourier transform. A20-fold improvement has been demonstrated in compari-son to raw fast Fourier transform bin spacing by using theHamming window [15].

After determining the mode frequency, the pole radiusof the mode resonance model must be determined. Thisestimate can be based on one of two parameters, the Qvalue of the steady-state resonance or the actual measure-

ment of the decay time T60. While the Q value can be esti-mated for isolated modes, it may be difficult or impossibleto define it for modes closely spaced in frequency.Because of this we use the decay time measurable directlyin short-term Fourier transform data to determine themode pole radius. The method employed in this work [16]provides a direct estimate of the decay rate coefficient τ.This then enables the calculation of T60 to obtain a repre-sentation of the decay time in a form readily related to theconcept of reverberation time,

..ln

τ τT

110

6 908 60

3 .` j (19)

The decay time for each detected potential room mode iscalculated by fitting an exponential decay noise modelinto time-series data formed by a particular frequency binof consecutive short-term Fourier transforms. The modedecay is modeled by an exponentially decaying sinusoid[Eq. (1) is reproduced here for convenience],

sin ω φeh t A t τm m m m

t m^ _h i (20)

where Am is the initial envelope amplitude of the decayingsinusoid, τm is a coefficient defining the decay rate, ωm isthe angular frequency of the mode, and φm is the initialphase of mode oscillation. We assume that this decay iscorrupted in practical measurements by an amount ofnoise nb(t),

n t A n tb n^ ^h h (21)

and that this noise is uncorrelated with the decay.Statistically the decay envelope of this signal is

.ea t A A τm n

t2 2 2 m^ h (22)

The values of Am, τm, and An are found by least-squaresfitting this model to the measured time series of short-termFourier transform bin values. For further details of thismethod, the reader is referred to [16]. A sufficientdynamic measurement range is required to allow reliabledetection of the room mode parameters, although theleast-squares fitting method used has been shown to beresilient to rather high noise levels. Noise level estimatesby using the least-squares fitting method across the fre-quency range provide a measurement of the frequency-dependent noise level A( f ), and this information can beused later to check data validity.

An alternative method employing ARMA modeling hasbeen proposed in [12], enabling the direct determinationof the pole model coefficients instead of going through theintermediate stage of determining first a center frequencyand a decay rate or resonance Q value.

In practice some further error checking of the identifiedmodes can be useful before moving onward to designmode equalizers in order to discard obvious measurementartifacts and not consider them as modes to be equalized.A candidate mode may have to be rejected if the noiselevel estimated at that mode frequency is too high, imply-ing an insufficient signal-to-noise ratio for reliable meas-urement. Also, candidate modes that show unrealistically



slow decay or no decay at all can be rejected if they rep-resent technical problems in the measurement, such asmains hum, ventilation noise in the room, or other unre-lated stationary error, and not any true mode decay. Theserejections can be confirmed by inspecting the short-termFourier transform data and with additional measurements.

3.2 Discrete-Time Representation of a ModeTo obtain a discrete-time representation of a mode, con-

sider now a second-order all-pole transfer function havingpole radius r and pole angle θ,

.cos θ

H zr r

r z r z

1 1

1

1 2

1

j j 1 1

1 2 2

e ez z^

` `

hj j

(23)

Taking the inverse z transform yields the impulse responseof this system as

sin

sin

θ

θh n

r nu n

1

n

^^

^hh

h8 B

(24)

where u(n) is a unit step function.The envelope of this sequence is determined by the term

rn. To obtain a decay rate equal to T60, we require that thedecay of 60 dB be accomplished in N60 steps given thesample rate fs,

, .log r N T f20 60 sN

60 6060` j (25)

We can now solve for the pole radius r of the discrete-timerepresentation,

.rT f

103

s60

(26)

We have now obtained a discrete-time model of a roommode described by two parameters, the mode frequency ωyielding the pole phase angle θ ωTs (where Ts denotesthe sampling interval) and the pole radius r.

We can also determine the desired location of a polehaving the target resonance decay time by using the sameapproach, selecting the same mode frequency but a modi-fied decay time T60. This will be needed later when con-structing a mode equalizer filter.

4 MODE EQUALIZER DESIGN

For the sake of simplicity only the design of the type Imodal equalizer is presented here. The type I equalizer isbased on the concept of modifying the primary sound suchthat target modes decay faster.

The mode resonance is modeled in discrete time by apole pair z F(r, θ) determined in short-term Fourierdata using the method presented. The desired decaytime is produced by a modified pole pair zc F(rc, θ).The correction filter of Eq. (6) for a single modebecomes

.

H zA z

A z

r r

r r

1 1

1 1

c

cj

cj

j j

1 1

1 1

c ce e

e e

z z

z z

l^

^

^

` `

` `

hh

h

j j

j j

(27)

To give an example of the equalizer function, considera system defined by a pole pair (at radius r 0.95 andangular frequency ω 0.l8π) and a zero pair (at r 1.9and ω 0.09π) In order to reduce the decay time of thepoles, we want to shift the location of the poles to theradius r 0.8. The resulting type I equalizer has a notch-type magnitude response (Fig. 4) because the numeratorgain of the correction filter is larger than the denominatorgain. Cascaded with the system response, poles at radiusr 0.95 are canceled and new poles created at the desiredradius (Fig. 5). Impulse responses of the two systems (Fig.6) demonstrate a reduction in the mode resonance Q value.The decay envelope of the impulse response (Fig. 7) nowshows a rapid decay.

The precision of a mode pole location estimate deter-mines the success of mode equalization. Errors in the esti-mated mode frequency and decay rate will displace thezero compensating an actual room mode pole and reducethe accuracy of control. For example, an estimation errorof 1% in the mode pole radius (Fig. 7) or angle (Fig. 8)greatly reduces control. Despite the error we still see rapidinitial decay, but then the decay follows the original decayrate of the mode. This demonstrates that accurate model-ing of the pole locations is paramount to the success ofmodal equalization.


Fig. 4. Effect of modal equalization on example system and magnitude response of modal equalizer.



Fig. 5. Poles () and zeros (&) of mode-equalized example system. Equalization of example system mode is produced by cancelingoriginal system poles creating the mode resonance and replacing them with a pole pair with less gain, located at the same frequencybut closer to origin.

Fig. 8. Effect of mode pole frequency modeling error on decay envelope control by modal equalizer in example system.

Fig. 7. Effect of mode pole radius modeling error on decay envelope control by modal equalizer in example system.

Fig. 6. Impulse responses of original and mode-equalized example systems.

Sample


5 CASE STUDIES

In this section we present case studies to enlighten themodal equalization process. In the first two examples weuse artificially added modes with known center frequencyand decay time to explain and demonstrate the principlesof modal equalizer design. In the third example a realroom measurement is used and the robustness of modalequalization for offset locations is demonstrated.

The short-term Fourier transforms used and plotted inthese case examples are calculated using a tapered timewindow, or an “apodized rectangular window,” as Buntonand Small called it originally [13]. Prior to this calculationthe 850-ms impulse response is decimated to a sample fre-quency of 1179 Hz, resulting in 1003 sample values. Theimpulse response is zero-padded for short-term Fouriertransform calculation. The time window is constructed byattaching a half Hanning time window as the initial taper

(length of taper is 4.24 ms) at the beginning of a time win-dow extending to the end of the data. This one-sided timewindow is slid over the impulse response data while cal-culating Fourier transforms (4096 bins) at 30-ms timeintervals.

Plots of short-term Fourier data (typically called water-fall plots) are limited to 40 dB. This floor is useful as itenhances the visibility of the initial decay rate increaseproduced by modal equalization, and it makes it easier toroughly judge the initial decay rate.

The ensemble average level decay within the frequencyband of interest is visualized by backward integrating thebandpass-filtered impulse response magnitude [17].

5.1 Cases with Artificial ModesCase I demonstrates the mode equalizer design. It is

based on a free-field response of a compact two-way loud-speaker measured in an anechoic room (Fig. 9). Someslow low-level decay can be observed in the waterfall.This is a combination of the decay of loudspeaker reso-nances and the fact that the sound absorption in the ane-choic room is no longer perfect at very low frequencies,below 50 Hz.

An artificial mode with a decay time T60 1.0 s isadded to the data at the frequency f 100 Hz (Fig. 10).The target frequency band for mode equalization is30–200 Hz. An equalizer is designed to limit the modedecay time to the 70th percentile of decay time distribu-tion measured within the target frequency band. Afterequalization we obtain a waterfall response very similar tothe original (Fig. 11). This trivial example highlights thebasic steps of mode equalizer design:

1) The short-term Fourier transform is calculated andscaled to set the maximum value within the target fre-quency range to 1.0.

2) Using the short-term Fourier transform, decay pro-files are detected. We used several detection levels (10,20, 25, 30, 35, and 40 dB). Some resulting


Fig. 9. Anechoic waterfall plot of two-way loudspeaker responseused in cases I and II.

Fig. 10. Case I. Free-field response of compact two-way loud-speaker with added artificial room mode at f 100 Hz.

Fig. 11. Case I. Mode-equalized free-field response of compacttwo-way loudspeaker with added artificial room mode at f 100Hz.


decay profiles for case I are plotted in Fig. 123) The mode equalization decay time target is set using

the decay profiles. We use the decay profile P20 detectedat level 20 dB. The 70th percentile of the P20 value dis-tribution is extrapolated to 60 dB and set as equalizedtarget decay time T60. An alternative target can be setbased on the reverberation time of the center frequenciesof 500 Hz to 2 kHz.

4) The mode detection function C is calculated as aweighted sum of the decay profiles. The detection func-tion for case I is plotted in Fig. 13

5) The mode detection function is used to determinethe number of modes to equalize. To determine this num-ber, a limit is set in the statistical distribution of the modedetection function values. In the case studies we used the65th percentile value. The count of peaks above this levelin the mode detection function data is taken as the numberof modes to equalize.

6) Recording the frequency of the mode detection func-tion maximum produces the first estimate of the mode fre-quency. Note that the mode detection function value onlyexists at Fourier transform bin frequencies. We use nearest-neighbor fitting with a second-order function [see Eq. (18)]to find an accurate frequency of the maximum interpolatedbetween bin frequencies. This is the estimate of the modefrequency based on the decay profile.

7) The second estimate for the mode frequency isobtained using the maximum magnitude of the Fouriertransform. This mode frequency estimate is made moreaccurate by fitting a second-order function to the magni-tude values of the bin where the maximum was found

and its two neighboring bins [see Eq. (18)]. This is theestimate of the mode frequency based on the maximummagnitude.

8) Now a choice must be made as to which mode fre-quency estimate to use in subsequent modal equalizerdesign. It is also possible to combine these two mode fre-quency estimates, or to take actions based on the similar-ity of their values. We made the choice prior to the design,based on the type of problem. Decay-profile-based modefrequency estimates may be more reliable in real roomswith complex closely spaced modes because in theserooms magnitude maxima do not necessarily correctlyindicate the frequency of mode resonance. The case Idesign presented in this paper uses maximum-magnitude-based mode frequency estimation, but both maximum-magnitude and decay-profile-based estimates yielded verysimilar designs for case I.

9) The impulse response is bandpass filtered to excludepossible other modes. The corner frequencies of the band-pass filter are chosen as 0.8ωm, 1.2ωm, where ωm is thepreviously estimated mode frequency.

10) The decay time is estimated using the nonlinear fit-ting approach [see Eq. (22) and [16].

11) The radii of the mode compensating filter poleand zero are determined based on the estimated centerfrequency and decay time and the decay-time target setearlier.

12) A second-order mode equalizer of Eq. (27) has nowbeen determined. After this, the room impulse response isfiltered with the equalizer to equalize this mode. Steps 1–11 are repeated a predetermined number of times to equal-


Fig. 13. Case I. Initial value of mode detection function.

Fig. 12. Case I. Initial decay profiles detected at l0-dB, 20- dB, 25-dB, and 35-dB levels (bottom to top).


ize all modes, excluding steps 3, 5, and 8, which are per-formed only the first time the algorithm is run.

The resulting equalizer for case I has the magnituderesponse of a simple notch at the mode frequency (Fig. 14)and equalization speeds up to the initial decay (Fig. 15).The reduction of the decay speed at time t 0.2 to 0.4 sin Fig. 15 was shown to be a result of noise in the meas-urement file showing up as an increase in the backward-

integrated level, and does not indicate a reduction in decayspeed. This is also evident in the waterfall plots of Figs.9–11.

Case II uses the same anechoic two-way loudspeakermeasurement. Five artificial modes with slightly differingdecay times are added at mode frequencies (decay times inparentheses) of 50 Hz (1.4 s), 55 Hz (0.8 s), 100 Hz (1.0s), 130 Hz (0.8 s), and 180 Hz (0.7 s). After modal equal-


Fig. 14. Case I. Modal-equalizer magnitude response.

Fig. 15. Case I. Impulse responses. (a) Original system with artificial mode at f 100 Hz. (b) Mode-equalized system. (c) Decayenvelopes for original (– · –) and equalized (–––) systems within frequency band of interest (30–200 Hz). Apparent reduction in decayspeed at t 0.2–0.4 s is caused by noise in anechoic measurement.

(c)

(b)

(a)


ization the magnitudes of the mode resonances have beendecreased (Figs. 16 and 17). There is an initial fast decay(Fig. 18). The same anechoic measurement as in case I

was used for generating the synthetic case II. Again, thereduction of the decay speed at time t 0.2 to 0.4 s in Fig.18 was shown by inspection to be a result of noise in the


Fig. 16. Case II. Five artificial modes added to impulse responseof compact two-way loudspeaker anechoic response. Modes areadded at frequencies (decay times in parentheses) of 50 Hz (1.4 s),55 Hz (0.8 s), 100 Hz (1.0 s), 130 Hz (0.8 s), and 180 Hz (0.7 s).

Fig. 17. Case II. Waterfall of mode-equalized system.

Fig. 18. Case II. Impulse responses. (a) Original system with five artificial modes. (b) Mode-equalized system. (c) Decay envelopes fororiginal (– · –) and equalized (–––) systems within frequency band of interest (30–200 Hz). Apparent reduction in decay speed at t 0.2–0.4 s is caused by noise in anechoic measurement.

(c)

(b)

(a)


measurement file, and is not a true indication of a reducedspeed of decay. We tried two mode frequency identifica-tion methods. Both the magnitude-maximum and thedecay-rate-based mode frequency identification methodsproduced similar results.

5.2 Case with a Real Room ResponseCase III is a measurement in a hard-walled approxi-

mately rectangular room with 40.9 m2 floor area (length7.18 m by width 5.69 m by height 2.5 m). The room isconstructed of concrete except for the front half of thefloor, which is a 30-mm chipboard construction on 200-by50-mm studs having 0.6-m spacing, covering a 0.6-m-deep concrete floor cavity in the front half of the room toform a flat floor surface.

The impulse response measurement for the filter designis taken in the center of the room (“design point”) at theheight of 1.2 m. The sound source is a three-way activestudio monitor loudspeaker (Genelec 1037B) placed on a

stand of 0.7-m height in the right front corner of the room.The vertical middle point of the front baffle is located 0.5m from both walls in the corner of the room, with theloudspeaker acoustical axis aimed to 1 m forward from thedesign point.

Additional measurements were taken 1 m to the left andright of the reference point, and also 1 m forward toward thefront of the room, and 1 m to the left and right from thatpoint, totaling six measurements, including the measure-ment of the design point. The floor area covered by themeasurements is 2 m2. The loudspeaker–microphonearrangement has front–back symmetry, but not left–rightsymmetry, and therefore the six measurements will give anidea about how local the effect of modal equalization is inthis room over a floor area of 4 m2. The listening location ofthe recording engineer is always in some way variable, andthe measurements outside the design point give a chance toget an idea of the spatial robustness of modal equalization.

To remove infrasound frequencies prior to processing,the original impulse responses are filtered by a fourth-order Butterworth high-pass filter with a corner frequencyat 20 Hz. The reverberation time of the room is 0.37 s inthe midfrequency band of 500 Hz to 2 kHz, and 1.2 s inthe target frequency band (30 to 200 Hz) measured in theinitial 30 dB of level decay. The low-frequency waterfallresponses (Fig. 19) reveal strong distinct modes, with thelongest decay time occurring at the second-order axial res-onance along the length of the room at the frequency of 47Hz. The first- and second-order eigenfrequencies for a rec-tangular cavity with the dimensions of the listening roomcan be used as an estimate of the expected modal frequen-cies (Table 1). It is typical that not all of the theoreticallyexisting eigenmodes are measurable in a real room.

The equalizer was designed with the algorithm pre-sented in case I, but incorporating additional optimizationsteps to be detailed. The decay-profile-based method wasused in estimating mode frequencies. Designs wheremode frequencies were identified using the magnitude-


(a)

(b) (c)

Fig. 19. Case III. Waterfall plots in real listening room. (a) 1.0 m forward and 1.0 m left of design point. (b) 1.0 m forward from designpoint. (c) 1.0 m right of design point. (d) 1.0 m left of design point. (e) Design point (center of room). (f ) 1.0 m right of design point.Measurements taken with omnidirectional microphone.


maximum identification were not successful. The case IIIroom response has some closely spaced modes, and is anexample of a real room where a single magnitude maxi-

mum is in fact constituted of several closely spaced moderesonances. Some of these resonances decay faster, causingthe apparent frequency of maximum magnitude to shift infrequency, rendering the use of maximum-magnitude-based identification problematic.

Additional optimization steps were incorporated in thecase III design to allow the interaction of closely spacedcorrection filter notches to be taken into account duringfilter design. The equalizer was designed with the iterativeapproach presented in case I, but incorporating the fol-lowing additional optimizing steps:

a) For every algorithm cycle, at step 12 of the basic iter-ative design algorithm presented for case I, calculate thedistance ε from the newly found mode center frequency ωto the center frequencies ωk of all N 1 previously deter-mined mode equalizers,

< , , , .εω

ω ωε for k N1 1

k

k0 f (28)

If the distance ε is smaller than ε0 (the value 0.01 wasused), consider this to be the same filter as the closest pre-vious filter. Discard the newly defined filter and changethe pole radius of the closest equalizer filter to move thefilter pole toward the origin by a design step to increase itsnotch depth. The design step is taken by modifying thepole radius r by an exponent a (the value 1.19 was used;note that r < 1.0),

.r n r n 1 a^ ^h h (29)


(d)

(e)

(f)

Fig. 19. Continued

Table 1. First- and second-order eigenfrequenciesin a rectangular cavity with case III listening

room dimensions.

Frequencyp q r [Hz] Type of Mode

1 0 0 23.96 Axial0 1 0 30.23 Axial1 1 0 38.57 Tangential2 0 0 47.91 Axial2 1 0 56.65 Tangential0 2 0 60.46 Axial1 2 0 65.03 Tangential0 0 1 68.80 Axial1 0 1 72.85 Tangential0 1 1 75.15 Tangential2 2 0 77.14 Tangential1 1 1 78.87 Oblique2 0 1 83.84 Tangential2 1 1 89.12 Oblique0 2 1 91.59 Tangential1 2 1 94.67 Oblique2 2 1 103.36 Oblique0 0 2 137.60 Axial1 0 2 139.67 Tangential0 1 2 140.88 Tangential1 1 2 142.90 Oblique2 0 2 145.70 Tangential2 1 2 148.81 Oblique0 2 2 150.30 Tangential1 2 2 152.19 Oblique2 2 2 157.75 Oblique

Note: p, q, r––order of eigenfrequency alonglength, width, and height of room.


The motivation of this optimization step is to handle thecase where in fact the algorithm picks up the same modeagain because the mode equalizer notch filter attenuationis not sufficiently large, and this optimization step allowsthe notch attenuation to be adapted to a sufficient level.

b) During the design process, after every M cycles (thevalue of 10 cycles was used), a simulated annealing step[18] is performed. This moves all previously designedequalizer poles outward, away from the origin, decreasingthe depth of their respective notch filters by an exponentderived from an annealing noise generator N having a log-arithmic positive value distribution,

, , > .r n r n T N n1 0 ( )TN n^ ^ ^h h h (30)

The magnitude scaling of the log noise variable N is deter-mined by a positive annealing temperature coefficient T,reduced after each annealing step. The purpose of the sim-ulated annealing step is to allow a readjustment of gains in

all mode filters during the design to accommodate theinteraction of closely spaced filters. This is the mechanismin the design process that reduces the notch depth of amode equalizer, and it can also eventually eliminate somemode equalizer filters by reducing their notch depth tozero if they are no longer needed to achieve equalization.Note that when the annealing step reduces the filter notchdepth, the basic algorithm reconsiders these mode fre-quencies if necessary, and will tend to return their notchdepth to a sufficient level.

c) Repeat the basic filter design loop with these twooptimization steps until the mode detection function C(ω)assumes a value less than a preset design target. This ter-minates the design process. When the optimization stepsdescribed here are used in the design process, the termi-nation condition based on a predetermined number ofmodes (step 5 described in case I) cannot be used.

The equalized waterfall response (Fig. 20) shows thatdecays are now well controlled within decay profile cal-


Fig. 20. Case III. Waterfall plots in mode-equalized real listening room. (a) 1.0 m forward and 1.0 m left of design point. (b) 1.0 m for-ward from design point. (c) 1.0 m right of design point. (d) 1.0 m left of design point. (e) Design point (center of room). (f ) 1.0 m rightof design point. Mode equalizer design based on omnidirectional microphone measurement at design point (e).

(b) (d)

(c)(a)


culation levels (20–40-dB decay levels) at the designpoint [Fig. 20(e)]. Decay below the target frequency range( f < 30 Hz) is not included in the equalizer filter design,and therefore is not controlled.

Moving away from the design position, the modaldecay remains controlled, although the amount of modalequalization reduces slightly. This demonstrates that forthe low frequencies considered in this work, the modalequalizer is able to control modes in an area large enoughfor practical sound engineering work, and can thereforepotentially bring an improvement to the sound qualitywithin a limited area in a room.

Note in this case study that the modal equalizer is notable to control a mode at about 90 Hz for positions outsidethe design point. This is because this mode is not measur-able at the design point [Fig. 19(e)], so the filter design isblind to this particular mode. This emphasizes the impor-tance of the point in space chosen as the design point, andthe significance of sufficient study of the modal structurein the room in the vicinity of the intended design point inorder to choose a response that is sufficiently representa-tive for successful modal equalizer design. If a mode hap-pens to have a null at the chosen design point, as in caseIII, this mode is not going to be controlled. This modemay then become disturbingly audible when other modes

are controlled at points close to the design point, where thelevel of this uncontrolled mode increases rapidly.

The reverberation time within the target frequency rangeafter equalization is 0.7 s measured in the initial 30 dB of leveldecay (original reverberation time was 1.2 s measured in thesame way). The resulting total equalizer (Fig. 21) is a collec-tion of notch filters, and has a very small positive gain outsidethe target frequency band. The equalized system magnituderesponse (Fig. 22) shows smaller magnitude deviation, andequalization has reduced the overall level close to mode fre-quencies. Shortening of the decay is also visible in thebandpass-filtered (30–200-Hz) impulse response [Fig.23(a) and (b)], and the energy decay shows fast initial decayand then follows the decay speed of the original system [Fig.23(c)].

The resulting equalizer filter contains nine second-ordersections. It is somewhat complex, but given that we areworking at low frequencies, multirate techniques allow anefficient and cost-effective implementation. A total of nineoptimization cycles were needed in the filter design. Forthis design, a simulated annealing step was not usedbecause the design converged so rapidly, but an equalizergain optimization step was used. The notch frequencies ofthe modal equalizer (Table 2) coincide fairly well with theeigenmodes calculated for a rectangular cavity with the


Fig. 21. Case III. Modal equalizer magnitude response.

Fig. 20. Continued

(e) (f)


dimensions of the room (Table 1). Naturally, only a part ofthe modes are measurable at the design point, thereforebecoming subject to equalization.

6 MODAL EQUALIZER IMPLEMENTATION

6.1 Type I Filter ImplementationTo correct N modes with a type I mode equalizer, we

need an order-2N IIR transfer function. The most immedi-

ate method is to design a second-order filter defined byEq. (27) for each mode. The final order-2N filter is thenformed as a cascade of second-order subfilters,

.H z H z ,c c kk

N

1

%^ ^h h (31)

Another formulation allowing the design for individualmodes is served by the formulation in Eq. (10), a parallelformulation of a cascaded correction filter. This leads to a


Fig. 22. Case III. Magnitude response of design point in listening room. Unequalized (– · –) and after modal equalization (–––).

Fig. 23. Case III. Impulse responses. (a) Original system. (b) Mode- equalized system. (c) Decay envelopes for original (– · –) and equal-ized (–––) systems within frequency band of interest (30–200 Hz).

(c)

(b)

(a)


parallel structure where the total filter is implemented as

.H z H z1 ,c c kk

N

1

! l^ ^h h (32)

6.2 Asymmetry in Type I EqualizersAt low angular frequency the maximum gain of a reso-

nant system may no longer coincide with the pole angle[19]. This is also true with modal equalizers, and must becompensated for in the design of an equalizer. A basic typeI modal equalizer becomes increasingly unsymmetrical asthe frequency approaches ω 0. A case example (Fig. 24)shows a standard design with pole and zero at ωp,z 0.01rad/s, zero radius rz 0.999, and pole radius rp 0.995.There is a significant gain change for frequencies belowthe resonant frequency. This asymmetry may cause aproblematic cumulative change in gain.

It is possible to avoid asymmetry by decreasing thesampling frequency in order to bring the mode resonanceshigher on the discrete frequency scale. If sufficient samplerate alteration is not possible, we can symmetrize a modeequalizer by moving the pole slightly downward in fre-quency. This will symmetrize the response, but it will alsoshift the notch frequency. These effects must be accountedfor when symmetrizing a mode equalizer at low frequen-cies. This can be handled by an iterative fitting procedurewith a target to achieve the desired mode decay timesimultaneously with a symmetrical response.

6.3 Type II Filter ImplementationThe type II mode equalizer uses one or more secondary

radiators in producing additional audio in the room to con-trol the decay time of modes. The type II equalizerrequires a solution of Eq. (8) for each secondary radiator.The correcting filter Hc(z) can be implemented directlyaccording to Eq. (8) as a difference of two transfer func-tions convolved by the inverse of the secondary radiatortransfer function. Then it is important to recognize therequirement of Eq. (11) that the secondary radiators beable to couple sufficiently to the room at frequencies ofinterest in order to control the sound field.

A more optimized implementation can be found by cal-culating the correcting filter transfer function Hc(z), thendesigning an FIR or IIR filter approximating this transferfunction. This filter can then be used as the correcting fil-ter. There are several standard filter design techniques thatcan be used in designing such a filter.

In the case of more than one secondary radiator thesolution becomes more convoluted as the contributions ofall secondary radiators must be considered if they haveoverlapping frequency ranges. For example, the solutionof Eq. (12) for the correction filter of the first secondaryradiator is

m.H z

H z

H z H z H z H z

,

,

, ,c

m c n nn

N

11 1

12!l^

^

^ ^ ^ ^h

h

h h h h

(33)

All secondary radiators interact to form the correction.Therefore the filter design process becomes a multidimen-sional optimization task where all correction filters mustbe optimized together. A suboptimal solution is to designone secondary filter at a time such that the subsequentsecondary sources only handle those frequencies not con-trollable by the previous secondary sources, for instance,because of poor radiator location in the room.

7 DISCUSSION

In this paper we have presented two different methodsto apply a mode equalizer in a room, type I modifying thesound input into the room using the primary loudspeakers,and type II using separate loudspeakers to output a soundthat controls the modal decay. Type I systems are typicallyminimum-phase systems. Type II systems, because thesecondary radiator is separate from the primary radiator,may have an excess phase component because of differingtimes of flight. As long as this is compensated in themodal equalizer implementation, type II systems can alsobe considered minimum-phase systems.

An optimization algorithm was presented for modalequalizer design. Several methods for finding and identi-fying modal resonances were presented. Both the maxi-mum magnitude and the information about decay timewere used as methods to find the dominant mode to beequalized in the short-term Fourier transform of the trans-fer function impulse response. Parameters describing themode were identified in the short-term Fourier transform


Fig. 24. Modified type I modal equalizer with symmetrical gainhaving zero radius r 0.999 at ω 0.01 rad/s and pole radiusr 0.995 at ω 0.0087 rad/s (–––), and standard type I modalequalizer having both a pole and a zero at ω 0.01 rad/s (– · –).

Table 2. Case III modal equalizer represented as a cascade ofsecond-order notch filters.

f0 G Zero PoleN [Hz] [dB] [rad] Radius Radius

1 30.6 8.38 0.13960 0.99131 0.976782 38.7 10.77 0.17639 0.99349 0.976783 45.4 13.55 0.20711 0.99539 0.976784 45.9 10.20 0.20920 0.99289 0.976785 46.7 1.35 0.21307 0.99550 0.994656 63.9 4.21 0.29148 0.98597 0.976787 139.6 8.73 0.63664 0.99168 0.976788 146.4 9.62 0.66728 0.99253 0.976789 152.9 7.15 0.69694 0.99013 0.97678

Note: f0––notch center frequency; G––notch gain; ω––angularfrequency of zero and pole.


data and used to obtain modal equalizer coefficients. Aniterative method of designing a modal equalizer was pre-sented, where the currently identified modes are equalizedin the impulse response to remove the effect of thesemodes, and the mode equalizer order is increased, includ-ing optimization of the already determined mode equalizercoefficients, until sufficient mode equalization is achieved.

Case examples demonstrated that mode equalizationcan achieve at least rapid initial decay with type I designs.At low frequencies the control of modal decay wasachieved also in a real room (case III) over a listening areareasonably sized for a sound engineer. Case III alsodemonstrated that the design method is able to deal withclosely spaced modes in a real room because of iterativedesign and optimization of modal filter coefficients.

The conventionally applied equalization target is toequalize a system to approximate a unity transfer function,causing no change to the magnitude or phase within a cer-tain bandwidth of interest. An example of a room having atransfer function very close to unity is the anechoic room.This may not be a very practical equalization target forloudspeakers in real rooms as the goal of room equaliza-tion is usually not to render a listening room anechoic byeliminating the room resonances entirely, but to improvethe perceived quality of sound reproduction in a listeningenvironment. In fact, listeners may even prefer to hearsome room response in the form of liveliness creating aspatial impression and some envelopment [20]. The con-ventional goal of practical listening room equalization isto minimize or eliminate the subjective experience of col-oration in audio due to room effects. Magnitude equaliza-tion can achieve a subjectively flat magnitude response atthe listening location for either the early arriving or thesteady-state sound. Both can improve the perceived audioquality for compromised loudspeaker–room systems sig-nificantly. However, coloration due to the reverberantsound field and slowly decaying sound due to high-Q roommodes cannot be sufficiently controlled with traditional-magnitude equalization. Such coloration may deterioratethe sound clarity and definition and become audible as achange in the temporal characteristics of the audio, or itmay cause a change in the frequency-domain characteris-tics, becoming audible as a change in the timbre of audio.

Modal equalization is a novel approach specificallyaddressing problematic modal resonances. By decreasingtheir Q value, excessively long decay rates in low fre-quencies can be reduced and set similar to decay rates athigher frequencies, in order to minimize the audibility ofenergy storage in low-frequency modes. Because modalequalization also reduces the gain of mode resonances, itdoes produce an amount of magnitude equalization aswell, although it does not guarantee magnitude equaliza-tion, and conventional magnitude equalization must typi-cally be performed after modal equalization to optimizethe audio quality.

It is important to note that conventional-magnitudeequalization does not achieve modal equalization as a by-product, nor does a modal equalizer necessarily achieve asystem response subjectively flat in frequency. There isno guarantee that zeros in a magnitude equalizer transfer

function have correct gain or frequency to achieve controlof mode resonance decay time. In fact, this is ratherimprobable. Modal equalization should be used withconventional magnitude equalization because modalequalization alone does not achieve a subjectively flatmagnitude response.

There are several reasons why modal equalization is par-ticularly interesting at low frequencies where passive meansto control the decay rate by room absorption may be pro-hibitively expensive, fail because of constructional faults, orsimply were not even considered at the time a facility wasconstructed. At higher frequencies methods to controlmodes are economical to implement and readily available.Furthermore, modal equalization becomes technically fea-sible at low frequencies, where the wavelength of soundbecomes large relative to the room size and objects in theroom, rendering the sound field no longer diffuse, andtherefore making (at least) local control of the sound fieldprogressively easier to achieve with decreasing frequency.

Modal equalization at low frequencies does not onlyaddress the steady-state sound in a room. By modifyingthe goodness of the modal resonances it actually modifiesthe transfer function in the room, at least in the vicinity ofthe design target point, causing related changes to thetransient properties of the room as well. At low frequen-cies, where the sound field is no longer diffuse, it may bemisleading to talk about the “direct sound” or “steady-state sound.” Modal equalization modifies both the tran-sient sound and the steady-state sound at low frequencies.This can be seen in the case examples in both the water-falls and the impulse responses.

The choice of the mode detection and parameter identi-fication method affects the success of modal equalizerdesign. The modal equalizer design method suggested inthis paper can also model closely spaced and coupledmodes with different decays. It tends to place a single or afew equalizers to account for very closely spaced modes,and may not be able to discern every individual modeaffecting the decay in that case. A modal equalizer tech-nique based on applying ARMA modeling has recentlybeen suggested [12], and may provide additional robust-ness in equalizing closely spaced or strongly coupledmodes with complex decay characteristics.

What should then be the design target for the decay rateat low frequencies? Recommendations [9]–[11] suggestthat it is psychoacoustically desirable to have approxi-mately equal reverberant decay rates over the audio rangeof frequencies, with possibly a modest increase towardlow frequencies. We have used this as the starting point todefine a target for modal equalization. We proposed alsoanother possible target, based on the statistical distributionof decay rates within the frequencies of interest, with asimilar psychoacoustical motivation. These targets mayserve as a starting point, but further research is needed inorder to find the limits to modal decay perception in orderto determine a psychoacoustically well-founded designtarget for modal equalization. Attempts toward this havebeen made quite recently in the area of psychoacousticstudy [21], [22] and modeling [23], [24].

Finally, because it appears clear that conventional-



magnitude equalization should be applied with a modalequalizer, there is the question of the incremental benefitof applying a modal equalizer once magnitude equaliza-tion has been performed. How much does the sound qual-ity increase by also applying modal equalization? Thisquestion is a subject of future study and particularly rele-vant. A recent preliminary listening test by the authorsappears to suggest that the incremental improvement ofapplying modal equalization may be small compared tothe initial incremental improvement provided by propermagnitude equalization, in itself not able to equalize themodes in a room [25].

8 SUMMARY AND CONCLUSIONS

In this paper we introduced the principle of modal equal-ization and formulations for type I and type II modal equal-ization filters. The type I system implements modal equal-ization by a filter in cascade with the main sound source,modifying the sound input into the room. The type II systemimplements modal equalization by one or more secondarysources in the room, requiring a correction filter producing acompensatory sound for each secondary source.

Methods for identifying and modeling modes in impulseresponse measurements were presented. The error sensitiv-ity of modeling and the implementation of system transferfunction poles were discussed. Case examples of modeequalizers with demonstrations of the design procedure weregiven of both simulated and real rooms. The spatial robust-ness of modal equalization in a real room was demonstrated.Finally, aspects related to the implementation of both type Iand type II mode equalizer filters were discussed.

Modal equalization is a method to control the reverber-ation in a room when conventional passive means are notfeasible, do not exist, or would present a prohibitivelyhigh cost. Modal equalization is an interesting designoption, particularly for low-frequency room reverberationcontrol when the mode density is not very high.Conventional-magnitude response equalization and modalequalization supplement each other in optimizing the psy-choacoustical quality of a monitoring space.

9 ACKNOWLEDGMENT

This study is part of the VÄRE technology program proj-ect TAKU (Control of Closed Space Acoustics) funded bythe National Technology Agency of Finland (Tekes). Untilthe fall 2001 the work of Vesa Välimäki was financed with apostdoctoral research grant by the Academy of Finland. Partof this work was conducted when he was with the PoriSchool of Technology and Economics, Tampere Universityof Technology, Pori, Finland, during the academic year2001–2002. In 2003 the work by Poju Antsalo was partiallyfunded by the Spatial Audio and Room Acoustics (SARA),Academy of Finland project no. 201050.

10 REFERENCES

[1] A. R. Groh, “High-Fidelity Sound System Equal-ization by Analysis of Standing Waves,” J. Audio Eng.

Soc., vol. 22, pp. 795–799 (1974 Dec.).[2] S. J. Elliott and P. A. Nelson, “Multiple-Point

Equalization in a Room Using Adaptive Digital Filters,” J.Audio Eng. Soc., vol. 37, pp. 899—907 (1989 Nov.).

[3] S. J. Elliott, L. P. Bhatia, F. S. Deghan, A. H. Fu,M. S. Stewart, and D. W. Wilson, “Practical Implemen-tation of Low-Frequency Equalization Using AdaptiveDigital Filters,” J. Audio Eng. Soc., vol. 42, pp. 988–998(1994 Dec.).

[4] J. Mourjopoulos, “Digital Equalization of RoomAcoustics,” presented at the 92th Convention of the AudioEngineering Society, J. Audio Eng. Soc. (Abstracts), vol.40, p. 443 (1992 May), preprint 3288.

[5] J. Mourjopoulos and M. A. Paraskevas, “Pole andZero Modelling of Room Transfer Functions,” J. SoundVibr., vol. 146, pp. 281–302 (1991).

[6] R. P. Genereux, “Adaptive Loudspeaker Systems:Correcting for the Acoustic Environment,” in Proc. AES8th Int. Conf. (Washington, DC, 1990 May), pp. 245–256.

[7] F. E. Toole and S. E. Olive, “The Modification ofTimbre by Resonances: Perception and Measurement,” J.Audio Eng. Soc., vol. 36, pp. 122–142 (1988 Mar.).

[8] S. E. Olive, P. L. Schuck, J. G. Ryan, S. L. Sally, andM. E. Bonneville, “The Detection Thresholds of Reson-ances at Low Frequencies,” J. Audio Eng. Soc., vol. 45, pp.116–127 (1997 Mar.).

[9] ITU-R BS.1116-1, “Methods for the Assessment ofSmall Impairments in Audio Systems Including Multi-channel Sound Systems,” International Telecommunica-tions Union, Geneva, Switzerland (1994).

[10] AES Tech. Comm. on Multichannel and BinauralAudio Technology (TC-MBAT), “Multichannel SurroundSound Systems and Operations,” Tech. Doc., version 1.5(2001).

[11] EBU Tech. Doc. 3276-1998, “Listening Conditionfor the Assessment of Sound Programme Material: Mono-phonic and Two-Channel Stereophonic,” 2nd ed. (1998).

[12] M. Karjalainen, P. A. A. Esquef, P. Antsalo, A.Mäkivirta, and V. Välimäki, “Frequency-Zooming ARMAModeling of Resonant and Reverberant Systems,” J. AudioEng. Soc., vol. 50, pp. 1012–1029 (2002 Dec.).

[13] J. D. Bunton and R. H. Small, “CumulativeSpectra, Tone Bursts, and Apodization,” J. Audio Eng.Soc., vol. 30, pp. 386–395 (1982 June).

[14] J. 0. Smith and X. Serra, “PARSHL: AnAnalysis/Synthesis Program for Non-Harmonic SoundsBased on a Sinusoidal Representation,” in Proc. Int.Computer Music Conf. (Urbana, IL, 1987), pp. 290–297.

[15] M. Desainte-Catherine and S. March, “High-Precision Fourier Analysis of Sounds Using SignalDerivatives,” J. Audio Eng. Soc., vol. 48, pp. 654–667(2000 July/Aug.).

[16] M. Karjalainen, P. Antsalo, A. Mäkivirta, T.Peltonen, and V. Välimäki, “Estimation of Modal DecayParameters from Noisy Response Measurements,” J.Audio Eng. Soc., vol. 50, pp. 867–878 (2002 Nov.).

[17] M. R. Schroeder, “New Method of MeasuringReverberation Time,” J. Acoust. Soc. Am., vol. 37, pp.409–412 (1965).

[18] S. Kirkpatrick, C. D. Gelatt, and M. P. Vecchi,



“Optimization by Simulated Annealing,” Science, vol.220, pp. 671–680 (1983).

[19] K. Steiglitz, “A Note on Constant-Gain DigitalResonators,” Computer Music J., vol. 18, no. 4, pp. 8–10(1994).

[20] R. Walker, “Equalisation of Room Acoustics andAdaptive Systems in the Equalisation of Small RoomAcoustics,” presented at the AES 15th Conf. (1998 Oct.),paper 15-005.

[21] T. I. Niaounakis and W. J. Davies, “Perception ofReverberation Time in Small Listening Rooms,” J. AudioEng. Soc., vol. 50, pp. 343–350 (2002 May).

[22] L. Fielder, “Analysis of Traditional and Reverberation-

Reducing Methods of Room Equalization,” J. Audio Eng.Soc., vol. 51, pp. 3–26 (2003 Jan/Feb.).

[23] T. Paatero and M. Karjalainen, “New Digital FilterTechniques for Room Response Modeling,” presented atthe AES 21st Int. Conf., St. Petersburg, Russia (2002 June1–3).

[24] M. Avis, “Q Factor Modification for Low-Frequency Room Modes,” presented at the AES 21st Int.Conf., St. Petersburg, Russia (2002 June).

[25] P. Antsalo, M. Karjalainen, A. Mäkivirta, and V.Välimäki, “Comparison of Modal Equalizer Design Methods,”presented at the 114th Convention of the Audio EngineeringSociety, Amsterdam, The Netherlands (2003 Mar.).


THE AUTHORS

Aki Mäkivirta was born 1960 in Jyväskylä, Finland. Hereceived Diploma Engineer, Licentiate in Technology, andDoctor of Science in Technology degrees in electricalengineering from Tampere University of Technology,Tampere, Finland, in 1985, 1989, and 1992, respectively.His doctor’s thesis described applications of nonlinearsignal processing methods for haemodynamic monitoringin medical intensive care.

In 1983, Dr. Mäkivirta joined the Medical EngineeringLaboratory of the Research Center of Finland atTampere, where he worked in various research positionsin the field of biomedical signal analysis. In 1990 hejoined Nokia Corporation Research Center in Tampere,Finland, where he served as a project manager, and after1992 was the research manager responsible for building aresearch group for DSP applications in television audioand high-quality loudspeaker reproduction. In 1995 hejoined Genelec Oy, Iisalmi, Finland, where he is a R&Dmanager responsible for digital system and signal pro-cessing applications.

Dr. Mäkivirta is a member of the AES and IEEE. Heholds 13 patents, and he is the author of more than 40journal and conference papers.

Poju Pietari Antsalo was born in Helsinki, Finland, in1971. He studied electrical and communications engi-neering at the Helsinki University of Technology, wherehe obtained a Master's degree in 1999. Since then, he hasbeen carrying out research in low-frequency roomacoustics at the Helsinki University of Technology,Laboratory of Acoustics and Audio Signal Processing.E-mail: [email protected].

The biography of Matti Karjalainen was published in the2003 January/February issue of the Journal.

Vesa Välimäki was born in Kuorevesi, Finland, in1968. He received Master of Science in Technology,Licentiate of Science in Technology, and Doctor ofScience in Technology degrees in electrical engineeringfrom the Helsinki University of Technology (HUT),Espoo, Finland, in 1992, 1994, and 1995, respectively.

Since 1990, Dr. Välimäki has worked mostly at theHUT Laboratory of Acoustics and Audio Signal Process-ing with the exception of a few periods. In 1996 he spentsix months as a postdoctoral research fellow at theUniversity of Westminster, London, UK. During the aca-demic year 2001–2002 he was professor of signal pro-cessing at the Pori School of Technology and Economics,Tampere University of Technology, Pori, Finland. InAugust 2002 he returned to HUT, where he is currentlyprofessor of audio signal processing. In 2003 he wasappointed a docent in signal processing at the Pori Schoolof Technology and Economics. He lectures on digitalaudio signal processing at the HUT, the Pori School ofTechnology and Economics, and the Centre for Music andTechnology, Sibelius Academy, Helsinki, Finland. Hisresearch interests include sound synthesis and digital fil-ter design.

Dr. Välimäki is a senior member of the IEEE SignalProcessing Society and is a member of the AES, theAcoustical Society of Finland, and the Finnish Music-ological Society. He was papers chair of the AES 22ndInternational Conference on “Virtual, Synthetic, andEntertainment Audio,” 2002 June, Espoo, Finland.

A. Mäkivirta P. P. Antsalo V. Välimäki

PAPERS

0 INTRODUCTION

In a sound field the sound intensity can be determinatedfrom two acoustical measurements, for example, pressureand particle velocity. In the p-p method, nowadays themethod used most commonly, two spaced microphones areused to calculate the sound intensity. Another method is thep-u method, which uses a pressure sensor to measure thesound particle velocity. A micromachined sensor known asthe Microflown is used for measuring the particle veloc-ity. The phase relationship between the two sensors is oflittle influence on the measurement since the in-phase partof the pressure and the particle velocity determines theactive sound intensity directly. In the case of the p-pmethod the out-of-phase pressures have to be used asmeasures of the active sound intensity while the pressureand particle velocity signals are almost in phase. Due to themicromachined particle velocity sensor, the size of theprobe is as small as a 1⁄2-in microphone, which makes close-range sound intensity measurements feasible. Since the p-umethod is less sensitive to errors in phase than the p-pmethod, the need for high-resolution data-acquisition toolsis not as important as in the p-p method. The paper presentssound intensity measurements performed with the low-costintensity probe using a computer soundcard combined withopen-source software. The results obtained with the p-uprobe show good agreement with those of a p-p probe.

1 SOUND INTENSITY AND ITS MEASURINGTECHNIQUES

For the determination of sound intensity, the two com-ponents of a sound wave, particle velocity and sound pres-

sure, have to be known [1], [2]. The measurement of thesound pressure (or particle velocity) only gives the sum ofthe free and diffuse sound fields. However, the free-fieldproperties are obtained from the (time-averaged) productof the instantaneous pressure p(t) and the correspondinginstantaneous particle velocity u(t) at the same position,

t dIT

p t u t1

T

0

:# ^ ^h h (1)

where the intensity I and the velocity u are vectors. Themeasured intensity corresponds in fact to the net flow ofacoustic energy at a given position. If the intensity arounda sound source is measured at a number of positions, theradiated sound power can be determined. Even in the pres-ence of reverberation or background noise one can deter-mine in this way the (free-field) radiated power of a soundsource [2].

Since the relationship between pressure and particlevelocity is not unique but depends on the sound field, anintensity measurement system should contain transducersto measure both quantities independently. The sound pres-sure can be measured in a reliable manner. However, thetechnical difficulties of designing a suitably stable, linearwide-frequency-band transducer for an accurate conver-sion of the fluid particle velocity into an analog electricsignal make sound intensity measurements difficult [2].

Instead of using transducers for particle velocity andsound pressure in order to measure the sound intensity, itwas shown that the sound intensity can also be determinedusing two nominally identical microphones [2]. We willrefer to the former method as the p-u method and the lat-ter as the p-p method.


A Low-Cost Intensity Probe*

R. RAANGS, W. F. DRUYVESTEYN, AES Life Member, AND H. E. DE BREE, AES Member

University of Twente, 7500 AE Enschede, The Netherlands

The sound intensity in a sound field can be determinated from two acoustical measure-ments such as pressure and particle velocity. In the p-p method, the most commonly usedmethod, the sound intensity is calculated by means of two spaced microphones. Anothermethod, the p-u method, uses a pressure sensor to measure the sound particle velocity. Usingthe p-u method, measurements performed with a low-cost intensity probe and a computersoundcard combined with an open-source software show good agreement with resultsobtained with a p-p probe.

* Manuscript received 2001 December 20; revised 2003January 30.

Microflown is a registered trademark of MicroflownTechnologies BV, Zevenaar, The Netherlands.

PAPERS INTENSITY PROBE

1.1 p-p Sound Intensity ProbeIn the early 1980s the first sound intensity measuring

systems became available and consisted of two closelyspaced identical microphones. These intensity probes arestill commonly used. In this p-p method the particle veloc-ity is calculated by the linearized equation of momentumconservation,

ρ ∆∆ du x p x p x t

1

xx

0#^ _ ^h i h (2)

where ρ0 is the density, p(x) the instantaneous sound pressureat position x, and ∆x the spacing between both microphones.

1.2 p-u Sound Intensity Probe: UltrasonicSolution

A totally different way of sensing the particle velocityis the principle of ultrasonic transduction: two parallelultrasonic beams are sent in opposite directions. The trav-eling time from the transmitter to the receiver is inverselyproportional to the speed of sound in the air. When the airis moving, this movement should be added to the speed.The probe (Fig. 1) consists of two transmitter–receiverpairs that are positioned in opposite directions. The differ-ence in the traveling times of the ultrasonic sound wavesis proportional to the particle velocity.

The sound intensity probe that is based on this principleis, however, no longer available because of its sensitivityto dc flows (wind). Furthermore, just like the p-p probe, itis a distributed sensor with the problems associated withthis type of sensing. (For example, the maximum fre-quency is limited by the spacing.) Finally, because of itsphysical dimensions, the probe is difficult to calibrate.

1.3 p-u Sound Intensity Probe: MicroflownSolution

Due to the development of the so-called Microflown atthe University of Twente in 1994, a novel way of soundintensity measurement became possible [3]–[5]. TheMicroflown is an acoustic sensor measuring particlevelocity instead of sound pressure. Thus a sound intensityprobe is created by the combination of a pressure micro-phone with this particle velocity sensor.

1.4 Microflown Operating PrincipleThe temperature sensors of the Microflown are imple-

mented as two closely spaced thin wires (1500 2.5 0.4 µm) of silicon nitride with an electrically conductingplatinum pattern (see Fig. 2). An electric current in thewires is dissipating electric power, thereby heating thewires. The wires therefore act as heaters and temperaturesensors. The sensors have a typical operational tempera-ture of about 200 to 400°C. When particle velocity is pres-ent, the temperature distribution around the wires changesasymmetrically. Fig. 3 illustrates the effect of a tempera-ture change in one wire. In this case the wire itself will notcool down. In the Microflown two wires are being used.The two wires are heated and cooled due to the changes inthe temperature profiles, as seen in Fig. 4. To calculate thisaltered temperature profile, perturbation theory can beused [6], [7]. Since the resistance of the wires is tempera-ture dependent, this change in temperature profile causesa change in the temperature of the resistors, and thus achange of resistance. The difference in resistance betweenthe two sensors quantifies the particle velocity (see Fig. 4).Due to this differential operation principle, the Micro-


Fig. 2. Bridge type Microflown. At the top a wire bond is visible.Sample is glued on a printed-circuit board; glue can be seen atthe side. This type is now used in commercial products.

Fig. 3. Temperature distribution around one wire. ––– no flow;g perturbed temperature profile due to convection (wire doesnot cool down); lower line––perturbation due to convection (atthe position of the wire the perturbation is zero; wire will notcool down).

∆T

[k]

velocity

distance 0

Fig. 1. Type 216 p-u intensity probe (Norwegian electronics).

RAANGS ET AL. PAPERS

flown can distinguish between positive and negativevelocity directions.

1.5 Frequency Response of the MicroflownThe Microflown is most sensitive toward the low-

frequency components of the sound particle velocity. Forhigher frequencies the sensitivity decreases. The fre-quency response of a Microflown can be described by thefollowing parameters: 1) a low-frequency sensitivity, and2) two characteristic frequencies. This high-frequencyrolloff is caused by diffusion effects. The effect can beestimated by a first-order low-pass frequency response,which has a diffusion corner frequency fdiff on the order of500 Hz to 2 kHz (depending on geometry and operatingtemperature). The second high-frequency rolloff is causedby the heat capacity of the wires (thermal mass) andshows an exact first-order low-pass behavior that has aheat capacity corner frequency fheat cap on the order of 2 to15 kHz for modern Microflowns (depending on geometryand operating temperature) [6], [7]. The frequencyresponse of the Microflown can be roughly described by

/ /output

LFS

f f f f1 1

heat cap diff2 2 2 2

(3)

LFS being the low-frequency sensitivity, the output signalat frequencies below the thermal diffusion corner frequencyfdiff. The measured phase response can be found from

tanphase Af

f

ph

1J

L

KK

N

P

OO

(4)

where A and fph are arbitrary constants, which can befound by fitting the calibration data.

1.6 Package GainApart from the protection of Microflown’s fragile sen-

sors, packaging has several effects. The particle velocitylevel rises considerably (10 to 30 dB depending on geom-etry and frequency) and the phase response changesslightly. The increase in the particle velocity level insidethe package (the so-called package gain) is caused by a

channeling effect, that is, the particle flow is “forced”through the package.

1.7 Pros and Cons of the p-p and p-u MethodsThe well-known p-p method has the advantage of its rela-

tively simple calibration and the fact that it is well describedand standardized. A disadvantage of the p-p method is thatone must change spacers (and microphones) in order to beable to measure the sound intensity over a broad frequencyband. A p-p probe is rather expensive as two identical micro-phones are used. It should further be noted that the bandwidthof the sound intensity measurements depends on the reactiv-ity of the sound field. (The low end of the bandwidthincreases if the reactivity increases.) The high-frequencylimit also depends on the reactivity of the sound field and thespacing between the two microphones.

Because the proposed p-u probe is quite different fromthe p-p probe, the advantages and disadvantages also differ.The main advantage of the p-u method is that all broad-band measurements (for example, 20 Hz to 20 kHz) can beperformed using the same probe configuration (and post-processing) without the need to change the spacers, whichis needed using the p-p method. Another main advantage isthat the phase mismatch is not very critical because in thep-u method we are not interested in small differencesbetween two large signals, which is the case in the p-pmethod, but we are simply interested in the product of thetwo signals. We still need to know the phase of the pressureand particle velocity sensors, but the measured soundintensity is not influenced much due to small errors in thisphase calibration. For example, phase error of 12 degwould result in an error in the active sound intensity of lessthan 0.1 dB if the sound field is mainly active.

Since the proposed p-u probe is relatively small, we canuse standard 1⁄2-in accessories and take measurements veryclose to the radiating surfaces. In addition the sensory is arelatively small acoustical obstacle due to its small size.Furthermore the p-u technique will give reliable results atvery short distances from a sound source, whereas the p-pmethod may introduce some errors since the linearizedEuler equation as used in the p-p technique may not bevalid close to the surfaces. Disadvantages of the p-u probeare difficulties in calibration and the fact that these cali-bration data have to be accounted for and the p-u methodis not a standardized method.

2 LOW-COST INTENSITY PROBE

A low-cost intensity probe is built by combining a pack-aged Microflown with a small microphone. The total costof operation of the 1⁄2-in p-u sound intensity probe is opti-mum because of the following reasons:

1) Reduced measurement time (personnel cost). The com-plete acoustic spectrum (mostly 50 Hz to 10 kHz) canbe measured instantaneously.

2) No special analyzer is needed; the soundcard of a PCcan be used.

3) The same accessories (mountings, windscreen and soon) as a 1⁄2-in microphone can be used; it is simply one


Fig. 4. g temperature distribution due to convection for twoheaters; both heaters have same temperature; ––– sum of twosingle temperature functions; a temperature difference occurs.

output

∆T

[k]

velocity

distance 0

upstream downstream


1⁄2-in probe.4) The (open-source) software for data acquisition is avail-

able as freeware.5) Simple low-cost calibration is possible, with no need

for special equipment.

2.1 Design ConsiderationsIn the beginning only a specialized 1⁄2-in Microflown

was used to perform the p-u measurements, as shown inFig. 5. The combination of this bowed version of a 1⁄2-inprobe and a standard 1⁄2-in microphone seemed to be themost cost-effective realization of a low-cost intensityprobe. Those interested in sound intensity are likely toown a 1⁄2-in microphone, so only the 1⁄2-in Microflown isneeded to create an intensity probe. Although this soundslogical, it seemed not to be the right way to go. It wasfound that the fabrication of a holder to position theMicroflown and the microphone in a sound field intro-duced mechanical problems, and standard accessoriessuch as windscreens did not fit. Hence a 1⁄2-in probe con-taining a Microflown and a miniature microphone wasrealized (see Figs. 6 and 7). Although the miniature micro-phone introduced new costs (it must be fitted into the 1⁄2-inMicroflown) and called for new calibration efforts (nowboth Microflown and miniature microphone need to becalibrated), this realization proved to be most convenient.Standard 1⁄2-in accessories can be used, and the miniaturemicrophone can be calibrated at the same time as theMicroflown. This type of realization will determine soundpressure and particle velocity at the same position.

3 CALIBRATION

The calibration of acoustic sensors is difficult becauseof the acoustic environment is not known and a referenceparticle velocity sensor is not available. Several calibra-tion techniques can be used, each with its limitations.These calibration techniques include

1) A standing-wave tube2) An anechoic environment3) A reverberant environment4) Electrical characterization.

3.1 Standing-Wave TubeThe standing-wave tube is preferred because the

acoustic impedance inside the tube is well understood, andthis calibration yields the sensitivities of both the pressureand the particle velocity sensor against a known (cali-brated) pressure microphone with little effort. It is possi-ble to fit the sensitivity on the maxima in the calibrationcurves, or we can compare the measurement against amodel of the standing-wave tube, including the effects ofviscosity. The latter method also yields the phase of thepressure and velocity sensors. The phase can also bederived from the mean of two measurements as the probeis rotated by 180 deg, measuring in the opposite direction.The latter method is allowed since the total time averageintensity in the tube is zero.

The disadvantage of the standing-wave tube is its lim-ited frequency range. Due to the physical size of our 1⁄2-inpackage we are not able to use a standing-wave tube thatfunctions above about 4 kHz. The lowest frequencydepends on the length of the standing-wave tube, so twotubes should be used if low frequencies (<100 Hz) are tobe implemented correctly.

Above the cutoff frequency the specific acoustic imped-ance changes due to the existence of standing waves per-pendicular to the direction of the sound wave. The setupwill then be very difficult to use since the acoustic imped-ance is no longer real and constant. For a tube this cutofffrequency is given by [8].

.f

d

c

1 71c (5)

where d represents the diameter of the tube and c is thespeed of sound (approximately 330 m/s). The sound wavecan only travel in one direction. In a standing-wave tube,a rigidly terminated tube with rigged sidewalls, all the


Fig. 7. Cross section of 1⁄2-in sound intensity probe.Fig. 6. 1⁄2-in sound intensity probe.

Fig. 5. Sound intensity probe consisting of 1⁄2-in Microflown withbowed mounting and 1⁄2-in microphone.


sound will be reflected at the end of the tube.The specific acoustic impedance inside the standing-

wave tube can be calculated by solving the wave equation.The fluid is excited by a piston (or a loudspeaker) withamplitude U at the left end and is terminated by a rigidboundary at the right end (see Fig. 8). The sound pressureand the particle velocity at any place in the tube are givenby the equations

.

sin

cos

sin

sin

ρip x cUkl

k l x

u x Ukl

k l x

0^^

^

^^

^

hh

h

hh

h

8

8

B

B

(6)

A good place to put a reference microphone is at the endof the tube since there the sound pressure is at its maxi-mum. Thus, pref (x l) pref (l) ≡ pref.

If a p-u sound intensity probe is put at a certain positionx in the tube, the relationship between the pressure micro-phone of the p-u sound intensity probe and the reference(pressure) microphone at the end of the tube is given by

cosp

pk l x

ref

probe^ h8 B (7)

where pprobe pprobe(x). This relation turns out to be a sim-ple cosine function. The distance l x can easily beobtained by measuring two minima of the cosine function.

Analogously, almost the same applies for the particlevelocity,

sinρ

ip

u

ck l x

ref

probe

0^ h8 B (8)

where uprobe u(x). The relationship between the particlevelocity and the reference sound pressure at the end of thetube turns out to be a simple sine function. The phase shiftbetween them equals 90 deg.

It shows that when calibrating the p-u probe to obtainthe sound intensity only the phase mismatch ϕp ϕushould be determined and not the individual phase mis-match of each probe. The p-u probe phase mismatch ϕp ϕu can be determined by measuring the ratio of the parti-cle velocity and the sound pressure in the tube, given by

.tan arg degρ

ip

u

ck l x

p x

u x90

probe

probe

0

"^^

^h

h

hR

T

SSS

8

V

X

WWW

B

(9)

This relation shows that in a standing-wave tube particlevelocity and sound pressure are 90 deg out of phase. In

this way the p-u probe is phase calibrated in the arrange-ment as it is used: the pressure microphone is positionedface to face with the Microflown.

In practice the sound in the tube is slightly damped.This is caused by two effects, the viscosity or stickiness,of the air and the absorption in air. The particle velocity iszero at the boundaries of the tube. In the middle of thetube the sound wave propagates as a plane wave. The fric-tion in the air between zero velocity and the plane wavecauses damping. This effect is noticed most at low fre-quencies. At high frequencies the damping of the air itselfwill be noticeable. If the standing-wave tube is properlydimensioned, the effects will be hardly noticeable.

If thermal viscous effects (damping) are taken intoaccount, Eq. (8) will turn into Eq. (10) [9],

sinh

ρ Γ

Γip

u

c

k x l

ref

probe

0

^ h8 B(10)

where Γ is the viscothermal wave propagation coefficient,given by

.Γγ σ

ii

s2

1 1

σe o (11)

Here γ 1.4 is the ratio of the specific heat of air, σ 0.845 the square root of the Prandtl number σ2, and s is theshear wave number,

ωµρ

sd

d f2

346 (12)

where µ 17.1 106 Pa · s is the dynamic viscosity andρ0 1.3 kg/m3 the density. To use this model, the shearwave number must be much larger than unity, which istrue in practical cases. The viscothermal wave propagationcoefficient can be simplified to

.Γ ii

d f332

1

(13)

As can be seen, for higher frequencies or large diameters,Γ will reach i and the simple model may be used.

The diameter and length of the tube should be deter-mined in such way that the factor 332d f is as large aspossible. For low frequencies we use a tube of 160-mmdiameter and 8-m length. In this case, at 20 Hz the effectof damping is not noticeable. The tube can be used up to1 kHz due to its corner frequency. For higher frequencieswe use a tube that is 45 mm in diameter and 1 m long. Thistube is used in a 50-Hz to 4-kHz bandwidth. For 1⁄2-inprobes it is not possible to use smaller tubes.

Although the thermal viscous effects are relatively small,they are observed most evidently in the phase response. Thephase response in the tube should be 90 deg. However,the measured response shows “rounded edges” (see Figs. 9and 10). This is due to these thermal viscous effects.

3.2 Anechoic EnvironmentAnother method is calibration in an anechoic environ-

ment. For this purpose we use a small 1-m3 box which is


Fig. 8. Tube rigidly terminated at x l, in which fluid is drivenby a vibrating piston at x 0.

x l=

)lx( −

( )u x

x = 0

Probe

pref)lx(

u x


acoustically damped for frequencies above about 1 kHz. Boththe standing-wave tube and the anechoic measurements canbe combined to achieve a useful calibration in the range of 10Hz to 12 kHz, where the upper frequency mainly relies on theeffort made in lining up the sensors in the anechoic cabinet.Also if weather conditions are good, we can use an open win-dow as an anechoic environment with surprisingly goodresults. As an alternative for the anechoic room, we carriedout measurements with the sound probes situated outside anopen window. The results are remarkably consistent.

3.3 Reverberant RoomThe last technique is reverberant room calibration.

Using a single sound source in a reverberant room and thep-u probe, we are able to determine how much sound isdirect and how much belongs to the reverberant soundfield. Ideally this is diffuse sound [10], [11]. From meas-urements of the particle velocity in the direction of thedirect sound wave, combined with measurements of thesame velocity in the perpendicular direction (only dif-fuse/reverberant sound is measured with the particlevelocity sensor), we can calculate the ratio between directand reverberant sound, knowing that in case of a diffuse

sound field only one-third of the power in the diffusesound field is measured with the particle velocity probedue to the directivity (cos θ) of the Microflown,

(14)u u u

u u

13

13

;; dir rev

rev9

2 2 2

2 2(15)

where u ;;2 and u9

2 are the auto spectra of the measured par-ticle velocities in the two directions in V2 or m2s2 if thesensitivity of the particle velocity probe is known. Notethat the units are not important here since ratios are used.We can rewrite Eqs. (14) and (15) in terms of the ratio ofdirect (free field) to reverberant sound as

u

u

u

u u

3

;;

rev

dir

9

92

2

2 2J

L

KK

N

P

OO (16)

and

.u u u ;;dir 92 2 2 (17)

Using a microphone, all power in both the direct andthe reverberant fields is measured if the microphonemeasures unidirectional and the reverberant sound is


Fig. 10. Calibration measurement of Microflown phase. ––– phase response in the standing-wave tube for two Microflown orientations;––– average phase response; g model.

100 1000

-200

-100

0

100

Frequency [Hz]

Pha

se [D

EG

]

Fig. 9. Calibration measurement of miniature microphone phase in a standing-wave tube; g model.

100 1000-300

-200

-100

0

100

Frequency [Hz]

Pha

se [

DE

G]


uniformly diffuse,

p p p dir rev2 2 2 (18)

where p 2 is the auto spectrum of the measured pressureand p 2

dir and p 2rev that of the direct and reverberant

pressures.

.p pu

u1 dir

rev

dir2 2

21

J

L

KK

N

P

OO

R

T

SSSS

V

X

WWWW

(19)

Since the microphone is unidirectional, we can use thisratio to calculate the pressure on site with the p-u probecaused by the direct sound. Assuming that all soundwaves will behave as plain waves, we can conclude thatthe ratio between direct and reverberant particle veloc-ities equals the ratio between direct and reverberantpressures,

.p

p

u

u

rev

dir

rev

dir (20)

The combination of the pressure due to the direct soundfield, and the particle velocity due to the direct sound fieldyields the sensitivity of the particle velocity sensor. Thepressure sensor can be calibrated for phase and sensitivityagainst a known (calibrated) pressure sensor if the direc-tivity is comparable.

Combining the pressures and particle velocities of thederived direct sound we are able to calibrate the particlevelocity sensor sensitivity Su* [V/Pa*] or Su [Vs/m] bydividing by ρ0c,

(21)

.ρ

Sp

uS

S cp

uS

dir

dir

dir

dir

u p

u p

2

2

0 2

2

*

(22)

Calibrating the phase between pressure and particle veloc-ity sensors is more difficult in a reverberant environmentbecause the sound field is not known. The methoddescribed yields only the auto spectra of the free-fieldpressure and particle velocity.

In a purely diffuse sound field there is no acoustic powerin the plane perpendicular to the direct sound field becausea purely diffuse sound does not contain net power, and thuspressure and particle velocity are out of phase by 90 deg.Secondary acoustic paths, however, such as floors, ceilings,walls, and furniture, can cause some acoustic power in theplane perpendicular to the direct sound field. Avoiding thesesecondary paths as much as possible, it should be possibleto measure the phase between pressure and particle velocitysensors. Using a pulse and time-window measurement it ispossible to determine the direct sound anechoic but in mostpractical situations the distances are short, which leaves lit-tle time for measurements and thus results in a very poorfrequency resolution.

The phase may be determined in the following manner.

Using the method described previously, the free-field autospectra of the sound pressure and the particle velocity aredetermined. Thus from the product of free-field pressureand velocity the sound intensity can be calculated. On theother hand, from the product of the auto spectra of themeasured pressure and velocity and the cosine of thephase of the sound field the sound intensity is obtained.Combining the phase of the sound field with the phasebetween the measured pressure and particle velocity, weshould be able to calculate the cosine of the phase betweenthe two acoustical sensors. In this way at least the activepart of the sound intensity can be measured using thesesensors.

3.4 Microflown Electrical CharacterizationThe acoustical sensitivity and the frequency behavior of

a Microflown sensor can be found using only electricalmeasurements. It can be shown [12] that there is a rela-tionship between the output of the wires (see Fig. 2) whenthe sensor is placed in an electronic setup, and the acousti-cal frequency response of the sensor. This electronic cali-bration method is a fast, relatively uncomplicated way tocharacterize the Microflown’s sensitivity behavior and caneasily cover a wide frequency range. Nearby acousticobstacles that might be used to enhance the signal, such asthe package gain in the p-u probe, are not taken intoaccount. Therefore the electrical characterization is notfurther discussed in this paper.

3.5 Calibration Data Acquisition HardwareCalibration is likely to be the most essential part of

this measurement technique because the results differgreatly depending on the calibration. Several parts in thesystem should be calibrated in order to be able to meas-ure accurately. We should calibrate the pressure and par-ticle velocity sensors and the analog-to-digital converterin the soundcard or an external analog-to-digital con-verter. It is also possible to introduce a preamplifier sothat we use the resolution of the analog-to-digital con-verter more efficiently.

The analog-to-digital converter can easily be calibratedusing a known signal because we can assume the sound-card used to be almost perfectly linear in the frequencydomain. A preferably white or pink signal can be con-nected to the inputs with known power. With little dataacquisition and processing, the sensitivity and a phase dif-ference between the channels can be calculated and imple-mented in the software. Also a sinusoidal signal can beused, although this will only yield the calibration for a sin-gle frequency. We should note that the gain and panning ofthe input should not be altered on the personal computersince the gain and panning will be applied on the acquiredsignals. If a preamplifier is being used, it also should notbe adjusted after calibration. The preamplifier can be cal-ibrated using the same method used for the soundcard.Note that depending on the quality of the soundcardrespectively, the analog-to-digital converter, the phasebetween two channels can be extremely large, for exam-ple, up to 140 deg at 20 kHz for a standard PC soundcard,or only 2.5 deg for an external analog-to-digital converter.



4 SOFTWARE

The software is programmed in Matlab around a graph-ical user interface. Matlab is widely used and is an easyprogram language so that users are able to modify the soft-ware according to their needs. The program is capable ofworking in three different modes. The instantaneous modeis very useful in examining the acoustical situationwhereas the other modes are intended for measurements towave files and data processing from wave files. In theinstantaneous mode it is possible to compute and showvarious windows containing:

• The auto spectra of the pressure and particle velocities• The coherence between the two signals• The intensity (active, reactive, and phase)• The sound energy• The active and reactive intensity divided by the energy

(indicating a new power transport velocity and reflection)• The acoustic impedance (magnitude, real and imaginary

parts).

All axes can be chosen as linear or logarithmic scales,and the vertical axis is by default in dB. Frequency spec-tra can be displayed in terms of frequency (per Hz) andin one-third-octave bands. Calculated spectra can besaved to the work space, or as Matlab.mat files and asASCII files for further processing. It should be noted thatthe one-third-octave band of 20 kHz is an underestima-tion in case a sample frequency of 44.1 kHz. is used.

This underestimation is due to the fact that the power isnot known in the whole one-third-octave band, and theanalog-to-digital converter will be less accurate above 20kHz. If sample frequencies of 48 kHz and higher areused, the 20-kHz one-third-octave band will be com-puted correctly.

In the instantaneous mode it is not convenient to uselarge frames because of computing time and power,resulting in a small frequency resolution. The default sizeof the frame is set at 2048. For higher resolution it isadvisable to save the signals into wave files so that theycan be processed later with higher frequency resolution.We should note that a good frequency resolution is nec-essary if we want to use the advantage of low-frequencymeasurements, which is possible due to the physics of thep-u probe. The frame size for the fast Fourier transformcan be altered according to the required frequency reso-lution and computation time (default 16834). By defaulta Hanning window is used in order to increase the spec-tral resolution.

It is possible to record measurements using externalhardware such as analog-to-digital converters connectedto digital soundcards, CD recorders, or DAT recorders.This is also very convenient since most of these are quitemobile as measuring becomes rather easy and postpro-cessing can be automated with the use of the availablesoftware.

5 DATA PROCESSING

Data processing is performed on the raw data accordingto the flowchart given in Fig. 11. The left-hand signal con-tains data from the pressure sensor, whereas the right-handsignal contains the data from the particle velocity sensor.First the signals are converted to the frequency domainbecause in this domain it is much easier to correct for thesensitivities and phases in the signals than it would be inthe time domain. Second, the data are converted from arbi-trary units to volts by means of the calibration of the sound-card (or external analog-to-digital converter)–amplifiercombination. Note also that the phase shift due to thesoundcard is implemented here so that the signals, nowexpressed in volts, are in the right phase with reference toeach other. If the signals are given in volts, the programconverts them to the pressure and particle velocity signalsusing the calibration of the sensitivities and phases of theacoustical probes. The signals are converted to pascals forthe pressure signal, and to meters per second for the parti-cle velocity signal using values for ρ0 and c, which can bealtered in the parameter settings (default 1.219 kg/m3 and343 ms). From the pressure and particle velocity signalsthe various auto and cross spectra are calculated (Gpp, Guu,Gpu).

The sound intensity in a sound wave is the product ofpressure and particle velocity. This product is calledinstantaneous intensity. Sound intensity is defined as themean of the integral of the instantaneous sound intensityintegrated over all times. In practical cases it is never pos-sible to integrate over all times, and it is convenient tointegrate from t T to t T or from t 0 to t T. The


Fig. 11. Schematic representation of software.

Make wav -file30 sec 44.1kHz Left=pressure, right=velocitystereo 16bit

DisplayAuto spectrum pressure & velocity

Phase pressure and velocityCoherence between channels

Active & reactive intensitySound energy

Active & reactive intensity/ sound energyReal, Imaginary & Magnitude impedance

Save display as: *.txt *.ma *.wav

read sound card read wav file

Model p, model u, model soundcard

ProcessingFFT left & right, correction sound card

correction freq. resp. pressure & velocity correction phase response pressure & velocity

calculate coherencecalculate intensity, Energy and Impedance

Instantaneous


intensity is defined by

.lim dIT

p t u t t1

oT

T

" 3#e ^ ô h h (23)

Eq. (23) can be rewritten as a cross correlation for σ 0,

.lim dIT

p t u t t R1

0

oT

pu

T

" 3#e ^ ^ ô h h h

(24)

Note that the subscripts of the cross spectra are not useduniquely. In this text a definition often found in signal pro-cessing is used [13]. The advantage of this usage is that thereactive part of the intensity equals the imaginary part of thecross spectrum, whereas otherwise a factor of 1 is neededin order to keep up with the convention [2], [14], [15].

The frequency distribution is given by the Fourier trans-form of the cross-correlation function, which is termedcross-spectral density S,

.expωπ

τ ωτ τi dS R2

1 pu pu

3

3

#^ ^ ^h h h (25)

For practical reasons it is convenient to redefine the spec-tral density function S as a single-sided function of fre-quency G. Because the signals of the pressure and particlevelocity sensors are both real, we can represent only halfthe frequency span without losing any information. Theadvantage of the single-sided spectral density function isthat fewer data have to be stored and computing time andeffort are decreased,

, > .ω ω ωforG S2 0pu pu pu^ ^h h (26)

From this cross spectrum the intensity spectrum is easilyobtained because the active intensity equals the real part ofthe cross spectrum between pressure and particle velocity,whereas the reactive intensity equals the imaginary part,

(27)

.

Re

Im

I G

I G

active

reactive

pu

pu

`

`

j

j (28)

Integrating over the frequency spectra we can calculate theintensity over the frequency spans,

, .dI f f G f pu

f

f

1 2

1

2

f#_ _i i (29)

The program calculates the intensity spectrum both in thefrequency domain and in one-third-octave bands.

The energy density is defined as [2]

.ρ

ρE tc

p tu t

2 2

1

02

2

02

^^

^hh

h (30)

Using the cross spectra calculated earlier, the time-averaged energy density can be calculated as

.ρ

ρEc

GG

2 2

1

ppuu

02 0 (31)

The acoustic impedance is defined as the ratio betweenpressure and particle velocity,

.ZU

P (32)

From the computed cross correlations we can derive theacoustic impedance, which the program computes as

.ZU

P

G

G

uu

pu(33)

6 MEASUREMENTS

6.1 Standing-Wave Tube and AnechoicCalibration

The 1⁄2-in sound probe is calibrated in a short standing-wave tube and a small 1-m3 anechoic chamber. Thestanding-wave tube can be used for frequencies up to 4kHz, the anechoic chamber for frequencies higher than1–2 kHz. To estimate the frequency response, a simpli-fied model is used. Fig. 12 shows the measurements inthe standing-wave tube and the anechoic chamber. At 4-kHz the model underestimates by 4 dB, and for frequen-cies higher than 6 kHz the model overestimates by 3 dB.It is our opinion that we can correct for the resulting errorsmanually later. We use a constant value of 177 deg as amodel for the phase response of the miniature microphone(see Fig. 11).

Measurements of the Microflown sensitivity as a func-tion of frequency are shown in Fig. 13. The model used toestimate the frequency response of the Microflown is

/ *mV PaHf f

6501

25001

35

u

2

2

2

27 A

(34)

A Microflown is sensitive to particle velocity ratherthan sound pressure, so the sensitivity cannot be given inmillivolts per pascal (mV/Pa). As a reference we use thesensitivity of a Microflown in mV/Pa*, where Pa* repre-sents 1/ρc m/s.

The phase of the Microflown can also be determined in


Fig. 12. Calibration measurement of miniature microphone instanding-wave tube and small anechoic room. – – – simplifiedmodel: Hp 22.3 ( f 2/3 kHz2 1)0.5 ( f 2/8 kHz2 1)0.5 (25Hz2/f 2 1)0.5 mV/Pa.

100 1k 10k-70

-60

-50

-40

-30

-20

Frequency [Hz]

Out

put r

e. 1

Pa

[dB

]


the standing-wave tube. Two measurements were madewhere the microphone was rotated by 180 deg so that themean phase of the two sound fields in the standing-wavetube equaled zero (see Fig. 10). The phase was fitted by

. .arg tanHf

1 52914

u1

# - (35)

6.2 Calibration Measurements––Open Windowand Reverberant Room

Although the frequency and phase responses are knowndue to previous calibration measurements, we presentalternative methods. The advantage of these alternativemethods is that no special facilities are needed.

With the acoustic sensors placed outside an open win-dow we can create an anechoic sound field. Depending onthe weather conditions, this can be very effective. Whiletaking the measurements a breeze was present (Figs. 14–16). This is visible below 100 Hz. The use of windscreenswill account for this, but the distance between the p-uprobe and the reference microphone should not be thatlarge.

We also applied the so-called reverberant room calibra-tion technique. The two signals of the p-u probe weremeasured while a reference microphone was positioned asclose as possible. The sensitivity of the pressure was cal-culated using the transfer function between the two micro-

phones and the sensitivity of our reference microphone.The sensitivity of the particle velocity probe was obtainedusing the calculated auto spectra of the free-field pressureand particle velocity, according to the method presented,combined with the sensitivity of the microphone used.

The calculated sensitivities of both the pressure sensor(Fig. 17) and the particle velocity sensor (Fig. 18) wereconsistent with the models obtained from the standing-wave tube and the anechoic measurements. Above 10 kHz


Fig. 15. Calibration measurement of Microflown amplituderesponse measured with open window method. g model.

100 1k 10k-80

-60

-40

-20

Frequency [Hz]O

utpu

t re.

1P

a* [d

B]

100 1k 10k-40

-35

-30

-25

-20

Frequency [Hz]

Out

put r

e. 1

Pa

[dB

]

100 1k 10k-50

-40

-30

-20

Fre

Out

put r

e. 1

Pa

[dB

]

quency [Hz]

100 1k 10k-70

-60

-50

-40

-30

Frequency [Hz]100 1k 10k

-70

-60

-50

-40

-30

Frequency [Hz]

Out

put r

e. 1

Pa*

[dB

]

100 1k 10k-300

-200

-100

0

100

Phas

e [D

EG]

Frequency [Hz]

Fig. 14. Calibration measurement of microphone amplituderesponse measured with open window method. g model (seeFig. 12).

Fig. 17. Calibration measurement of miniature microphoneamplitude response measured with reverberate room method;g model (see Fig. 12).

Fig. 13. Calibration measurement of Microflown in standing-wave tube and small anechoic room; g model.

Fig. 16. Calibration measurement of Microflown phase response(top) and miniature microphone (bottom) measured with openwindow method; g model. Microflown, ––– model microphone.


the sensitivities calculated from the reverberant measure-ments were no longer valid (see Fig. 18) because noacoustic power was available in the sound field since amidrange loudspeaker was used (see also Fig. 19).

6.3 Intensity MeasurementsIntensity measurements are performed in a reverberant

room using a p-p probe and the p-u probe described. Thep-p probe is a Brüel & Kjær intensity probe, type 2260Investigator, using two 1⁄2-in microphones (type 4181).Measurements were made twice, using spacers of 50 mm(type UC 5270) and 12 mm (type UC 5269) so that wecould measure up to one-third-octave center frequenciesof 1.25 kHz and 5 kHz, respectively. The intensities in theone-third-octave frequency bands are calculated using theInvestigator B&K 2260.

The signals of the p-u probe were measured using a per-sonal computer combined with a digital soundcard and anexternal analog-to-digital converter as well as a preampli-fier set at about 40 dB for both channels. The signalswere acquired using Cool Edit Pro. A white noise of 5.1mV rms was applied in order to calibrate the hardware.The wave file created was read into Matlab and the power

was calculated so that we were able to determine the sig-nal in volts for our calculations. As a result of an imped-ance mismatch we had to correct for a change in sensitiv-ities of the two sensors. If the acoustical sensors arecalibrated using the same data acquisition hardware, thisis not necessary. Using the software described, we calcu-lated the active intensity in one-third-octave bands, asshown in Figs. 19 and 20.

First the intensity in front of a small loudspeaker wasmeasured using the p-p and p-u techniques. In Fig. 19 wenotice that the intensity measured with the p-u methodslightly overestimates the measurements of the p-pmethod in all frequency bands. The cause of this overesti-mation is very likely to be found in the calibration of theacoustical sensors or the preamplifier–soundcard combi-nation, or both. A more accurate calibration of the acousti-cal sensors and the data acquisition hardware, includingthe p-p probe, should minimize this discrepancy.

It should be noted that below 150–250 Hz the loud-speaker radiates almost no acoustic power and the meas-ured intensities are mainly from the background. Thenoise levels in the p-p measurement mainly cause the dis-crepancies in the results for frequencies below 100 Hz.


Fig. 19. Measured sound intensity of simple loudspeaker. p-u probe (black); p-p probe 50 mm (dark gray); p-p probe 12 mm (lightgray).

Fig. 18. Calibration measurement of Microflown amplitude response measured with reverberate room method; ––– model.

0

10

20

30

40

50

60

70

25 50 100 250 500 1k 2k 4k 8k 16k

Frequency in 1/3 octave bands

Sou

nd In

tens

ity [d

B] puprobe

pp 50mmpp12mm

100 1k 10k-80

-70

-60

-50

-40

-30

-20

Frequency [Hz]

Out

put r

e. 1

Pa*

[dB

]


The intensities calculated using the p-u method for fre-quencies above 10 kHz overestimate the actual intensitybecause the model for the microphone sensitivity is nolonger valid (see Figs. 12 and 17).

Second, we performed intensity measurements inbetween two identical loudspeakers, which were the sameas described before. The power to loudspeaker 1 (S1) waskept at a constant level, whereas the power to loudspeaker2 (S2) was varied. A linear relationship was expectedbetween the active intensity and the ratio between theacoustic powers of both loudspeakers [16]. The minussign indicates that the power is coming from the directionof loudspeaker 2. From the measurement at, for example,1 kHz (see Fig. 20) we can note that this is the case forboth measurement techniques. We notice a small overesti-mation for the p-u measurement, although the deviation inthe p-p measurements using 12-mm and 50-mm spacers isof the same order.

6.4 Signal-to-Noise Ratios of Both MethodsThe self-noise of two separate microphones is not cor-

related. In the case of the p-p method, the imaginary partof the cross spectrum is taken to get the active intensity; inthe case of the p-u method, the real part of the cross spec-trum is used to obtain this figure. In either case the crossspectrum is used, and two noncorrelating sources (theself-noise of the microphones) will result theoretically ina zero output. Practical measurements have been per-formed for the p-u probe (Fig. 21), and the p-p probe used(Fig. 22). Both probes were placed in a completely silentenvironment and the intensity output was measured.

7 DISCUSSION AND CONCLUSIONS

Intensity can be measured using the p-u probe and aconventional soundcard, as shown for the measurements


Fig. 21. Measured noise levels of 1⁄2-in p-u probe. Noise levels ofsound pressure microphone (–––) in dB SPL (re 20 106

Pa)/Hz1⁄2, Microflown (g) in dB PVL (re 20 nm/s)Hz1⁄2 and soundintensity signal (– – –) in dB SIL (re 1 pW)/Hz1⁄2.

Fig. 22. Measured noise levels of 1⁄2-in p-p probe. Noise levels ofsound pressure microphone (g) in dB SPL (re 20 106

Pa)/Hz1⁄2, sound intensity signal (–––) in dB SIL (re 1 pW)/Hz1⁄2.From frequencies above 5 kHz 1⁄4-in microphones should be usedso that self-noise of intensity determination as displayed herewill no longer be valid. For frequencies above 10 kHz no inten-sity measurement can be performed.

Fig. 20. Sound intensity measurements between two identical loudspeakers at 1-kHz one-third-octave band versus ratio of electricpower sent to loudspeakers.

100 1k 10k-40

-20

0

20

40

F

u

I

p

Noi

se le

vel i

n dB

/ S

qrt (

Hz)

requency [Hz]

100 1k 10k-40

-20

0

20

40

F

Noi

se le

vel i

n dB

in 1

Hz

band

wid

th

requency [Hz]

-300

-200

-100

0

100

200

300

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

S2/S1 [-]

Sou

nd In

tens

ity [r

e. 1

0 W

m]

-9

-2

pupp 50mmpp 12mm


performed. The measured active intensity is somehowoverestimated compared to the p-p measurements, butadequate hardware and calibration should account for this.

The main advantage of the p-u probe versus the p-pprobe is its low cost of operation. Measurements over thewhole frequency range can be made at the same setting,without changing the probe itself. The main disadvantageis the calibration of the p-u probe. Because the two signalsmeasure two different acoustical properties, it is not pos-sible to use the same method for both sensors unless thesound field is known well. Some calibration methods arepresented with their own accuracy and simplicity.

The combination of p-u probe and software creates anaffordable and easy-to-use intensity measuring device.Other acoustical parameters, such as reactive intensity,phase of sound field, energy density, and acoustic imped-ance, are being calculated without extra effort becausethey are also obtained from the auto and cross spectra ofthe pressure and particle velocity signals.

8 ACKNOWLEDGMENT

The authors wish to thank the Dutch TechnologyFoundation STW for financial support.

9 REFERENCES

[1] H. F. Olson, “System Responsive to the EnergyFlow of Sound Waves,” US patent 1,892,644 (1932).

[2] F. Fahy, Sound Intensity (E&FN Spon, London,1989).

[3] H. E. de Bree, P. Leussink, T. Korthorst, H. Jansen,T. Lammerink, and M. Elwenspoek, “The Microflown,”Sensors and Actuators A––Phys., vol. 54, pp. 552–557(1996).

[4] H. E. de Bree, “The Microflown,” Ph.D. thesis,University of Twente, The Netherlands (1997).

[5] H. E. de Bree, T. Lammerink, M. Elwenspoek, andJ. H. J. Fluitman, “Use of a Fluid Flow-Measuring Deviceas a Microphone and System Comprising such a

Microphone,” patent PCT/NL95/00220 (1995).[6] J. W. van Honschoten, G. J. M. Krijnen, V. B.

Svetovoy, H. E. de Bree, and M. Elwenspoek, “Optimi-sation of a Two-Wire Thermal Sensor for Flow and SoundMeasurements,” in Proc. MEMS (Interlaken, 2001), pp.523–526.

[7] V. B. Svetovoy and I. A. Winter, “The µ-Flown,”Sensors and Actuators, vol. 86, pp. 171–181 (2000).

[8] A. Hirschberg and S. W. Rienstra, “An Introductionto Acoustics,” IWDE Rep. 92-06, version 1/06/92 (1992Feb.).

[9] J. W. van Honschoten, H. E. de Bree, F. J. M. vander Eerden, and G. J. M. Krijnen, “The Influence of Visco-thermal Effects on Calibration Measurements in a Tube,”presented at the 109th Convention of the Audio Engin-eering Society, J. Audio Eng. Soc. (Abstracts), vol. 48, p.1095 (2000 Nov.), preprint 5182.

[10] R. Raangs, W. F. Druyvesteyn, and H. E. de Bree,“A Novel Two-Dimensional Sound Particle Velocity Probefor Source Location and Free Field Measurements in aDiffuse Field,” in Proc. Internoise (The Hague, 2001), pp.1919–1922.

[11] W. F. Druyvesteyn, H. E. de Bree, and M.Elwenspoek, “A New Acoustic Measurement Probe, theMicroflown,” in Proc. IOA (London, 1999), pp. 139–149.

[12] J. W. van Honschoten, V. B. Svetovoy, T. S. J.Lammerink, G. J. M. Krijnen, and M. C. Elwenspoek,IEEE Conference on Sensors (Orlando, FL., 2002).

[13] J. A. Cadzow, Foundations of Digital Signal Pro-cessing and Data Analysis (Macmillan, New York, 1987).

[14] F. J. M. van der Eerden, “Noise Reduction withCoupled Prismatic Tubes,” Ph.D. thesis, University ofTwente, The Netherlands (2000).

[15] Brüel & Kjær Tech. Rev. no 3, 4 (1982), 4 (1985),4 (1986), 1 (1996).

[16] W. F. Druyvesteyn and H. E. de Bree, “A NovelSound Intensity Probe; Comparison with the Pair of Pres-sure Microphones Intensity Probe,” J. Audio Eng. Soc.(Engineering Reports), vol. 48, pp. 49–56 (2000Jan./Feb.).


Ron Raangs received a B.Sc. degree in physics fromthe Polytechnic High School of Enschede, TheNetherlands. After that he studied experimental physics

at the University of Nijmegen, The Netherlands. He dis-covered the interesting field of acoustics and thus wrotehis thesis on biological applications of trace gas detection

R. Raangs W. F. Druyvesteyn H. E. de Bree

THE AUTHORS



using photo-acoustics in 1996. Since he was still veryinterested in and attracted to acoustics and music, hestarted studying for his Ph.D. degree at the University ofTwente in 2000. His main interests include Microflownand its usage, 3-D sound intensity probes, acousticalsource location, digital signal processing, and combina-tions of the above.

W. F. Druyvesteyn completed his studies in physics atthe Technical University in Delft, The Netherlands, at theend of 1961. In 1962 he joined the Philips ResearchLaboratories in Eindhoven, where he worked on super-conductivity especially on hard super conductors. Hewrote a thesis on this subject in 1965. From 1965 to 1970

he studied helicon waves in pure metals. He later workedin the area of magnetic bubbles and magnetic recordingheads. Since 1983, Dr. Druyvesteyn has concentrated onacoustics. In 1998 he retired from the Philips ResearchLaboratories and became professor at the University ofTwente in The Netherlands.

H. E. de Bree completed his studies in electronics atthe Technical University in Enschede, The Netherlands,in 1994. In that year he invented the Microflown and twoyears later he received his Ph.D. degree for his work onthe device. Since then he has been working on the devel-opment of the Microflown at the Mesa ResearchInstitute.

ENGINEERING REPORTS

0 INTRODUCTION

On 2001 November 29 the DAB Subcommittee of theNational Radio Systems Committee (NRSC) adopted anevaluation report on iBiquity Digital Corporation’s FM-band in-band on-channel digital audio broadcasting(IBOC DAB) system. On 2002 April 6 it adopted an eval-uation report on iBiquity’s AM-band IBOC DAB system.This engineering report summarizes the DAB Subcom-mittee’s findings concerning the audio quality ofiBiquity’s IBOC DAB systems, as well as the audio qual-ity of the existing AM and FM systems.

1 BACKGROUND

The NRSC is jointly sponsored by the ConsumerElectronics Association (CEA) and the National Associ-ation of Broadcasters (NAB). It has been studying IBOCDAB since the early 1990s, when broadcasters in theUnited States first proposed this technology for the terres-trial radio broadcast bands. The United States has a verycompetitive commercial broadcast industry. The geo-graphic location and maximum authorized power of aradio broadcast station have a significant impact on thecompetitive position of that station. When consideringmethods to convert their transmission systems to digital,many members of the competitive commercial broadcastindustry in the United States have been very concernedabout upsetting the competitive balance between existingstations. They much prefer a system that will maintain thestatus quo, in terms of service area, to the greatest extent

possible. Thus the IBOC DAB system, which permitsbroadcasters to use their existing transmission facilities,has been the preferred choice for most U.S. broadcasters.

The DAB Subcommittee approved laboratory and fieldtest procedures for testing IBOC DAB systems in 2000and 2001. Using these test procedures, testing of iBiquityDigital Corporation’s AM-band and FM-band IBOC DABsystems was performed in 2001. These tests were carriedout by iBiquity and its contractors under the supervisionof the NRSC. iBiquity submitted a report on the testresults for its FM-band IBOC DAB system to the NRSC’sDAB Subcommittee in 2001 August, and it submitted areport on the test results for its AM-band system to thesubcommittee in 2002 January [1], [2]. The DABSubcommittee adopted evaluation reports on the FM-bandand AM-band systems on 2001 November 29 and 2002April 6, respectively [3], [4].

The laboratory and field tests, together, were designed todetermine whether or not the digital audio quality of theAM-band and FM-band IBOC DAB systems is a signifi-cant improvement over the audio quality of today’s analogsystems. They were also designed to predict the degree towhich the addition of IBOC DAB digital signals to the AMand FM bands would degrade the reception of analog AMand FM broadcasts on today’s receivers, and whether ornot this interference would cause listeners to stop listening.

Objective and subjective audio quality tests were per-formed in the presence of increasing co- and adjacentchannel interference, and under impaired reception condi-tions such as impulse noise and multipath reception. Eachtest was performed with no IBOC DAB signal present,and again with one or more IBOC DAB signals present, inorder to compare today’s audio quality with that whichcan be expected with the introduction of IBOC DAB.


Industry Evaluation of In-Band On-Channel DigitalAudio Broadcast Systems*

DAVID WILSON, AES Member

Consumer Electronics Association, Arlington, VA 22201, USA

The National Radio Systems Committee’s testing and evaluation program for in-band on-channel digital audio broadcast systems is described. The results of laboratory and field testsperformed during 2001 on iBiquity Digital Corporation’s AM-band and FM-band IBOCDAB systems are reported. The conclusions drawn from the laboratory and field test resultsare also reported, and implications for the future are discussed.

* Presented at the 113th Convention of the Audio EngineeringSociety, Los Angeles, CA, 2002 October 5–8.

ENGINEERING REPORTS EVALUATION OF DIGITAL AUDIO BROADCAST SYSTEMS

The NRSC concluded that the AM-band and FM-bandIBOC DAB systems do provide digital audio that is sig-nificantly improved over today’s analog AM and FM radiosystems. It also concluded that IBOC DAB will have alimited negative impact on listeners’ ability to receiveusable analog audio. It recommended that the implemen-tation of AM-band IBOC DAB be initially restricted todaytime use only pending further investigation into theimpact of IBOC DAB on nighttime reception.

2 FM AND AM IBOC DAB SYSTEM DESIGN

The iBiquity FM IBOC hybrid system has a pair of dig-ital sidebands, one above the host analog signal and onebelow it [5]. In service mode MP1, which was the modetested by the NRSC, each of these sidebands is 69 kHzwide. The sidebands are independent, so the system willstill perform when one of them is lost. The digital signalsuse orthogonal frequency division multiplexing (OFDM)modulation, and provide a digital audio data rate of 96kbit/s, plus 3–4 kbit/s of auxiliary data (1 kbit/s plus 2–3kbit/s of opportunistic data). The FM IBOC system alsohas an all-digital mode in which the analog FM signal isreplaced with a digital signal. Neither the auxiliary datacapability nor the all-digital mode of the iBiquity systemwas tested by the NRSC. The FM IBOC system RF spec-tral occupancy is illustrated in Fig. 1.

The iBiquity AM IBOC hybrid system has three pairs ofdigital sidebands [6]. Each sideband is 5 kHz wide. Thetwo sidebands in each pair are independent, so the systemwill still function when one of them is lost. However, botha secondary and a tertiary sideband are needed to producestereo audio. If both are not available, the system will pro-duce mono digital audio as long as a primary sideband isavailable. The digital signals use OFDM modulation, andprovide an audio data rate of 36 kbit/s, plus 0.4 kbit/s ofauxiliary data. At the broadcaster’s option, the audio ratemay be decreased to 20 kbit/s to provide additional auxil-iary data capacity, or raised to 56 kbit/s with reduced errorcorrection. Only the 36-kbit/s audio rate was tested by theNRSC. The iBiquity AM IBOC system also has an all-digital mode in which the analog AM signal is replaced bya digital signal. Neither the auxiliary data capability northe all-digital mode of the iBiquity system was tested bythe NRSC. The RF spectral occupancy of the hybrid AMIBOC system is illustrated in Fig. 2.

3 NRSC TEST PROGRAM

The NRSC test program for IBOC DAB was designedto test both the performance and the compatibility of thehybrid IBOC DAB systems. The performance tests deter-mined the performance of the digital audio signals interms of coverage and signal quality. The compatibilitytests determined the effect that the IBOC digital signalswould have on the existing analog signals in the AM andFM broadcast bands. Both the performance and the com-patibility tests were conducted using objective and subjec-tive evaluation criteria. Also, most tests were conductedboth in the field and in the laboratory.

The objective tests typically measured quantities suchas audio signal-to-noise (S/N), bit error rate (BER), and/orRF signal level. The subjective tests determined the audioquality as perceived by typical (not expert) listeners. Forthe subjective tests each audio cut was reviewed by at least40 listeners, and a median score and confidence intervalwere determined from these data. The age groups 16–24,25–32, 33–42, and 43–50 were represented approxi-mately equally, and there were an approximately equalnumber of males and females in each group [7].

4 FIRST ADJACENT COMPATIBILITY

First adjacent compatibility is a significant challengefor the AM and FM IBOC DAB systems because theIBOC digital signals of one station share the RF spectrumwith the analog signals of stations one channel above andbelow that station. This is illustrated in Figs. 3 and 4 forthe FM and AM bands, respectively. To minimize the


Fig. 1. FM IBOC system RF spectral occupancy.

-20 dBTAP

-198 -129 +129 +198 kHz

-20 dBTAPAnalog

IBOC IBOC

fc

TAP = total analog power

Fig. 3. FM-band first adjacent channel overlap.

Fig. 2. AM IBOC system RF spectral occupancy.

1st Adj.

AnalogAnalog

IBOC IBOC

-198 -129 fc +129 +198 kHz

Analog-13 dBTAP -13 dBTAP

Pri.

IBOC

Pri.

IBOC

-27

dBTAP

-27

dBTAP

-30.8 dBTAPSec.

IBOC

Sec.

IBOCTer. IBOC

fc +5 +10 +15 kHz-15 -10 -5

TAP = total analog power

WILSON ENGINEERING REPORTS

impact of first adjacent channel interference from theIBOC DAB digital signals to adjacent channel analog sig-nals, the digital signal power level must be limited.

The U.S. Federal Communications Commission (FCC)has always required AM and FM radio stations to be sep-arated by enough distance, geographically, from stationson first adjacent channels so that the stations will notinterfere with one another to an unacceptable degree. TheFCC separation requirements stipulate that AM and FMradio signals should always be at least 6 dB stronger thanany interfering first adjacent channel signals anywherewithin their FCC protected contours [8]. Thus minimizingthe interference from IBOC DAB digital signals to firstadjacent channel AM and FM analog signals that are 6 dBstronger was an important design goal for the IBOC DABsystems.

The first adjacent channel compatibility test results arepresented graphically in Figs. 5 through 36. Data fordesired-to-undesired (D/U) signal ratios that are lowerthan 6 dB (including negative D/U values) represent datapoints that are predicted to lie outside the FCC-protectedcontour of the analog station whose audio quality is beingevaluated. In each of these charts there is a dashed line at2.0 MOS (for rock programming) or at 2.3 MOS (for

speech programming). This represents the point, deter-mined during the subjective evaluation process, where atleast half of the typical listeners would turn off the desiredanalog station due to poor audio quality [9].

The results presented in Figs. 5 and 6 suggest that OEMautomobile radios can be expected to provide listenableaudio well outside the protected contour of an FM stationprogramming rock music in the presence of first adjacentchannel interference. They also suggest that this audio willremain listenable after the addition of IBOC DAB to thefirst adjacent channel interferer, although the analog audiowill be somewhat degraded in terms of perceived quality.

The results presented in Figs. 7 and 8 suggest that OEMautomobile radios can be expected to provide listenableaudio well outside the protected contour of an FM stationprogramming speech in the presence of first adjacentchannel interference. However, they indicate that in manysituations outside the protected contour the analog audioquality will become unlistenable once IBOC DAB isadded to the first adjacent channel interferer. It is interest-


Fig. 5. OEM auto FM rock (field). Fig. 7. OEM auto FM speech (field).

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re

= analog audio with 1st adjacent analog-only interferer

= analog audio with 1st adjacent hybrid IBOC interferer

1 = “bad;” 2 = “poor;” 3 = “fair;” 4 = “good;” 5 = “excellent”

Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

Fig. 4. AM-band first adjacent channel overlap.

1st Adj.

AnalogAnalog

Pri.

IBOC

Sec.

IBO

Sec.

IBO Ter. IBOC

-15 -10 -5 fc +5 +10 +15 kHz

Fig. 6. OEM auto FM rock (lab).


ing to note that the field test results indicate that analogaudio quality begins to improve as the level of the inter-fering signal becomes stronger in the 5- to 10-dB D/Urange. This is because most OEM automobile radios havecircuitry in them that narrows their RF filtering circuitryand automatically switches the audio to mono to reducethe perceived impact of the interference.

The results presented in Figs. 9 and 10 suggest thataftermarket automobile radios can be expected to providelistenable audio well outside the protected contour of anFM station programming rock music in the presence offirst adjacent channel interference. They also suggest thatthis audio will remain listenable after the addition ofIBOC DAB to the first adjacent channel interferer,although the audio will be somewhat degraded in terms ofperceived quality.

The results presented in Figs. 11 and 12 suggest thataftermarket automobile radios can be expected, to someextent, to provide listenable audio outside the protectedcontour of an FM station with speech programming that is


Fig. 9. Aftermarket auto FM rock (field) Fig. 12. Aftermarket auto FM speech (lab).

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

Fig. 8. OEM auto FM speech (lab).Fig. 11. Aftermarket auto FM speech (field).

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

Fig. 10. Aftermarket auto FM rock (lab).


experiencing first adjacent channel interference. They alsosuggest that this audio will become unlistenable shouldIBOC DAB transmissions commence on the first adjacentchannel.

The results presented in Figs. 13 and 14 suggest thatportable radios provide listenable audio at the FCC pro-tected contour when the desired station is broadcastingrock programming in the presence of first adjacent channelinterference, but that reception of listenable audio muchbeyond the protected contour is probably not possible.They also suggest that this would still be the case if IBOCDAB were added to the first adjacent channel interferer.

The results presented in Figs. 15 and 16 suggest that,when the desired station is broadcasting speech program-ming and experiencing first adjacent channel interference,portable radios generally cannot provide listenable audiobeyond the protected contour, except perhaps for a verylimited area just outside the protected contour. Theseresults also suggest that adding IBOC DAB to the firstadjacent channel interferer will have little impact outside

the protected contour, primarily because the perceivedaudio quality is already severely degraded.

The results presented in Figs. 17 and 18 suggest thathome hi-fi receivers can provide listenable audio beyondthe protected contour of an FM station broadcasting rockprogramming, even in the presence of first adjacent chan-nel interference. They also suggest that the addition ofIBOC DAB to the first adjacent channel interferer willhave a minimal impact on the perceived audio qualityfrom the desired station.

The results presented in Figs. 19 and 20 suggest thathome hi-fi receivers can provide listenable audio qualityat, but not much beyond, the protected contour of an FMstation broadcasting speech programming in the presenceof first adjacent channel interference. They also suggestthat the ability to receive listenable audio at the protectedcontour will likely be lost with the addition of the IBOCDAB signal to the first adjacent channel interferer.

The results presented in Figs. 21 and 22 suggest thatOEM automobile receivers can provide listenable audio


Fig. 14. Portable FM rock (lab). Fig. 16. Portable FM speech (lab).

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

Fig. 13. Portable FM rock (field). Fig. 15. Portable FM speech (field).



Fig. 19. Home hi-fi FM speech (field). Fig. 22. OEM auto AM rock (lab).

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

Fig. 18. Home hi-fi FM rock (lab). Fig. 21. OEM auto AM rock (field).

Fig. 17. Home hi-fi FM rock (field).Fig. 20. Home hi-fi FM speech (lab).

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

6 -4 -6 -8 -10

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

-14-14

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re


within, but not far beyond, the protected contour of an AMstation broadcasting rock programming, in the presence offirst adjacent channel interference. They also suggest thatthe addition of IBOC DAB to the first adjacent channelinterferer will have a noticeable impact on the perceivedaudio quality from the desired station, but that this impactwill not be sufficient to cause the typical listener to turnoff the station.

The results presented in Figs. 23 and 24 suggest thatOEM automobile receivers provide listenable audiowithin the FCC protected contour of an AM station broad-casting speech programming in the presence of first adja-cent channel interference, but that this audio will becomeunlistenable, or nearly unlistenable, with the introductionof IBOC DAB to the first adjacent channel station.

The results presented in Figs. 25 and 26 suggest thataftermarket automobile receivers provide listenable audiowithin the FCC protected contour of an AM station broad-casting rock programming in the presence of first adjacentchannel interference. They also suggest that the addition

of IBOC DAB to the first adjacent channel interferer willmake this listenable audio unlistenable in some situations.

The results presented in Figs. 27 and 28 suggest thataftermarket automobile receivers provide listenable audiowithin, but not much beyond, the FCC protected contourof an AM station broadcasting speech programming in thepresence of first adjacent channel interference. They alsosuggest that the addition of IBOC DAB to the first adja-cent channel interferer will make listenable audio unlis-tenable in some situations.

The results presented in Figs. 29 and 30 suggest thatportable receivers may provide limited instances of listen-able audio at D/U ratios of 16 dB or less within the FCCprotected contour of an AM station broadcasting rock pro-gramming in the presence of first adjacent channel inter-ference. They also suggest that the addition of IBOC DABto the first adjacent channel interferer will cause only lim-ited degradation to the existing audio quality in the mindsof typical listeners.

The results presented in Figs. 31 and 32 suggest that


Fig. 24. OEM auto AM speech (lab). Fig. 26. Aftermarket auto AM rock (lab).

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

Fig. 23. OEM auto AM speech (field). Fig. 25. Aftermarket auto AM rock (field).



1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

Fig. 27. Aftermarket auto AM speech (field). Fig. 30. Portable AM rock (lab).

Fig. 29. Portable AM rock (field). Fig. 32. Portable AM speech (lab).

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

Fig. 28. Aftermarket auto AM speech (lab). Fig. 31. Portable AM speech (field).

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re


portable receivers may provide limited instances of listen-able audio at D/U ratios of 16 dB or less within the FCCprotected contour of an AM station broadcasting speechprogramming in the presence of first adjacent channelinterference. They also suggest that the addition of IBOCDAB to the first adjacent channel interferer will, in manycases, cause only limited degradation to existing audioquality in the minds of typical listeners.

The results presented in Figs. 33 and 34 suggest thathome hi-fi receivers may provide limited instances of lis-tenable audio at D/U ratios of 16 dB or less within theFCC protected contour of an AM station broadcastingrock programming in the presence of first adjacent chan-nel interference. They also suggest that the addition ofIBOC DAB to the first adjacent channel interferer willcause only limited degradation to the existing audio qual-ity in the minds of typical listeners.

The results presented in Figs. 35 and 36 suggest thathome hi-fi receivers may provide limited instances of lis-tenable audio at D/U ratios of 16 dB or less within the

FCC protected contour of an AM station broadcastingspeech programming in the presence of first adjacentchannel interference. They also suggest that the additionof IBOC DAB to the first adjacent channel interferer will,in many cases, cause only limited degradation to the exist-ing audio quality in the minds of typical listeners.

5 OTHER TEST RESULTS

This engineering report focuses mainly on the first adja-cent channel compatibility results because first adjacentchannel compatibility is arguably the most difficult chal-lenge for the IBOC DAB systems. However, many otheraspects of IBOC DAB performance and compatibilitywere evaluated by the NRSC. The results of these evalua-tions are summarized here.

5.1 FM-Band ResultsFor the FM IBOC hybrid system, the NRSC determined

that the digital audio quality was “significantly improved”


Fig. 34. Home hi-fi AM rock (lab). Fig. 36. Home hi-fi AM speech (lab).

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 16 10 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re




Mea

n O

pin

ion

Sco

re

1

2

3

4

5

20 0

D/U Ratio (dB)

Mean

Op

inio

n S

co

re







Mea

n O

pin

ion

Sco

re

Fig. 33. Home hi-fi AM rock (field). Fig. 35. Home hi-fi AM speech (field).


compared to existing analog FM in mobile listening envi-ronments, and that this was true when the IBOC DAB sys-tem used either MPEG-2 AAC perceptual audio coding ora proprietary iBiquity codec based on Lucent Technol-ogies’ perceptual audio codec (PAC) [10]. It based thisconclusion largely on the fact that when all the subjectiveevaluation results from all the field test conditions wereaggregated, the listeners found the digital signal to be inthe “good to excellent” range, whereas the analog-onlysignal was typically in the “fair” range. The digital signalshowed a slightly greater improvement over the analogsignal for speech and classical music programming than itdid for rock and country programming.

The NRSC determined that the hybrid FM IBOC digi-tal coverage would be comparable to analog coverage, andthat the digital signal’s improved resistance to varioustypes of interference (co- and adjacent channel, impulsenoise, and multipath fading in particular) would make itavailable in areas where analog service currently cannotbe adequately received due to these types of interference.

The NRSC also concluded that FM IBOC receiversshould have better stereo separation than analog automo-tive receivers because analog automotive receivers typi-cally switch to monophonic reception to reduce the impactof interference on the audio, and this will not be necessarywith the digital signal because including stereo in the dig-ital bit stream does not reduce the S/N ratio of the audio inthe same manner that it does with the analog FM stereosystem.

Regarding compatibility with existing analog FM sig-nals, the NRSC found that listeners should not perceiveany impact on the host station’s analog signal with theaddition of the hybrid IBOC DAB signals. It also foundthat the addition of the digital signals would not increaseco-channel interference in the FM band because the digitalsignals do not occupy any of the spectrum used by a sta-tion’s analog FM signal, and therefore by definition do notoccupy any of the spectrum used by co-channel stations. Itfound that typical listeners within the FCC protected serv-ice area of an FM station should not perceive any impacton their analog audio quality when an IBOC DAB signal isadded to a first adjacent channel station. However, it didconclude that a limited number of listeners outside theFCC protected contour of an FM station may experiencesome degradation in analog audio quality in this situation.It found that listeners using analog automotive receiversshould not notice any impact when an IBOC DAB signal isadded to a second adjacent channel station, but that listen-ers using home hi-fi receivers might notice a negativeimpact if they are particularly close to (within a few kilo-meters of) the second adjacent station’s transmitter.

5.2 AM-Band ResultsThe NRSC found that the hybrid AM IBOC system

with MPEG-2 AAC perceptual audio coding demonstrated“significantly improved” audio quality over existing ana-log signals in the AM band. It based this conclusionlargely on the fact that when all the subjective evaluationresults from all the laboratory test conditions were aggre-gated, the typical listeners found the digital signal to be in

the “good” range, whereas the analog-only signal wasalways between “fair” and “poor.” The digital signalshowed a slightly greater improvement over the analogsignal for classical and rock programming than it did forspeech. At this time the NRSC has not yet completed testsof the AM-band system with the proprietary iBiquitycodec based on Lucent Technologies’ PAC.

The NRSC found that the coverage of the hybrid AMIBOC digital signal during the daytime would be compa-rable to the coverage of the analog signal. As with the FMsystem, it concluded that the digital signal’s improvedresistance to various types of interference (such as co- andadjacent channel, impulse noise, power lines) shouldmake it available in areas where analog service is cur-rently of unacceptable quality due to this interference.However, concerning nighttime coverage, the NRSCfound that the digital signal could not provide coverage asfar out as the analog signal due to interference to the dig-ital signal from adjacent channel analog signals.

Regarding stereo separation, the NRSC concluded thatAM IBOC digital receivers should provide superior stereoseparation compared to analog AM receivers, the vastmajority of which do not have any stereo capability at all.Digital stereo is only available from the AM IBOC DABsystem when the secondary and tertiary digital signals areavailable. If the only digital signal available is the primaryone, then the digital receiver will produce mono audio.

The NRSC concluded that the AM IBOC digital signalwill have little impact on the reception of the host analogsignal, and what impact there is will be receiver depend-ent. It found that the narrow-bandwidth automobilereceivers were the least sensitive to the digital signal,whereas the wider bandwidth hi-fi and portable radioswere found to be more sensitive to the digital signal. TheNRSC concluded that although the introduction of AMIBOC will be noticeable to some listeners of the host ana-log station, particularly those using hi-fi and portablereceivers, these listeners are not expected to find theiraudio quality sufficiently degraded to impact listening. Itshould be emphasized, however, that the NRSC did notconclude that the digital signals would have no impact onthe host analog signal, and some have suggested that theamount of interference to the host signal that the NRSC iswilling to accept is too high [11].

As with the FM system, the NRSC found that the addi-tion of the digital signals would not increase co-channelinterference in the AM band because the digital signals aredesigned to minimize interference to the host analog sig-nal, and therefore will not have an impact on other stationsthat operate on the same frequency as the host.

First adjacent compatibility was determined to be achallenge for the AM IBOC system, as was discussed inmore detail earlier. The NRSC found that analog automo-bile radios, which typically have narrower RF filters, areless susceptible to interference from a first adjacent IBOCdigital signal than home hi-fi and portable radios.

The NRSC found that second adjacent compatibilitywas not a significant problem for automobile receivers,but that typical listeners would notice some new interfer-ence on home hi-fi and portable analog receivers as a



result of IBOC DAB transmissions on a second adjacentchannel.

Interference to analog reception from stations transmit-ting IBOC DAB on third adjacent channels was deter-mined not to be a problem by the NRSC.

6 CONCLUSION

The IBOC DAB systems are compromise solutions forintroducing terrestrial digital audio broadcasting. In orderto have the benefit of a system that requires no new spec-trum allocations and causes minimal disruption to theexisting competitive balance between radio broadcasters,some new interference to analog reception must beaccepted. For the FM band this new interference is outsidethe FCC protected contour, although still within the rangeof mainly automobile receivers. For the AM band it is out-side the FCC protected contour and, to some extent, alsoinside the protected contour. This is to be expectedbecause AM-band channels are packed closer together onthe dial than FM-band channels, and they have a narrowerbandwidth.

The NRSC has concluded that IBOC DAB is a viablemethod for transitioning terrestrial radio broadcasts in theUnited States to digital broadcasting.

7 REFERENCES

[1] iBiquity Digital Corp., “FM IBOC DAB Laboratoryand Field Testing,” (FM IBOC Test Data Report), Reportto the National Radio Systems Committee (2001 Aug.).

[2] iBiquity Digital Corp., “AM IBOC DAB Laboratoryand Field Testing,” Third Report to the National RadioSystems Committee (2002 Jan. 4).

[3] National Radio Systems Committee DAB Subcom-mittee, “Evaluation of the iBiquity Digital CorporationIBOC System Part 1––FM IBOC” (2001 Nov. 29).

[4] National Radio Systems Committee DAB Subcom-mittee, “Evaluation of the iBiquity Digital CorporationIBOC System Part 2––AM IBOC” (2002 Apr. 6).

[5] FM IBOC Test Data Report, Appendix A.[6] AM IBOC Test Data Report, Appendix A.[7] FM IBOC Test Data Report, Appendix H.[8] Title 47 of the Code of Federal Regulations, Part 73,

Sections 73.37 and 73.215 (2001 Oct. 1).[9] FM IBOC Test Data Report, Appendix J.[10] H. D. Messer, NRSC DAB Evaluation Working

Group Chair, Memo to M. Smith, NRSC DAB Subcom-mittee, Chair (2002 May 16).

[11] R. C. Crane, President, C. Crane Company Inc.,Letter to NRSC (2002 Mar. 15).


THE AUTHOR

Dave Wilson has a B.S. degree in electrical engineering andan M.S. degree in accounting, both from the University ofVirginia.

Mr. Wilson is director of engineering at theConsumer Electronics Association (CEA) in Arlington,Virginia, where he oversees CEA’s efforts on variousstandards-setting activities affecting consumer electron-ics products. He joined CEA in 2000 after six years atthe National Association of Broadcasters (NAB), wherehe was a staff engineer and then manager of Technical

Regulatory Affairs. Prior to NAB he spent four yearsin the Office of Engineering and Technology at theFederal Communications Commission providing engi-neering support on equipment authorization and spec-trum allocation issues. Before that he spent six yearsat WUVA-FM in Charlottesville, Virginia, serving forseveral years as the station’s chief engineer beforebecoming its general manager and president. He cur-rently serves on the board of directors of WUVA,Incorporated.


1 AES46-2002

2003-04-14 printing

AES standard for network and file transfer ofaudio — Audio-file transfer and exchange —

Radio traffic audio delivery extensionto the Broadcast Wave file format

Published byAudio Engineering Society, Inc.Copyright ©2002 by the Audio Engineering Society

Abstract

This document provides a convention for communicating basic radio traffic and continuity data via adedicated chunk embedded in Broadcast-Wave-Format-compliant WAVE files.

An AES standard implies a consensus of those directly and materially affected by its scope and provisionsand is intended as a guide to aid the manufacturer, the consumer, and the general public. The existence ofan AES standard does not in any respect preclude anyone, whether or not he or she has approved thedocument, from manufacturing, marketing, purchasing, or using products, processes, or procedures not inagreement with the standard. Prior to approval, all parties were provided opportunities to comment or objectto any provision. Attention is drawn to the possibility that some of the elements of this AES standard orinformation document may be the subject of patent rights. AES shall not be held responsible for identifyingany or all such patents. Approval does not assume any liability to any patent owner, nor does it assume anyobligation whatever to parties adopting the standards document. This document is subject to periodic reviewand users are cautioned to obtain the latest edition. Recipients of this document are invited to submit, withtheir comments, notification of any relevant patent rights of which they are aware and to provide supportingdocumentation.


2 AES46-2002

2003-04-14 printing

Contents

Foreword..........................................................................................................................................................................................................30 Introduction................................................................................................................................................................................................40.1 Rationale .................................................................................................................................................................................................40.2 Conventions............................................................................................................................................................................................41 Scope............................................................................................................................................................................................................52 Normative references............................................................................................................................................................................53 Definitions and abbreviations............................................................................................................................................................54 Coding conventions................................................................................................................................................................................64.1 Coding examples.................................................................................................................................................................................64.2 Octet ordering........................................................................................................................................................................................75 Character set .............................................................................................................................................................................................76 cart extension chunk..........................................................................................................................................................................76.1 Chunk ordering......................................................................................................................................................................................76.2 Contents of cart extension chunk.............................................................................................................................................86.2.1 General..................................................................................................................................................................................................86.2.2 Default values and empty data................................................................................................................................................116.2.3 Short strings......................................................................................................................................................................................116.3 Other relevant information............................................................................................................................................................116.4 Broadcast Wave Format usage...................................................................................................................................................117 Private and application-specific information...........................................................................................................................118 Assignment of coding..........................................................................................................................................................................11Annex A Recommended parameter names...................................................................................................................................12A.1 Category names.................................................................................................................................................................................12A.2 Mark timer identification..............................................................................................................................................................12Annex B Informative references........................................................................................................................................................14Annex C Currently approved resource locator, identifier, and format references......................................................15


3 AES46-2002

2003-04-14 printing

Foreword

[This foreword is not a part of AES standard for network and file transfer of audio — Audio-file transfer andexchange — Radio traffic audio delivery extension to the Broadcast Wave file format, AES46-2002.]

This document was written by a task group, headed by D. Pierce and G. Steadman, of the SC-06-01 WorkingGroup on Audio-File Transfer and Exchange of the SC-06 Subcommittee on Network and File Transfer ofAudio, under project AES-X87 of the AES Standards Committee. The members of the task group were G.Novacek, D. Pierce, G. Steadman, G. Uzelac, and J. Zigler.

Mark Yonge, chairBrooks Harris, vice-chairSC-06-012002-02-21

NOTE: In AES standards documents, sentences containing the verb “shall” are requirementsfor compliance with the standard. Sentences containing the verb “should” are strongsuggestions (recommendations). Sentences giving permission use the verb “may.” Sentencesexpressing a possibility use the verb “can.”


4 AES46-2002

2003-04-14 printing

AES standard for network and file transfer ofaudio — Audio-file transfer and exchange —

Radio traffic audio delivery extensionto the Broadcast Wave file format

0 Introduction

0.1 Rationale

The radio broadcast industry utilizes a variety of production, on-air, and other equipment in daily operation.No single vendor dominates the industry. Users have long complained about the inability to transport audioand traffic-continuity data between systems in a uniform and easy fashion. This complaint is due to the lackof any uniform agreement about an exchange standard for communicating this information between systems.Often, different on-air delivery systems use proprietary audio-file formats and incompatible access methodsto manage audio storage and playback, yet the scheduling, continuity, or traffic information they use to labelaudio files share many common attributes. Furthermore, audio data themselves are represented in variousoften proprietary formats. To simplify the communication among different systems such as audio productionand on-air delivery systems, a common representation for both continuity or traffic information and audio datais desirable.

The resource interchange file format (RIFF) WAVE format has emerged as a dominant audio representation.It supports a wide variety of audio formats such as linear pulse-code modulation (PCM), Moving PicturesExperts Group (MPEG) formats, different sampling frequencies and sample sizes, multiple tracks, and so on.The RIFF conventions allow the arbitrary addition of other data without impacting the ability of diverse RIFF-compliant applications from reading and interpreting needed data. Thus, adding an extension to a WAVE fileallows the inclusion of needed continuity or traffic data to a widely accepted representation.

The RIFF specification requires all readers to be able to read all compliant RIFF files. When such anapplication encounters data that it is not prepared to handle, it can simply ignore the data and move on. SomeRIFF consumer applications are intolerant of new and unknown chunks. For this reason alone, theseapplications are not RIFF compliant; but they may be front-ended by so-called chunk-stripper utilities, theproduct of which is then RIFF compliant.

The radio traffic data (commonly called CART) format described in this document utilizes a widely usedaudio file format (WAVE and Broadcast Wave Format files). It incorporates broadcast-specific cartridge-labeling information into a specialized chunk within the file itself. As a result, the burden of linking multiplesystems is reduced to the producer applications writing a single file and the consumer applications readingit. The destination application can thereby extract information and insert it into the native databaseapplication as needed.

0.2 Conventions

0.2.1 Decimal pointsAccording to IEC directives, the comma is used in all text to indicate the decimal point. However, inspecified coding, including the examples shown, the full stop is used as in IEC programming languagestandards.

0.2.2 Data representationAll coding and data representations are printed in an equally spaced font.

0.2.3 Non-printing ASCII charactersNon-printing characters are delimited by angle brackets, as in <CR> for carriage return.

0.2.4 Reserved bitsUnless otherwise indicated, bit assignments shown as reserved are reserved for future standardization by theAES and only by means of amendment or revision of this document.


5 AES46-2002

2003-04-14 printing

1 Scope

This document provides a means for communicating basic radio traffic and continuity data via a dedicatedchunk embedded in broadcast-wave-compliant WAVE files. The new RIFF chunk supports most common dataused in radio traffic and continuity systems, whereas the WAVE format itself supports most samplingfrequencies, sample widths, and audio formats.

This document does not specify the representation of these or other data within a specific application’s space,only in the public interchange between disparate systems. Any such private representation may be coveredby other standards or by a particular vendor’s best judgment.

2 Normative references

The following standards contain provisions which, through reference in this text, constitute provisions of thisdocument. At the time of publication, the editions indicated were valid. All standards are subject to revision,and parties to agreements based on this document are encouraged to investigate the possibility of applyingthe most recent editions of the indicated standards.

ISO/IEC 646:1991, Information technology — ISO-7-bit coded character set for information exchange. Geneva,CH: International Organization for Standardization.

RIFF file structure. See the resource locator on the databases page of www.aes.org/standards.

3 Definitions and abbreviations

3.1RIFFresource interchange file format, a file representation upon which the WAVE file format is based

3.2chunka data package within RIFF files containing related data

3.3EBU Broadcast Wave FormatBWFWAVE files containing the EBU bext chunk as described in EBU Tech. Document 3285 (see 8)

3.4bextbroadcast wave or BWF extension chunk

3.5cartextension chunk to WAVE files containing CART format data as described in this document

3.6ASCII7-bit coded character set, according to ISO 646

3.7<NUL>character code specified as 0/0 in ISO 646 wherein all bits of the code are set to zero

3.8<CR>character code specified as 0/13 in ISO 646 for carriage return

3.9<LF>character code specified as 0/10 in ISO 646 for line feed


6 AES46-2002

2003-04-14 printing

3.10MPEGcompressed audio data as specified within ISO 11172-3:1993, originally formulated by the Motion PictureExperts Group as MPEG I audio

3.11WAVE fileAudio ‘waveform’ file format using the RIFF file representation (see 8)

4 Coding conventions

4.1 Coding examples

Coding examples and data layouts described herein use the syntax and conventions of the C-programminglanguage. These examples and layouts may be used for illustrative purposes only and neither constrain norrecommend a particular implementation style or method.

The mnemonics that describe the data types used in these examples shall be as shown in Table 1.

Table 1 — Example data typesAtomictype

Meaning C type

CHAR 8-bit signed integer, representing integer values from –128 to +127 signed character

BYTE 8-bit unsigned integer, representing integer values from 0 to +255 unsigned character

INT 16-bit signed integer in Intel format, representing integer valuesfrom –32768 to +32767

signed short integer

WORD 16-bit unsigned integer in Intel format, representing integer valuesfrom 0 to +65535

unsigned short integer

LONG 32-bit signed integer in Intel format, representing integer valuesfrom –2147483648 to +2147483647

signed long integer

DWORD 32-bit unsigned integer in Intel format, representing integer valuesfrom 0 to +42949672951

unsigned long integer

The tag type used in RIFF files may be further defined as

typedef DWORD FOURCC; // Four-character code

A macro operator, cvtFOURCC, can be defined whose purpose is to convert a character string or fourcharacters into a FOURCC representation in a system-independent, portable fashion, for example:

tag4cc = cvtFOURCC("ABCD");tag4cc = cvtFOURCC('A', 'B', 'C', 'D');

When the string // appears in a structure layout or other illustrative example, it can designate the start ofan explanatory comment. Neither // nor any text following are part of the actual layout or data.


7 AES46-2002

2003-04-14 printing

4.2 Octet ordering

The octet ordering used for the storage of multi-octet numeric data (INT, WORD, DWORD, LONG, and so on)in RIFF files shall be the least significant octet first.

16-bit values (with bits numbered 00 though 15) shall be stored in files as in Table 2.

Table 2 — 16-bit valuesOctet 1 1

51

41

31

21

11

00

90

8

Octet 0 07

06

05

04

03

02

01

00

32-bit values (with bits numbered 00 through 31 from least significant to most significant) shall be stored infiles as in table 3.

Table 3 — 32-bit valuesOctet 3 3

13

02

92

82

72

62

52

4

Octet 2 23

22

21

20

19

18

17

16

Octet 1 15

14

13

12

11

10

09

08

Octet 0 07

06

05

04

03

02

01

00

5 Character set

The cart extension chunk shall use the ASCII character set for all text strings.

The first character of the identifier FOURCC shall be an upper- or lowercase alphabetic character, followedby one to three upper- or lowercase alphabetic or numeric characters. If the identifier is less than fourcharacters long, the remaining characters shall be <NUL> characters.

Uppercase FOURCC reserved identifiers, when used for chunk identifiers, are by convention reserved forspecific registered RIFF identifiers (see annex C). Other chunk identifiers shall use lowercase alphabetic andnumeric characters.

Internal data within chunks using FOURCC identifiers may use upper- or lowercase alphabetical and numericcharacters

6 cart extension chunk

6.1 Chunk ordering

The first chunk in a CART WAVE file shall be the format (fmt ) chunk.

If required, the fact chunk should be second. The last chunk should be the data chunk. Other chunks inthe file may be in any order.

NOTE: This chunk order is expected to provide optimum speed of access to the radio traffic data witha wide range of computer filing systems. Any RIFF-compliant chunk sequence may be encountered inpractical interchange.

An example of an MPEG-encoded WAVE file with both bext and cart chunks, in addition to the factchunk and mpeg chunk, is shown.

EXAMPLE<WAVE-form>RIFF 'WAVE'<fmt-ck> // required for all WAVE files<fact-ck> // required for non-PCM data<bext-ck> // EBU BWF chunk<mpeg-ck> // EBU MPEG-extension data chunk<cart-ck> // cart information<data-ck> // audio data, required for all WAVE files


8 AES46-2002

2003-04-14 printing

6.2 Contents of cart extension chunk

6.2.1 General

The cart extension chunk shall have the contents shown in table 4. The example uses C-programmingnotation for illustration.

EXAMPLE

typedef struct cartchunk_tag DWORD ckID; // FOURCC chunk ID: cart DWORD ckSize; // chunk data length in octets BYTE ckData[ckSize]; // data, as cart_EXTENSION typetypedef struct cart_extension_tag CHAR Version[4]; // Version of the data structure CHAR Title[64]; // ASCII title of cart audio sequence CHAR Artist[64]; // ASCII artist or creator name CHAR CutNum[64]; // ASCII cut number identification CHAR ClientID[64]; // ASCII client identification CHAR Category[64]; // ASCII category ID, PSA, NEWS, etc CHAR Classification[64]; // ASCII classification or auxiliary key CHAR OutCue[64]; // ASCII out cue text CHAR StartDate[10]; // ASCII YYYY-MM-DD CHAR StartTime[8]; // ASCII hh:mm:ss CHAR EndDate[10]; // ASCII YYYY-MM-DD CHAR EndTime[8]; // ASCII hh:mm:ss CHAR ProducerAppID[64]; // Name of vendor or application CHAR ProducerAppVersion[64]; // Version of producer application CHAR UserDef[64]; // User defined text DWORD dwLevelReference // Sample value for 0 dB reference CART_TIMER PostTimer[8]; // 8 time markers after head CHAR Reserved[276]; // Reserved for future expansion CHAR URL[1024]; // Uniform resource locator CHAR TagText[]; // Free form text for scripts or tags CART_EXTENSION;

typedef struct cart_timer_tag // Post timer storage unit FOURCC dwUsage; // FOURCC timer usage ID DWORD dwValue; // timer value in samples from head CART_TIMER;

Table 4 — CART extension chunk contents

Field DescriptionVersion A 4-character ASCII numeric string giving the version of the cart data

structure, particularly the contents and usage of the reserved area. Thefirst two numbers shall give the major release level (with leading 0) from00 to 99 and the last two shall give the revision level (with leading 0) inthe range of 00 to 99. The version number of the cart data structure asdescribed in this document shall be version 1.01, and thus is representedby the string 0101.

Title An ASCII string, 64 characters or less, representing the title of the cut.The title should be a descriptive summary of the audio contents of thefile, and may be used as an entry into a table of contents, and so on.


9 AES46-2002

2003-04-14 printing

Applications that do not support a 64-character title may truncate thefield as needed.

Artist An ASCII string, 64 characters or less, holding the artist or creator namefor the audio cut.

CutNum An ASCII string, 64 characters or less, representing the cut number, orunique cut key. The string shall be left justified. Some consumer systemscan have restricted cut number lengths or allowable character set. Theseapplications should provide some means of synthesizing a usable cutidentifier if it has such restrictions.

ClientID An ASCII string, 64 characters or less, holding a client or customeridentification or name.

Category An ASCII string, 64 characters or less, holding a category name. Thecategory name may be application dependent. Applications should usecommon category names. See annex A for a list of recommendedcategory names.

Classification An ASCII string, 64 characters or less, holding a classification key. Thiskey can be used for general classification, selection, or sorting based onlanguage, locale, or other similar applications.

OutCue An ASCII string, 64 characters or less, holding the optional out cuephrase to be displayed when the cut is being played. This shall be a userreadable cue string.

StartDate An ASCII date string, 10 characters, of the form YYYY-MM-DD, such as1998-12-25, holding the start date.

Year (YYYY) shall be defined as 0000 to 9999.Month (MM) shall be defined as 01 to 12.Day (DD) shall be defined as 01 to 28, 29, 30, or 31 as applicable.The separator between date fields shall be a hyphen (-).

NOTE: This format complies with ISO 8601 and is compatible with otherdates in BWF files.

To signify an immediate start date, applications shall use 1900-01-01.StartTime An ASCII time string, 8 characters, of the form hh:mm:ss, such as 12:31:45,

representing the 24-hour time-of-day for the start time on the assigned startdate.

Hour (hh) shall be defined as 00 to 23.Minutes (mm) and seconds (ss) shall be defined as 00 to 59.

If blank, applications shall assume a start time of 00:00:00..The separator between time fields shall be a colon (:)


10 AES46-2002

2003-04-14 printing

Table 4 (continued)EndDate As in start date, but shall indicate the date after which the sequence

will no longer be active. If the sequence is to run forever, the date shallbe 9999-12-31. There shall be no default for this field.

EndTime This code shall indicate the time of day on the appointed end date afterwhich the sequence becomes inactive. If blank, applications shallassume an end time of 23:59:59.

ProducerAppID An ASCII string, 64 characters or less, containing the vendor name,product name, or both, of the program or application that produced theWAVE file with this cart chunk.

ProducerAppVersion An ASCII string, 64 characters or less, containing the version of theprogram or applications that produced the WAVE file containing thecart chunk. Because this string is informational only, the applicationmay represent the version in any convenient format.

UserDef An ASCII string, 64 characters or less, whose use and contents may bedefined by the user of the system.

dwLevelReference A 32-bit signed (2’s complement) integer word that shall hold the samplevalue of the 0-dB reference level for the originating system. This referencecan facilitate scaling and metering consistency across disparate systems.As an example, a 16-bit linear PCM system that has its meters calibratedas 0 corresponding to maximum signed digital value shall have the valueset to 32768 (800016).

The peak value shall be the absolute value of the largest sample valuepossible before saturation. In the example given, that of a 16-bit linearsystem using 2’s complement notation, the range of allowable values is -32768 to 32767. Thus the maximum peak value is 32768 in the examplegiven.

PostTimer Eight CART_TIMER structures representing time marks. The time unitsshall be in sample periods at the sampling frequency of the associatedaudio data and shall be referenced to the first sample of the audio data.

The timer range shall be 232 or 4,294,967,295 sample periods. These periodsallow timer ranges at a sampling frequency of 48 kHz, for example, toextend beyond 24 h (24:51:18).

These timers may be used to activate events in the cart system.

Each timer entry shall consist of a FOURCC timer usage identifier(dwUsage) and a 32-bit unsigned integer DWORD timer in sample periods(dwValue) as described above. Applications should use FOURCC usageidentifiers as described in Annex A.3

If a timer is not used, or is not set, its usage identifier should be set to all<NUL> characters (0000000016) and its timer value set to 0(0000000016).

Reserved This area, 276 octets, shall be reserved.URL An ASCII string, 1024 characters or less, representing a universal

resource locator (URL) referencing or referenced by the audio program.The URL field contents should conform to the URL syntax as shown inAnnex C.

TagText Non-restricted ASCII characters containing a collection of strings, eachterminated by <CR><LF>. This text may be system- or user-defineddescriptive text for the sound, such as live tag, script information,descriptive text special instructions, and so on.


11 AES46-2002

2003-04-14 printing

6.2.2 Default values and empty data

Text fields that do not require strings shall have a zero-length string: the first octet of the field shall be a<NUL> character, and all subsequent data in the field shall be ignored.

Binary value fields not requiring a specific value will be set to a value of 0.

The action to be taken on encountering a blank, or empty data, may be implementation and site-installationdependent.

6.2.3 Short strings

In cases where the string content, in octets, is shorter than the specified field size, the string shall beterminated by a <NUL> octet following the last significant character. This terminator and the remainder ofthe field shall be ignored. Strings shall be left justified.

6.3 Other relevant information

All the other information regarding WAVE audio characteristics can be found in the mandatory fmt chunk.This includes sampling frequency, number of tracks, sample width, and sample format. For other than PCMformat, the fact chunk and the EBU mpeg chunk can contain further information (see EBU Tech.Document 3285 for information on these chunks (see annex B).

6.4 Broadcast Wave Format usage

The broadcast wave bext chunk may be used in conjunction with the cart data described in this document.Its data does not conflict with nor replicate data here.

Refer to EBU Tech. Document 3285 for more specifics on the recommended official usage of these fields.

7 Private and application-specific information

Private and application-specific data not contained in the cart chunk data described here is outside thescope of this document.

8 Assignment of coding

The coding of resource locators, identifiers, and formats shall conform to recognized industry practice asdetermined by the AES Standards Committee (AESSC) according to its rules. This determination is shownas of the printing date of this standard in Annex C, which shall be published and kept current on the AESSCWeb site databases page.


12 AES46-2002

2003-04-14 printing

Annex A

(informative)

Recommended parameter names

A.1 Category names

Categories and aliases should be according to Table A.1. The actual categories and aliases may be sitedependent, implementation dependent, or both.

Table A.1 – Recommended category names and aliases

Category names AliasesAll ALLBeds BED, BEDSSound bits BIT, BITSCommercials COM, COMMContests CON, CONTDaily play lists DAYEmergency broadcast EBSound effects EFXFillers FIL, FILLStation ID IDIntros INT, INTRJingles JIN, JINGLiners LIN, LINELogos LOG, LOGOMagic call MAG, MAGIMusic MUS, MUSCNetwork delay NET, NETWNews NEW, NEWSPromos PRO, PROMPublic service announcements PSASegues SEGShows SHW, SHOWSound effects SNDSpots SPO, SPOTSports SPR, SPRTStagers STG, STAGAnnouncer stack STK, STAKSweeps SWP, SWEPTest tones TST, TESTTemporary TMP, TEMP

A.2 Mark timer identification

Timer types, along with their FOURCC identification, should be according to Table A.2. The interpretationand behavior of systems on encountering timer information may be site dependent, implementationdependent, or both.


13 AES46-2002

2003-04-14 printing

Table A.2 – Basic timer types

Timer ID Description Start–End Enumerated MultiplesUnused No No Yes

SEG Segue timer Yes Yes YesAUD Audio boundary Yes No NoINT Introduction Yes Yes YesOUT Epilog Yes Yes YesSEC Secondary Yes Yes YesTER Tertiary Yes Yes YesMRK Generic marker No Yes YesEOD End-of-data No No Yes

Timers may be qualified in one of three ways:

a) as start or end timers, by appending a lowercase ASCII letter s for a start timer or a lowercaseASCII letter e for an end timer; for example, the timer identification AUDs designates the start ofaudio following silence, whereas AUDe designates the end of the audio segment;

b) as enumerated timers, by appending an ASCII numeric character; for example, SEC1 can bedesignated secondary timer number 1, SEC2 can be secondary timer number 2, and so on;

c) as multiple timers, by having multiple instances of the same timer ID; one could have, forexample, multiple instances of MRK.

Each application may prioritize the order of the timers.


14 AES46-2002

2003-04-14 printing

Annex B

(informative)

Informative references

Microsoft Software Developers Kit Multimedia Standards Update, rev 3.0 15 April 1994, MicrosoftCorporation.

Microsoft Multimedia Programmer’s Reference 1991–1992, Microsoft Corporation.

EBU Tech. Document 3285 — Specification of the Broadcast Wave Format, Supplement 1 — MPEG audio.Geneva, CH: European Broadcasting Union.

ISO 11172-3 — Part 3: Audio. Information technology —- Coding of moving pictures and associated audio fordigital storage — Geneva, CH: International Standards Organization.


15 AES46-2002

2003-04-14 printing

Annex C

(normative)

Currently approved resource locator, identifier, and format references

See the resource locator on the databases page of www.aes.org/standards for updated information.

IETF RFC1738 — Uniform Resource Locators (URL). Internet Engineering Task Force.

EBU Tech. Document 3285 — Specification of the Broadcast Wave Format. Geneva, CH: EuropeanBroadcasting Union.

Julian DunnI was greatly saddened to learn that Julian had failed inhis battle against leukemia, in his inimitably quiet but de-termined way, he had been optimistic and courageous inhis fight.

I first met Julian when I worked at the BBC and he wasa young engineer in Designs Department. However, itwas only after Julian left the BBC and he started at-tending the AES standards meetings that I really gotknow him. In spite of his quiet manner I soon realizedJulian was an excellent engineer and his input to a dis-cussion always warranted careful consideration. At timeshe could appear to be stubborn and determined but thiswas because he felt a principle was at stake and notbecause he wanted to enhance his personal position; hewas usually right. Julian rightly earned an internationalreputation in the field of digital audio from many of theworld’s experts.

Julian's contribution to the AES standards work hasbeen immense, particularly the AES3 digital audio in-terface standard, and it is fitting that his work on thisstandard will provide a lasting legacy to his memory.Outside of the formal meetings Julian was always willingto help. I remember an occasion at a convention in Pariswhen, because of problems with the photocopying ma-chines, he and Chris Travis helped Dan Queen andmyself photocopy many hundreds of documents so theywould be ready for meetings the next day. We did notfinish until 10:00 pm but that did not concern Julian; hejust wanted to ensure the meetings would run smoothly.

I and many other members of the AES StandardsCommittee will greatly miss Julian, not just for histechnical expertise but because he was also a witty, prin-cipled and caring colleague.

John NunnChairman, AES Standards Committee

Call for Comment on WITHDRAWAL ofAES19-1992 Published 2003-03-10Call for Comment on WITHDRAWAL of AES19-1992(r1997), AES-ALMA standard test method for audio engi-neering—Measurement of the lowest resonance frequencyof loudspeaker cones has been published.

Withdrawal of a document results in AESSC no longersupporting it with periodic reviews or updates of its asso-ciated database. The document will remain in the catalogand on the Web site but designated as withdrawn. By ar-rangement with ALMA International this document willcontinue as an ALMA standard.

This document was developed by a writing group of theAudio Engineering Society Standards Committee(AESSC) and has been prepared for comment according toAES policies and procedures. It has been brought to the at-tention of International Electrotechnical CommissionTechnical Committee 100. Existing international standardsrelating to the subject of this document were used and ref-erenced throughout its development.

Address comments by e-mail to the secretariat [email protected] or by mail to the AESSC Secretariat,Audio Engineering Society, 60 E. 42nd St., New York,NY 10165. E-mail is preferred. Only comments so ad-dressed will be considered. Comments that suggestchanges must include proposed wording. Comments mustbe restricted to this document only. Send comments toother documents separately.

This document will be approved by the AES after anyadverse comment received within three months of the pub-lication of this call on www.aes.org/standards 2003-03-10,has been resolved. All comments will be published on theWeb site. Persons unable to obtain this document from theWeb site may request a copy from the secretariat at: AudioEngineering Society Standards Committee, DraftComments Dept., Woodlands, Goodwood Rise, Marlow,Bucks SL7 3QE, UK.


COMMITTEE NEWSAES STANDARDS

Information regarding Standards Committee activi-ties including meetings, structure, procedures, re-ports, and membership may be obtained viahttp://www.aes.org/standards/. For its publisheddocuments and reports, including this column, theAESSC is guided by International ElectrotechnicalCommission (IEC) style as described in the ISO-IECDirectives, Part 3. IEC style differs in some respectsfrom the style of the AES as used elsewhere in thisJournal. For current project schedules, see the pro-ject-status document on the Web site. AESSC docu-ment stages referenced are proposed task-groupdraft (PTD), proposed working-group draft (PWD),proposed call for comment (PCFC), and call forcomment (CFC).


AAEESS 111144thth

CCOONNVVEENNTTIIOONN2003 March 22–25

RAI Conference and Exhibition CentreAmsterdam, The Netherlands


hesitated to come to the convention at this time due to eco-nomic worries and concerns about the political situation inthe Middle East; but he pointed out that there were evenbetter reasons to come. Peter Swarte was pleased to an-nounce that 250 exhibitors were participating and delegatesfrom over 70 countries had preregistered for the event.

Because the chair of the Awards Committee, Roy Pritts,was grounded in Denver, Colorado, by over four feet (about1.2 meters) of snow, Past President Garry Margolis waspleased to announce the AES awards. The PublicationsAward, announced by Journal Editor Daniel von Reckling-hausen, was given this time to a large group of authors whohad jointly contributed to an influential paper entitled “TheITU Standard for Objective Measurement of Perceived Au-dio Quality,” published in the Journal of the Audio Engi-neering Society, volume 48, number 1/2. Thilo Thiede,William Treurniet, Roland Bitto, Christian Schmidmer,Thomas Sporer, John Beerends, Catherine Colomes,

Bathed in glorious spring sunshine, the AES114th Convention brought thousands to Ams-terdam to meet, learn, and share in the broad

community that is the audio industry. A more extensiveand rich technical program than ever before includednewly established educational seminars on basic audio en-

gineering issues as well as exhibitor seminars that enabled del-egates to discover more about the latest product developments.Peter Swarte, convention chair, worked tirelessly with his con-vention committee over the year leading up to the event to en-sure that there was something for everyone at Amsterdam’sRAI Centre.

OPENING CEREMONY AND AWARDSPRESENTATIONOfficially opening the convention on the evening of thefirst day, AES President Kees Immink, himself a nativeNetherlander, acknowledged that some people might have


114th Convention

Michael Keyhl, Gerhard Stoll, Karlheinz Brandenburg andBernhard Feiten formed an impressive group for their pho-tograph on the stage, exemplifying the impressive fruition ofinternational cooperation and research.

The Board of Governors Award, given for outstandingcontributions to the Audio Engineering Society, was pre-sented to Nickolay Ivanov for chairing the AES 21st Inter-national Conference in St. Petersburg, Russia; Martin Wöhrfor chairing the 112th Convention in Munich; and Daniel

Zalay for chairing the 108th Convention in Paris.Fellowships were conferred upon John Beerends for suc-

cessful application of principles of perception and cognitionto objective quality measurement of audio-visual signalsand Rinus Boone for his work in the fields of outdoor noisepropagation, audio transducers, multichannel sound repro-duction, simulation, and auralization. The large crowd ac-knowledged the important contributions of the award win-ners with warm applause.

Kees Immink, AES president

Roger Furness, AES executive director

Kees Immink, center, surrounded by Publications Award winners, from left, Roland Bitto, ChristianSchmidmer, Karlheinz Brandenburg, Thomas Sporer, Michael Keyhl, John Beerends, Gerhard Stoll, andBernhard Feiten. Thilo Theiede, William Treurniet, and Catherine Colomes could not attend.

Daniel vonRecklinghausen,editor

Garry Margolis,filling in for RoyPritts, awards chair

Stuart Bruce, keynote speaker

Peter Swarte,convention chair


114th Convention

The keynote address at the 114th Convention was given byStuart Bruce, the successful British recording engineer andproducer who now runs an independent business as part ofReal World Studios in Bath, England. Bruce opened in com-bative style, asking the blunt question, “Why is the main-stream music business dead on its feet?” Customers are leav-ing us in droves, he suggested, and the endless diet ofmediocre pop music exemplified by the steady supply of“talent search” series on television are only feeding the phe-

nomenon. “We are living in an era of the blind leading thebland,” he contended, with many teenagers going back to themusic of the 1960s and 1970s for inspiration. Looking atwhat is needed if the music business is to regain its vitalityand credibility, Bruce proposed that producers and engineerscan develop talent—that is how the innovative record labelsbegan in the past. Good music always rises to the surface,even if promoted by relatively small labels, and a number ofsuch enterprises have really learned how to use the Inter-

Board of Governors Award recipients,clockwise from above, Nickolay Ivanov,

Martin Wöhr, and Daniel Zalay

Fellowship Awards recipients John Beerends (left) and Rinus Boone


114th Convention

net to their advantage in recent years. “Why not start smalland build things up again?” He pleaded for a better dialoguebetween the studios and educators, mediated by the AES,otherwise we are in danger of losing valued old techniques.

EXHIBITIONSupported for the first time by a popular series of 25 exhibitorseminars that enabled manufacturers to explain their recentdevelopments in greater depth to participants, the exhibitionfloor buzzed with interest in response to the range of productsand services on display. Some of the highlights are reviewedhere.

Top of the bill was SSL’s launch of two new digital mixingconsoles, the C100 and the C200, adhering to the company’sone-knob-per-function heritage designed to appeal to a widerange of engineers. The C200 is intended for mixing applica-tions, such as music and entertainment, with a special mobileversion available. Among its many features is the novel self-healing DSP that is designed to work around hardware fail-ures; it also has integrated workstation control, full automa-tion, and scalable DSP. Unlike some previous digital mixers,this one will work at both 48- and 96-kHz sampling frequen-

cies, and its 192-kHz-capable Centuri platformis optimized for DVD-Audio applications.

Innova SON showed its new Sy80 digitalmixer that is targeted to live applications suchas front-of-house and monitoring, as well aslive recording and on-air mixing. Also at theforefront of digital console developments wasCalrec’s new Zeta 100, intended as a live pro-duction console for the broadcast market, tar-geted at local TV and radio operations. Thismixer is claimed to enable broadcasters to in-vest in digital technology at a price similar tothat of comparable analog technology. Indeed,in the analog domain there are still new con-soles being developed, and Yamaha showed itsnew PM5000 mixer that is primarily intendedfor front-of-house live applications, with alarge number of independent scene memories

for automation purposes.Multichannel transmission has

finally come to the Europeanbroadcast ing industry in theform of broadcasts involving 5.1surround from ORF (AustrianBroadcasting Corporation) andSwedish Radio. ORF’s NewYear’s Day broadcast from Vi-enna was highlighted on theDolby stand, showing evidenceof Dolby Digital broadcasting bythe Austrian network. Further-more, Swedish Radio made abroadcast greeting to the AESconvention in 5.1 surround overa Sirius 2 satellite transmissionusing DTS encoding at a veryhigh rate of 1.5 Mbit/s, calling

Large group of poster presentations facilitated direct contact withauthors and laptop-computer demonstration of new concepts.

Ron Streicher (left), AES president-elect, and Stanley Lipshitzgiving tutorial seminars.

Attentive crowd listens to author Rozenn Nicol.

Among exhibitor seminars winning rave reviews were DVD-Audio: Exploring the Format(above) and Super Audio CD Procuction (inset).

Mono

Multichannel

Stereo

• Home Theater/Entertainment

• Wireless + Portable

• Telecom + Voice

• Gaming

• Internet + Broadcast

Technologies. Product Applications

World Wide Partners

• Circle Surround II

• FOCUS

• SRS 3D

• SRS Headphone

• TruBass

• TruSurround XT

• VIP

• WOW

The Future of Audio. Technical information and online demos at www.srslabs.com2002 SRS Labs, Inc. All rights reserved. The SRS logo is a registered trademark of SRS Labs, Inc.C

Aiwa, AKM, Analog Devices, Broadcom, Cirrus Logic, ESS, Fujitsu, Funai,

Hitachi, Hughes Network Systems, Kenwood, Marantz, Microsoft,

Mitsubishi, Motorola, NJRC, Olympus, Philips, Pioneer, RCA, Samsung,

Sanyo, Sherwood, Sony, STMicroelectronics, Texas Instruments, Toshiba

SRS Labs is a recognized leader in developing audio solutions for any application. Its diverse portfolio

of proprietary technologies includes mono and stereo enhancement, voice processing, multichannel

audio, headphones, and speaker design. • With over seventy patents, established platform partnerships

with analog and digital implementations, and hardware or software solutions, SRS Labs is the perfect

partner for companies reliant upon audio performance.


114th Convention

it an experiment in “high-definition radio.”Birthdays were in order at the 114th. Loudspeaker manu-

facturer Genelec celebrated its 25th anniversary at the con-vention by giving away a monitoring system every day. Thesystems got bigger each time and culminated in a choice ofthe 1029.LSE PowerPak or 6040A systems. In addition,Neumann launched its 75th-birthday Sound EngineeringContest with prizes including U87 and TLM103 micro-phones. Linn Products introduced its 328A professionalmonitor that has remarkable designer styling for near- and

mid-field applications.DSD and SuperAudio CD development is gathering

speed, including noise removal tools from PureNotes for theMerging Technologies DSD platform and the new Sony Su-perMac DSD interface that enables either standard PCM orDSD signals at a variety of different sampling frequencies tobe transmitted over CAT5 cabling. Philips described theDSD-IFF file format that enables DSD material to be inter-changed between systems. DVD-Audio also had a greaterpresence at this convention, with a demo room containing

Ronald Aarts (right), papers chair, and author JohnVanderkooy

Thierry Bergmans, exhibitsorganization

Jan Romijn (right), banquet chair

Werner de Bruijn,education events chair

Stan Tempelaars (left)and Alex Balster,

historical events chairs

From left, Menno van der Veen, technical tours chair; PeterSwarte, convention chair; John Beerends, workshops vice chair


114th Convention

Meridian top-of-the-range replay systems to demonstrateDVD surround reproduction from companies such as AIXRecords. This small independent label produces DVD-Au-dio disks containing mixes from different “standpoints,”such as one from within the ensemble and another from anaudience position, demonstrating the versatility of the for-mat. SADiE also showed its new Series 5 range of worksta-tions that incorporates DSD mastering and authoring as analternative to conventional PCM.

If recording up to 100 kHz is your aim, for SuperAudio

CD or DVD-Audio, then Sanken’s CO-100K omni micro-phone, shown for the first time here, may be suitable. Thecompany claims that this is the first microphone in theworld to offer 100-kHz frequency response outside of themeasurement microphone domain. It was designed togetherwith NHK’s research laboratory.

Networking and interchange were high on the agenda in anumber of places. Digigram’s new EtherSound distributiontechnology can handle 64 channels of audio on CAT5 cablingand take advantage of standard Ethernet switches for rout-

Michiel van Eeden (right), facilities chair, and RonStreicher, AES president-elect

Han Tendeloo,conventionprogramcoordinator and AES secretary

From left, Peter Swarte, convention chair; StefaniRenner, press relations, Roger Furness, executivedirector

Gisèle Clark, promotion

Erik Larsen (right), papers vice chair, with author MattiKarjalainen

Diemer de Vries,workshops chair

We Thank…Peter Swarte

chair

Ronald Aartspapers chair

Erik Larsenpapers vice chair

Diemer de Vriesworkshops chair

John Beerendsworkshops vice chair

Werner de Bruijneducation events chair

Stan TempelaarsAlex Balster

historical events chairs

Jan Romijnbanquet chair

Michiel van Eedenfacilities chair

Menno van der Veentechnical tours chair

Han Tendelooprogram coordinator


114th Convention

ing purposes. Lawo explained its mc2Net series of productsthat complements its digital mixers and enables them to be in-tegrated into networked studio complexes, while Otari adver-tised increased speeds for its ND-20 fibre optic network sys-tem based on IEEE 1394 mLAN technology. A range ofISDN and broadcast contribution encoding systems were alsoon display from a variety of manufacturers.

Further supporting the migration from dedicated audioformats to generic mass storage media for audio, Fostex’sPD-6 location recorder was introduced to enable multichan-nel audio recording along with timecode on a portableDVD-RAM recorder. This innovative product utilizes aform of mini-DVD-RAM that is only 8 cm in diameter, andenables audio from location recordings to be uploaded di-rectly to editing systems. Otari gave visitors a preview of itsDR-100 multichannel recorder that is backward compatiblewith RADAR-II material, and its DR-10 MO recorder de-signed for broadcasting applications. Nagra showed a newtwo-channel machine, the Nagra V, that uses high-resolu-tion disk recording. It is designed for location recording infilm and video applications, the latest in Nagra’s long tradi-tion of robust outdoor machines.

HEYSER LECTURE AND RECEPTIONThe Richard C. Heyser Memorial Lecture, given at each con-vention in honor of its namesake, organized by the TechnicalCouncil, was given this year by Professor Jens Blauert.Blauert is well known around the world for his research on

human perception, and spatial perception inparticular. His book Spatial Hearing hasbeen translated into many languages. Hislecture, “Communication Acoustics: AudioGoes Cognitive,” dealt with the new chal-lenges that face researchers in buildingmodels that take account of the higher levelperceptual organization of auditory infor-mation. He related these themes to thecomplementary fields of CASA (computa-tional auditory scene analysis) and VR (vir-tual reality).

Blauert’s lecture was packed to theseams, with over 250 people, including anumber of his former students, listening at-tentively. He concluded his lecture with alist of the 50 Ph.D. students who have con-tributed since 1980 to the work discussed.One of his students had helped to generatesome remarkable footage of shots from astorage oscilloscope that indicated sourcebroadening effects, which Blauert played

as a movie showing dynamic changes.The Technical Council reception that followed this event

enabled delegates to meet each other in an informal atmo-sphere over a glass of wine.

SEMINARS The AES took a bold step to further its educational mission atthe 114th by instigating a series of 10 seminars led by key fig-ures in the industry. The aim of these was to offer visitors anopportunity to obtain a good basic grounding in many of au-dio’s key issues. The seminars proved extremely popular withstudents and professional delegates alike.

An enthusiastic audience turned out to hear two of the in-dustry’s best-known academics, Stanley Lipshitz and JohnVanderkooy. Their seminar, Basics of Digital Audio, includ-ed numerous demonstrations using custom-made equip-ment. Two seminars were on surround: Stereo and SurroundMicrophone Techniques, chaired by Geoff Martin, and Howto Set Up 5.1 Surround, presented by Andrew Goldberg.Martin also ran another seminar, Microphone TechniqueTheory for Stereo and Surround, which was complementedby Ron Streicher’s Working with Microphones: A PracticalReview. Diemer de Vries’ excellent primer, Basics of RoomAcoustics, was also well received. Indeed there was some-thing for almost everyone—including quite a few seasonedpros who probably felt like sitting in on a refresher course.Other seminars included Hearing Damage in Musical Prac-tice, Grounding and Shielding, Mixing and Mastering, andBasics of Sound Reinforcement, given by Peter Swarte,114th convention chair.

WORKSHOPSIn addition to coorinating the new seminar series, Diemer deVries and John Beerends put together the well-establishedworkshop program that contained 13 separate events of vari-ous lengths covering diverse topics from recording technolo-gy to measurement techniques. Of particular interest at

Jens Blauert, 114th Richard C. Heyser Distinguished Lecturer,during lecture; receiving certificate from Technical Council officers,from left, Robert Schulein, Wieslaw Woszczyk, and Jürgen Herre;and answering questions at reception after lecture.


114th Convention

the present time was The Future of High Resolution Audio,chaired by Malcolm Hawksford, at which panelists debatedthe various formats and technologies that offer higher resolu-tion than that offered by the CD. Multichannel SurroundSound: A Chance for Enhanced Creativity, chaired by pro-ducer Martha de Francisco, brought together a number of toprecording engineers and producers from broadcasting and thepop and classical fields to discuss the ways in which sur-round sound may increase the creative options open to users.James Mallinson, for example, claimed that high-resolutionformats for the first time brought back the sense of emotionconveyed by music reproduction that he remembered fromanalog days, while others claimed that this was of relativelyminor importance compared with the enhancement providedby surround sound.

Other interesting workshops included Low Bit-Rate Codingof Spatial Audio, which looked at the different ways in whichinterchannel similarities and differences can be exploited indata reduction. It included a fascinating historical tour of spa-tial audio representation since the dawn of stereo, given byMark Davis of Dolby Labs. Additional workshops were onsampling rate convertors, MPEG-4, LAN delivery of audioand application to digital libraries, handset and headset test-ing, electronic reverberation for concert halls, correlation be-tween subjective and objective measurements for automotivesound systems, AES47, large-room acoustics, the value of in-formation, and wavefield synthesis applications.

PAPERS SESSIONSDuring the convention the papers sessions were often filledto capacity or beyond, in an extremely wide-ranging and in-teresting program coordinated by Ronald Aarts and EricLarsen. A strong emphasis was on psychoacoustics, percep-tion and listening tests, as well as spatial perception and pro-cessing, and multichannel sound. The topic of wavefield syn-thesis came up frequently, in conjunction with presentationsfrom members of the European CARROUSO project that isnearing its close. They also offered a demonstration of itstechnology in a room at the RAI Centre.

A paper from Matti Karjalainen and some of his studentsdiscussed the innovative concept of wearable augmented real-ity, involving small headphones mounted in the ear canals, inconjunction with miniature microphones and a real-time au-dio signal processing system that can feed the ears with syn-thetically processed spatial signals to complement naturalacoustic cues. This fascinating paper showed the direction formuch future work in this area and was reminiscent of somescience-fiction concepts wherein communications tools areintegrated seamlessly with the subject’s natural senses. Alsofascinating in this field was a paper by Daniel, Nicol, andMoreau comparing high-order ambisonics and wavefield syn-thesis, showing that there were more similarities than mightotherwise have been thought, since both attempt some formof accurate wavefront reconstruction in the listening area. In asubsequent session Thomas Sporer likened this situation tothat which had existed earlier in the field of low bit-rate cod-ing, when people argued about the merits of transform codingversus subband coding, eventually coming to realize thatthese were really just two sides of the same coin, so to speak.

Gilbert Soulodre’s excellent paper on spatial measure-ments corresponding to listener envelopment showed theresults of numerous correlation analyses to discover themost reliable combination of gain (G) and spatial (S) com-ponents, leading to a new term GSperc. The audience wasamused to note the initials of this term and wondered if

AES 114th is year’s premier Pro Audio event in Europe.

Historical Cornerdrew crowds todisplays of vintageequipment andlectures, such asTim de Wolf’s(above) on record-cutting techniques.

Ticketholders boarding bus for technical tour.


114th Convention

their derivation had been more than a coincidence.There was a strong emphasis on signal processing and

low bit-rate coding during the last two days of the conven-tion, including a number of papers on high-resolution con-version and novel modulators for sigma–delta applications.Papers from James Angus and Harpe, Reefman andJanssen dealt with a new type of SDM called a Trellis con-vertor that is designed to give lower distortion and S/N ra-tios than former types, in this case with lower computa-tional load than previously thought necessary.

Further sessions covered microphones; automotive audioand instrumentation and measurement; room acoustics andsound reinforcement; audio networking; and analysis andsynthesis of sound. There were also a large number ofposter presentations in the Topaz Lounge outside the con-ference rooms where authors met with interested delegatesto discuss their work. A full listing of all papers and

114th banquet was held in elegant Winter Garden of GrandHotel Krasnapolsky. Music was provided by Pumps in Blue.


tations were also given each day by experts in their fields ontopics such as the history of record cutting, vintage measuringequipment, ancient electronic music, and other topics.

SPECIAL EVENTSBoth the AES mixer party and the banquet provided vibrantopportunities for delegates to relax and enjoy themselves to-gether with their colleagues, thanks to the planning of JanRomijn. The banquet, held this year in the elegant Winter Gar-den of the Grand Hotel Krasnapolsky, was a highlight of theconvention, accompanied by the ensemble Pumps in Blue.

Beer is arguably a part of the staple diet of the Netherlan-ders, and a few people were able to enjoy it on a tour of theHeineken Brewery Experience at the Stadhouderskade in Am-sterdam. This tour took visitors through a voyage of discoveryinvolving the world of Heineken Beer, including its malt silosand brew house, enhanced by a state-of-the-art audio-visualsystem installed by Mansveld-Eindhoven, whose representa-tives were on hand to answer technical questions.

Student recordingcompetition drew largecrowd; Martha DeFrancisco (left) helpedwith judging.

Ron Streicher, second from left, gave copies of hisbook to winners of student recording competition:from left, Raphel Allain (jazz/folk), Kent Walker(classical), and Thomas Geiger (pop/rock).

their abstracts and the complete list of workshops begins onpage 405 of this issue. A CD-ROM of all the 114th Con-vention papers is available.

TECHNICAL TOURSAn appealing program of technical tours, organized by Mennovan der Veen, provided an opportunity to visit a number of ex-ternal venues. These included Polyhymnia Classical RecordingCentre in Baarn, where delegates were treated to a demonstra-tion of the company’s high resolution SACD surround record-ings. Visitors could also take trips to Dutch View and NOBCross Media Services in Hilversum; Record Industry vinylpressing plant in Haarlem; Philips Research Laboratories inEindhoven; and Delft University of Technology.

HISTORICAL PRESENTATIONSThe Historical Corner on the exhibition floor, coordinated byStan Tempelaars and Alex Balster, contained examples ofvintage recording and measuring equipment as well as a num-ber of ancient musical instruments. A number of short presen-

Newly elected SDA officers Natalia Teplova (secondfrom left) and Martin Berggren (right) are joined by

outgoing officers Feliz Dreher and Isabella Biedermann.


114th Convention

Conference Policy Committee meeting: from left, JanPedersen, Nick Zacharov, and Roger Furness.

Technical Council meeting: from left, Jürgen Herre, vicechair; Wieslaw Woszczyk, chair; and Bob Schulein, vice chair

John Nunn (left), Standards Committee chair, and MarkYonge, Standards manager

Publications Policy Committee meeting: Richard Small (inset), chair; fromleft, Stanley Lipshitz, Søren Bech, and Kees Immink;

STUDENT ACTIVITIESStudents were particularly involved in the tutorial semi-nars mentioned above. The seminars were a big attractionfor those at the earlier stages in their careers. Other educ-tion student activities, organized by Werner de Bruijn, in-cluded the Education Fair, which provided a forum fordiscovering information about training courses, and theJob/Career Seminar, which addressed the issue of entry-level employment for those just graduating. The StudentDelegate Assembly (SDA) met twice to discuss nomina-tions and to elect student representatives for Europe.There was also an exciting recording competition and aposter session demonstrating the exceptionally high stan-dard of practical and academic work created by studentsof AES sections around the world.

TECHNICAL COUNCIL AND STANDARDSCOMMITTEESThroughout the convention a full program of standards andtechnical committee meetings took place, thanks to the

sterling work of Mark Yonge and John Nunn for Stan-dards, and Wieslaw Woszczyk, Jürgen Herre, and BobSchulein for the Technical Council. A new technical com-mittee on semantic audio analysis was formed at the 114th.Chaired by Mark Sandler, this committee intends to dealwith topics such as music pitch and rhythm analysis,meaning extraction, and semantic content analysis. Thetiming is good for the creation of this new committee, fol-lowing the plea of Heyser Lecturer Jens Blauert for audioto “go cognitive.”

Everyone who attended the 114th is now more awareof the challenges/opportunities facing the industry. Thecity of Amsterdam, again, provided a welcome atmo-sphere for the exchange of ideas and information centralto the mission of the AES. Be cognitive of the next op-portunity to meet your colleagues and learn of the latestscientific and commercial advancements at the AES115th Convention in New York October 10–13. (For de-tails on all upcoming Society activit ies visitwww.aes.org.)

Regions and Sections meeting: SubirPramanik (left), chair, and Neville Thiele,vice president, International Region

Mark Yonge(left), Standards

manager, withGeorge

Massenburg,chair of Technical

Committee onStudio Practicesand Production

114th Convention

5


3

1

BoarBoard of Gd of Goverovernors nors MMeetseets

Kees Immink, president; Jay Fouts, legal counsel

David Robinson, governor; Roland Tan, governor;Wieslaw Woszczyk, Technical Council chair

Garry Margolis, past president and NominationsCommittee chair

Han Tendeloo, secretary

Curtis Hoyt, governor; Richard Small, PublicationsPolicy Committee chair; Mercedes Onorato, LatinAmerican Region vice president

Kunimaro Tanaka, governor

Annemarie Staepelaere, governor; Irv Joel,Historical Committee vice chair; Neville Thiele,International Region vice president

7

6

5

4

3

2

1

Meeting on March 26, members of the AES Board ofGovernors gather from around the world to hear reports

from AES officials and standing committees:

Søren Bech, Europe Northern Region vice presidentand Conference Policy Committee chair; MarkusErne, Europe Central Region vice president; DanielZalay, Europe Southern Region vice president; RonStreicher, president-elect

Marshall Buck, treasurer, Convention PolicyCommittee chair, and Finance Committee chair

Daniel von Recklinghausen, editor; Subir Pramanik,Regions and Sections Committee chair

John Nunn, Standards Committee chair; RogerFurness, executive director

Karl-Otto Bäder, governor; James Kaiser,USA/Canada Central Region vice president; JimAnderson, USA/Canada Eastern Region vicepresident

12

11

10

9

8

2

4

114th Convention


8 9

12

6

11

7

10


111144thth EEXXHHIIBITBITORSORSAmsterdam, The Netherlands

2003 March 22–25

A.P.R.S.Reading, UKwww.aprs.co.uk

ACME/Audio ProBrussels, Belgiumwww.cahiersacme.org

Acoustics EngineeringBoxmeer, The Netherlandswww.acoustics-engineering.com

Adam AudioBerlin, Germanywww.adam-audio.de

ADK MicrophonesVancouver, WA, USAwww.adkmic.com

Advanced ComputerSystems S.p.A.Rome, Italywww.acsys.it

AEQ, S.A. Léganes (Madrid), Spainwww.aeq.es

*AKG Acoustics GmbH Vienna, Austriawww.akg.com

Alcons AudioAmsterdam, TheNetherlandswww.alconsaudio.com

Allen & Heath Ltd. Penryn, UKwww.allen-heath.com

AM 3DÅlborg, Denmarkwww.am3d.com

AMPTEC bvba HASSELT, Belgiumwww.amptec.be

*AMS—Neve plc.Burnley, UKwww.ams-neve.com

Aphex Systems Ltd.Sun Valley, CA, USAwww.aphex.com

APT—Audio ProcessingTechnologyBelfast, Northern Irelandwww.aptx.com

Arbor AudioCommunicationsDoetinchem, TheNetherlandswww.arbor-audio.com

ArkamysParis, Francewww.arkamys.com

ASL Intercom bvUtrecht, The Netherlandswww.asl-inter.com

Audemat / AztecBordeaux-Merignac,Francewww.audemat-aztec.com

Audient Watford, UKwww.audient.co.uk

Audio EngineeringLondon, UKwww.micronwireless.co.uk

*Audio PrecisionBeaverton, OR, USAwww.audioprecision.com

Audio Pro Magazinewww.mi-pro.co.uk

Audio Pro NederlandAmsterdam, TheNetherlandswww.audio-pro.nl

Audio Video EchosParis, Francewww.edgnet.com

AudionicsSheffield, UKwww.audionics.co.uk

Audioscope 2KRome, Italywww.audioscope.it

Audio-Service UlrichSchierbecker GmbHHamburg, Germanywww.audio-service.com

*Audio-TechnicaLeeds, UKwww.audio-technica.co.uk

Avalon DesignSan Clemente, CA, USAwww.avalondesign.com

AviomWest Chester, PA, USAwww.aviom.com

AVT Nürnberg, Germanywww.avt-nbg.de

BeldenVenlo, The Netherlandswww.belden-europe.com

BrainstormGenval, Belgiumwww.plus24.com

BrainstormW. Hollywood, CA, USAwww.brainstormtime.com

Brauner MicrophonesHamminklen, Germanywww.brauner-microphones.com

Brüel & KjaerNaerum, Denmarkwww.bksv.com

*Cadac Electronics PLCLuton, UKwww.cadac-sound.com

*Calrec AudioHebden Bridge, UKwww.calrec.com

*Canford Audio plcTyne & Wear, UKwww.canford.co.uk

CB ElectronicsCharvil, UKwww.colinbroad.com

*CEDAR Audio Ltd.Fulborn, UKwww.cedaraudio.com

Cirrus Logic Inc.Austin, TX, USAwww.cirrus.com

Clyde BroadcastProducts Ltd.Clydebank, UKwww.clydebroadcast.com

C-MEXX SoftwareWeissenhohe, Germanywww.c-mexx.de

Coding TechnologiesNürnberg, Germanywww.codingtechnologies.com

ComrexDevens, MA, USAwww.comrex.com

CSTBSt. Martin D’heres, Francewww.cstb.fr

Cube-TecAlbstadt, Germanywww.cube-tec.com

d & b audiotechnikBacknang, Germanywww.dbaudio.com

D & RWeesp, The Netherlandswww.d-r.nl

D.A.V.I.D.Munich, Germanywww.digasystem.com

DaletParis, Francewww.dalet.com

Dangerous MusicNew York, NY, USAwww.dangerousmusic.com

Dateq AudioTechnologiesAlmere, The Netherlandswww.dateq.nl

Delec Audio-undVideotechnikGölheim, Germanywww.delec.de

DHD GmbHLeipzig, Germanywww.dhd-audio.de

DiGiCo UK Ltd.Epsom, UKwww.digiconsoles.com

________________*Sustaining Member of the Audio Engineering Society

*Digidesign, a division ofAvid TechnologyIver Heath, UKwww.digidesign.com

*DigigramMontbonnot, Francewww.digigram.com

DK-AudioHerlev, Denmarkwww.dk-audio.com

*Dolby LaboratoriesWiltshire, UKwww.dolby.com

Domino Design GmbHWiesbaden, Germanywww.domino-design.de

DPA Microphones A/SAllerød, Denmarkwww.dpamicrophones.com

DrawmerWakefield, UKwww.drawmer.com

*DTSTwyford, UKwww.dtsonline.com

Duran Audio bv.Zaltbommel, TheNetherlandswww.duran-audio.com

DVD-Audio Group c/oMeridian Audio GroupHuntingdon, UKwww.dvdaudio-info.com

Earthworks AudioProducts, Inc.Milford, NH, USAwww.earthworksaudio.com

EBH Radio SoftwareBremen, Germanywww.ebh.com

Emes Studio MonitorSystemsBurgau, Germanywww.emes.de

Empirical Labs, Inc.Lake Hiawatha, NJ, USAwww.empiricallabs.com

Euphonix Europe OfficeHarrow, UK

EventideLittle Ferry, NJ, USAwww.eventide.com

FAR—FundamentalAcoustic ResearchOugree, Belgiumwww.far-audio.com

Focal—JM Lab.Saint Etienne, Francewww.jmlab.fr

Fostex CorporationTokyo, Japanwww.fostex.co.jp

*Fraunhofer IISErlangen, Germanywww.iis.fraunhofer.de

Friend-ChipBerlin, Germanywww.friend-chip.de

Genelec OyIisalmi, Finlandwww.genelec.com

Genex Audio Inc.Santa Monica, CA, USAwww.genexaudio.com

Geoffrey Daking and Co.Inc.Wilmington, DE, USAwww.daking.com

GhielmettiBiberist, Switzerlandwww.ghielmetti.ch

George Massenburg LabsFranklin, TN, USAwww.massenburg.com

Harman Pro GroupNorthridge, CA, USAwww.jblpro.com

Hearsafe TechnologiesKöln, Germanywww.hearsafe.de

I.B.E.Redhill, UKwww.ibeweb.com

Imas PublishingSaint Ives, UKwww.audiomedia.com

*Innova SON SAPlougoumelen, Francewww.innovason.com

Institut fürRundfundtechnik GmbHMunich, Germanywww.irt.de

Interface Media BV—Shiva DMAAmsterdam, TheNetherlandswww.interface.nl

Inter-M / AlgorithmixKarlsruhe, Germanywww.algorithmix.com

J & C Intersonic AGRegensdorf, Switzerlandwww.jcintersonic.com

JK Audio, Inc.Sandwich, MA, USAwww.jkaudio.com

Jünger AudioBerlin, Germanywww.junger-audio.com

Jutel Oy Oulu, Finlandwww.jutel.fi

KeyboardsMontreuil, France

Klippel GmbHDresden, Germanywww.klippel.de

KM Studio SystemsKarlskoga, Swedenwww.kmstudiosystems.com

KS DigitalIllingen, Germanywww.ksdigital.de

Lake PeopleKonstanz, Germanywww.lake-people.de

LawoRastatt, Germanywww.lawo.de

Le Guide RadioDonzenac, Francewww.leguideradio.com

LectrosonicsRio Rancho, NM, USAwww.lectrosonics.com

LexiconBedford, MA, USAwww.lexicon.com

Line UpHorsham, W. Sussex, UK

LinkRome, Italywww.linkitaly.com

Linn ProductsGlasgow, UKwww.linn.co.uk

Listen Inc.Boston, MA, USAwww.listeninc.com

Live SoundSan Francisco, CA, USAwww.livesoundint.com

Logitek ElectronicSystemsHouston, TX, USAwww.logitekaudio.com

Lundahl TransformersNorrtälje, Swedenwww.lundahl.se

LydkraftVanløse, Denmarkwww.tube-tech.com

Mandozzi Elettronica sa.Ponte Capriasca,Switzerlandwww.mandozzi.ch

Manley LabsChino, CA, USAwww.manleylabs.com

Marantz ProfessionalEuropeLongford, UKwww.marantz.com

M-AudioHemel Hempsted, UKwww.maudio.co.uk

Mayah CommunicationsHallbergmoos, Germanywww.mayah.com

Maycom Audio SystemsElst, The Netherlandswww.maycom.nl

MB Quart GmbHObrigheim, Germanywww.mbquart.com

Mecastep OyVillahde, Finlandwww.mecastep.fi

Media UtilitiesNaarden, The Netherlandswww.media-utilities.nl

Mediarte GmbHDüsseldorf, Germanywww.mediarte.com

Merging TechnologiesPuidoux, Switzerlandwww.merging.com

Michael Stevens &Partners Ltd.Bromley, UKwww.michael-stevens.com

Milab Helsingborg, Swedenwww.milabmic.com

Millenia MediaPlacerville, CA, USAwww.mil-media.com

Miller & KreiselInternationalSolrød Strand, Denmarkwww.mkprofessional.com

Miller & Kreisel SoundInc.Chatsworth, CA, USAwww.mkprofessional.com

Minnetonka AudioSoftwareMinnetonka, MN, USAwww.minnetonkaaudio.com

mLAN AllianceIwata-Gun, Japanwww.yamaha.co.jp/tech/1394mLAN/english

MogamiTokyo, Japanwww.mogami-wire.co.jp

MT Gefell Gefell, Germanywww.microtechgefell.de

Musicam USAHolmdel, NJ, USAwww.musicamusa.com

Musikelektronik GeithainGmbHGeithain, Germanywww.me-geithain.de

Mutec GmbHBerlin, Germanywww.mutec-net.de

Nagra Nagravision SAKudelski GroupCheseaux, Switzerlandwww.nagra.com

National SemiconductorGmbHFürstenfeldbruck, Germany

NetiaClaret, Francewww.netia.net

*Georg Neumann GmbHBerlin, Germanywww.neumann.com

*Neutrik AGSchaan, Liechtensteinwww.neutrik.com

NHT ProBenicia, CA, USAwww.nhtpro.com

114114ththConventionConventionExhibitors


NOA Audio SolutionsVertriebsgesmbHVienna, Austriawww.noa-audio.com

NTISchaan, Liechtensteinwww.nt-instruments.com

NTP AV GroupHerlev, Denmarkwww.ntp.dk

NTP System EngineeringHerlev, Denmarkwww.ntp.dk

NXT—New TransducersLtd.Huntingdon, UKwww.nxtsound.com

OmniaCleveland, OH, USAwww.omniaudio.com

OpticomErlangen, Germanywww.opticom.de

ORBAN/CRL SystemsInc.San Leandro, USAwww.orban.com

Otari Europe GmbHMeerbusch, Germanywww.otari.com

Panphonics Ltd.Espoo, Finlandwww.panphonics.fi

Pearl MicrophonesAstorp, Swedenwww.pearl.se

Philips International /Super Audio CDEindhoven, TheNetherlandswww.superaudiocd.philips.com

Plus24W. Hollywood, CA, USAwww.plus24.net

PMI Torrance, CA, USAwww.pmiaudio.com

*Prism Media ProductsCambridge, UKwww.prismsound.com

Pro Audio AsiaTunbridge Wells, UKwww.proaudioasia.com

Pro Audio VisieBuren, The Netherlandswww.contekst.nl

Pro Sound & VisionAthens, Greecewww.avalonline.gr

ProdysLeganes (Madrid), Spainwww.prodys.net

ProkomBerlin, Germany

PSI Audio / RelecYverdon, Switzerlandwww.psiaudio.com

QuestedHarefield, UKwww.quested.com

Radikal TechnologiesDeutschlandMunich, Germanywww.radikaltechnologies.com

Radio LinkXalandri Athens, Greecewww.radio-link.gr

Reel DrumsRingwood, NJ, USAwww.reeldrums.com

Reference Laboratorys.r.l. Osimo, Italywww.referencelaboratory.com

Renkus-HeinzFoothill Ranch, CA, USAwww.renkus-heinz.com

ResolutionLondon, UKwww.resolution.com

Rohde & SchwarzMunich, Germanywww.rohde-schwarz-com

RosendahlUtting, Germanywww.rosendahl-studiotechnik.de

RTWKöln, Germanywww.rtw.de

Rycote MicrophoneWindsheilds Ltd.Stroud, UKwww.rycote.com

S.E.A. Vertrieb &Consulting (West)Emsbüren, Germanywww.sea-vertrieb.de

Salzbrenner StagetecButtenheim, Germanywww.stagetec.com

Sanken Microphone Co.,Ltd.Tokyo, Japanwww.sanken-mic.com

SchoepsKarlsruhe, Germanywww.schoeps.de

SD Systems InstrumentMicrophonesAmsterdam, TheNetherlandswww.sdsystems.com

Sek’DW. Hollywood, CA, USAwww.sekd.com

Sennheiser ElectronicCorporationWedemark, Germanywww.sennheiser.com

Servicios De RadioWavenetCedillo Del Condado, Spainwww.arrakis.es/~wavenet/qsi.htm

*Shure GmbHHeilbronn, Germanywww.shure.com

*Solid State Logic Ltd.Oxford, UKwww.solid-state-logic.com

Sonic SolutionsNovato, CA, USAwww.sonic.com

Sonic Studio LLCPlymouth, MN, USAwww.sonicstudio.com

SonifexNorthants, UKwww.sonifex.co.uk

Sono MagazineParis Cedex 19,Francewww.sonomag.com

Sonosax SAS SALe Mont, Switzerlandwww.sonosax.com

SonovisionParis, Francewww.sonovision.com

*Sony Europe / SuperAudio CDBadhoevedorp, TheNetherlandswww.superaudio-cd.com

Sony Overseas SASchlieren, Switzerlandwww.sonybiz.net

Sound and VisionLandsberg am Lech,Germanywww.modernezeiten.org

Sound KeysSuresnes, Francewww.soundkeysmag.com

*Sound On Sound Ltd.Cambridge, UKwww.sospubs.co.uk

SoundfieldWakefield, UKwww.soundfield.com

SoundmanagerInternational ASÅlesund, Norwaywww.ntp.dk

*Stage Accompany B.V.Hoorn, The Netherlandswww.stageaccompany.com

Stage TecEntwicklungsgesellschaftBerlin, Germanywww.stagetec.com

Steinberg MediaTechnologies AGHamburg, Germanywww.steinberg.net

*Studer ProfessionalAudio AGRegensdorf, Switzerlandwww.studer.ch

Studio 22 GmbHKarlsruhe, Germanywww.musictouch.de

Studio Audio + Video Ltd.Ely, UKwww.sadie.com

Studio Box GmbHWalzbachtal, Germany -www.studiobox.de)

Studio MagazinOberhausen, Germanywww.studio-magazin.de

SyphaLondon, UKwww.syphaonline.com

TamuraTokyo, Japanwww.qolle.com

TC Electronic A/SRiskov, Denmarkwww.tcelectronic.com

Teac / TascamWiesbaden, Germanywww.tascam-europe.de

Telos SystemsCleveland, OH, USAwww.telos-systems.com

Texas InstrumentsVilleneuve-Loubet, Francewww.ti.com

Thum + Mahr GmbHMonheim, Germanywww.thummahr.de

THX, Ltd.San Rafael, CA, USAwww.thx.com

TM Audio Holland BVUtrecht, The Netherlandswww.tmaudio.nl

Top FormatHaarlem, The Netherlands

Total ProductionStockport, UK

Truth AudioSanta Rosa Beach, FL,USAwww.truthaudio.com

TSLMaidenhead, UKwww.televisionsystems.ltd.uk

UltrasonePenzberg, Germanywww.ultrasone.com

*United EntertainmentMediaLondon, UKwww.cmpinformation.com

Universal AudioSanta Cruz, CA, USAwww.uaudio.com

VCS AktiengesellschaftBochum, Germanywww.vcs.de

VDTBergisch Gladbach,Germanywww.tonmeister.de

Walters-Storyk DesignGroup EuropeLiestal, Switzerlandwww.wsdg.com

Wave DistributionRingwood, NJ, USAwww.wavedistribution.com

Wave Mechanics Inc.Burlington, VT, USAwww.wavemechanics.com

WavesKnoxville, TN, USAwww.waves.com

Weiss Engineering Ltd.Uster, Switzerlandwww.weiss.ch

WysicomRomano D'Ezzelino, Italywww.wisycom.com

*Yamaha Corp.Hamamatsu-Shi, Japanwww.global.yamaha.com

Yamaha MusicVianen, The Netherlandswww.yamaha.nl

Yamaha Music CentralEurope GmbHRellingen, Germanywww.yamaha-europe.com

YellowtecMonheim, Germanywww.yellowtec.com

YOU/COMAudiocommunicatie bv.Delft, The Netherlandswww.youcom.nl

Zenon MediaWillstätt, Germanywww.zenon-media.com

ZPlane.DevelopmentBerlin, Germanywww.zplane.de


114114ththConventionConventionExhibitors


Thursday, March 20 13:30 h Room 1Standards Committee Meeting on SC-02-02 Digital Input/Output Interfacing

18:00 h Room 1Standards Committee Meeting on SC-02-05 Synchronization

Friday, March 21 10:30 h Room 2Standards Committee Meeting on SC-06-02 AudioApplications of IEEE 1394

13:00 h Room 1Standards Committee Meeting on SC-06-04 InternetAudio Delivery System

16:00 h Room 1Standards Committee Meeting on SC-05-05 Grounding and EMC Practices

Session A Saturday, March 22 09:00–12:30 hRoom C/D

MICROPHONES

Chair: David Josephson, Josephson Engineering, Santa Cruz, CA, USA

09:00 h

A-1 Autodirective Dual Microphone—Alexander A.Goldin, Alango Ltd., Haifa, Israel

The paper describes Autodirective Dual Microphone(ADM) technology and its applications. ADM digitalsignal processing technology developed by AlangoLtd. is an adaptive beamforming technology thatuses only two closely spaced omnidirectional soundpressure sensors. ADM technology provides the op-timal and variable directivity in every frequency re-gion. The adaptation time is very fast, leading tovery good improvement in signal-to-noise ratio infast changing noisy environments. Contrary to regu-lar directional microphones, an ADM technology-based microphone is much less sensitive to windnoises and does not effect proximity. Its DSP imple-mentation is relatively simple requiring very modestcomputational resources.[Paper not presented at convention, but ConventionPaper 5715 is available.]

09:30 h

A-2 Circular Microphone Array for Discrete Multi-channel Audio Recording—Edo Hulsebos,

Thomas Schuurmans, Diemer de Vries, RinusBoone, Delft University of Technology, Delft, TheNetherlands

Traditional stereo microphone pair techniques fornatural recording are quite capable for two-channelstereo reproduction. However, for multichannel re-production systems like 5.1, 7.1, Ambisonics, andwave field synthesis compromises in terms of cover-age, source localization, and channel separation are unavoidable. The main reason for this is that microphones currently used only have low order directivity patterns (omni, figure-eight, cardioidor hypercardioid) that cannot provide sufficient angu-lar resolution to avoid unwanted cross talk betweenthe recording channels. In this paper a discrete coin-cident 12-channel microphone is proposed in orderto solve these problems. This microphone consists ofa circular array with a radius of 1 meter using 288 mi-crophone capsules whose output signals are com-bined into 24 channels using simple analog electron-ics. These 24 channels are captured using amultitrack computer interface and post-processedinto up to 12 discrete reproduction audio channels.Convention Paper 5716

10:00 h

A-3 A New Comprehensive Approach of SurroundSound Recording—Arnaud Laborie, Rémy Bruno,Sébastien Montoya, ImmersiveSound Project,Paris, France

Many techniques have been developed concerning surround sound recording and the issue has turnedout to be a challenge without a comprehensive theory.This paper presents an approach based on a full 3-Dacoustic field theory and the use of a 3-D microphonearray. Our research work has led to a new spatial digi-tal processing technique which allows the use of freelypositioned capsules of any type such as omnidirec-tionals, bidirectionals, or cardioids. This technique canbe seen as an extension of full sphere, generalizedAmbisonics which provides a spatial resolution neverachieved before but requires neither high order direc-tivity capsules nor assumes that all capsules are coin-cident. The theory has been validated with a fullsphere third-order prototype using 24 omnidirectionalcapsules. The theory also allows for a fifth-orderspherical harmonic, multichannel 5.0 microphone.Convention Paper 5717

10:30 h

A-4 The Mechanisms Creating Wind Noise in Micro-phones—Stuart Bradley1, Tao Wu1, Sabine von

AAEESS 111144thth

CCOONNVVEENNTTIIOONN2003 March 22–25

RAI Conference and Exhibition CentreAmsterdam, The Netherlands

Technical Technical PPrrooggraramm


Hünerbein1, Juha Backman21Acoustics, University of Salford, Salford, Greater Manchester, UK

2Nokia Mobile Phones, Espoo, Finland

This paper identifies how wind noise is generated inmicrophones and how important different mechanismsare at different frequencies. Studies in a quiet windtunnel have been performed on the noise spectrumfrom microphones embedded in a turbulent air streamof known characteristics. The effect of the micro-phone’s housing in creating a turbulent boundary layeris included. Scaling relationships are found which giveinsight into noise-generating mechanisms and the ef-fects of microphone geometry and placement. Thenoise spectrum consists of thee regimes: a constantnoise level band at low frequencies; and 1/f behaviorbut with different generating mechanisms and scalingdependence on wind speed and other variables in twodistinct higher frequency bands. Appropriate noise re-duction schemes are explored.Convention Paper 5718

11:00 h

A-5 Integrated Circuits for High Performance Electret Microphones—Arie van Rhijn, NationalSemiconductor, Santa Clara, CA, USA

Electret condenser microphones (ECM) are used in al-most every consumer and communication audio appli-cation with a total yearly volume of well over one billionunits. Over the years, ECM innovation has concentrat-ed mainly on lower production cost and smaller sizes,but improvements of sensitivity, signal-to-noise ratio(SNR), linearity, and supply current have not been ad-dressed. In this paper we introduce significant innova-tions to ECMs though a new amplifier IC and packagedesigns. First, we describe current ECMs using junctionfield effect transistors (JFETs), focusing on advantagesand drawbacks. After that, we discuss two analog am-plifiers replacing the JFET in an ECM, showing the per-formance improvements achieved. Finally, further inno-vation though complete analog-to-digital conversioninside the ECM is presented.Convention Paper 5719

11:30 h

A-6 Radio Frequency Susceptibility of Capacitor Microphones—Jim Brown1, David Josephson21Audio Systems Group, Inc., Chicago, IL, USA2Josephson Engineering, Santa Cruz, CA, USA

Neil Muncy has shown that improper termination ofshield wiring, commonly called the pin 1 problem,couples noise currents flowing on a cable shield intoaudio circuitry though common impedance coupling.Inadequate bandwidth limiting of the microphone’sline driver and decoupling of its phantom power cir-cuits can also allow a path for radio frequency (RF)interference. This paper examines the susceptibilityof modern microphones, describes a simple test tofind problems, and offers simple solutions.Convention Paper 5720

12:00 h

A-7 Microphone Array Beamforming forMultichannel Recording—Juha Backman, Nokia Oy, Espoo, Finland; Helsinki University ofTechnology, Espoo, Finland

A new multichannel microphone technique, basedon using a microphone array with some inherent di-

rectivity, combined with DSP beamforming, is intro-duced. The advantage of the new method overpurely acoustical or simple analog polar patterncontrol is added freedom in defining the polar pat-tern shape, resulting in more precise control of pan-ning laws and side-lobe behavior. Alternative micro-phone arrangements, e.g., directional microphonearrays or sphere-mounted arrays, are discussed.Convention Paper 5721

Workshop 1 Saturday, March 22 09:00–11:00 hRoom A

LOW BIT-RATE CODING OF SPATIAL AUDIO

Chair: Chistof Faller, Agere Systems, Murray Hill, NJ, USA

Presenters: Mark Davis, Dolby Laboratories, San Francisco, CA, USAMichael C. Kelly, York University, Heslington York, UK Gerald Schuller, Fraunhofer-IIS, Ilmenau, GermanyThomas Sporer, Fraunhofer-IIS, Ilmenau, Germany

While entertainment brings high fidelity surround soundinto most homes, a need arises for coding technologiesfor efficiently storing and transmitting stereo and multi-channel audio signals. The history of spatial audio repro-duction and spatial audio coding is reviewed. An intro-duction to spatial perception in terms of binaural hearingand auditory scene analysis is given. Currently usedtechniques such as sum/difference coding and intensitystereo are described in detail. Recent techniques for sig-nificantly lowering the bit rate for stereo and multichannelaudio coding are described. An outlook is presented withrespect to further developments in terms of low bit-ratespatial audio coding. The workshop is complementedwith a variety of demonstrations.

Seminar 1 Saturday, March 22 09:00–11:00 hRoom B

BASICS OF DIGITAL AUDIO

Presenters: Stanley Lipshitz, John Vanderkooy,University of Waterloo, Canada

This is an introductory-level seminar aiming to explainand demonstrate with “live” examples the two fundamen-tal aspects of any digital audio system—sampling andquantization. These two operations will be discussed andillustrated in real time using a custom-built sampler andquantizer. This will enable us to present some of thepathologies of such systems, which should not normallybe audible, and also show that, when properly imple-mented, a digital system has analog characteristics. Thiswill make the presentation interesting to newcomers and“old pros” alike.

Topics to be covered will include:— Sampling only (without quantization)— Sampling artifacts (aliases and images)— Reconstruction— Quantization only (without sampling)— Quantization errors— Dither

The demonstrations will enable the audience to hearand see what is going on, both good and bad.

Session B Saturday, March 22 09:30–12:30 hRoom E/F


PSYCHOACOUSTICS, PERCEPTION, AND LISTENINGTESTS, PART 1

Chair: Steven van de Par, Philips, Eindhoven, The Netherlands

09:30 h

B-1 Listening Technology for Automotive SoundSystems —David Clark, DLC Design, Wixon, MI,USA

A procedure for managing listening tests of automo-tive sound systems is presented. Techniques areselected to reduce listener bias and simplify the lis-tening task. These include comparison to a fixedreference sound, consistent source material, break-down of the listening task into many independentabsolute judgments, and ade-quate listener training. The overallrating is derived from a weightedsum of all of the individual judg-ments, not listener opinion.Convention Paper 5723

10:00 h

B-2 Perceptual Quality Assessmentof Telecommunications Systems Including Terminals—A. W. Rix1, J. Berger2, J. G.Beerends31Psytechnics Limited, Ipswich, UK2T-Systems Nova GmbH Berkom, Berlin, Germany

3TNO Telecom, Leidschendam, The Netherlands

Perceptual quality measurementmodels such as PESQ (ITU-TRecommendation P.862) are nowin common use for evaluating thespeech quality of communicationsnetworks and systems. Howeverthese models are mainly designedfor use with electrical or digital—not acoustic—interfaces to thesystems under test. This limits themodels’ applicability to terminals,in particular where the effects oftransducers, acoustics, and signalprocessing in the terminals maybe combined with network proper-ties such as low bit-rate codingand channel errors. This paperdescribes work under way in ITU-T SG12 to develop a new modelfor evaluation of both networksand terminals using acoustic inter-faces, and reports the latest re-sults in the development of a newITU-T Recommendation for thisapplication.Convention Paper 5724

10:30 h

B-3 The Significance of SpectralOverlap in Multiple-Source Localization—Michael C. Kelly,Anthony I. Tew, University ofYork, Heslington, York, UK

The mechanisms of human local-

ization for a single sound source are well understood,but less is known about how we localize multiple, si-multaneous sound sources. In rendering a complexvirtual auditory space (VAS), localization cues are ap-plied separately to each sound source and the resultsare summed-up to create a multiple-source environ-ment. In this paper we investigate the relevance ofthe inter-source spectral overlap that arises in such aVAS. We do so by adjusting the spatial cues in theseregions and comparing a listener’s localization abilityfor the modified and unmodified cases. We show howeven total removal of the weaker spectral compo-nents in regions of overlap has no effect on localiza-tion ability. Finally, we discuss the exploitation of re-dundancies in the regions of spectral overlap withrespect to multiple-source localization.Convention Paper 5725

T E S T F A S T E R F O R L E S SW I T H D S C O P E S E R I E S I I I

Ideal for:• Research & Development• Automated Production Test• Quality Assurance• Servicing• Installation

F o l l o w i n g c o m p l e t i o n o f o u r e x t e n s i v e b e t a - t e s tp r o g r a m , R e l e a s e 1 . 0 0 i s n o w a v a i l a b l e

Prism Media Products LimitedWilliam James House,Cowley Road, Cambridge. CB4 0WX. UK.

Tel: +44 (0)1223 424988Fax: +44 (0)1223 425023

[email protected]

Prism Media Products Inc.21 Pine Street, Rockaway, NJ. 07866. USA.

Tel: 1-973 983 9577 Fax: 1-973 983 9588

www.prismsound.com

dScope Series III issimply the fastest way to test.

Call or e-mail NOW to find out just how fast your tests can be!

DSNet I/O Switcher 16:2now available



11:00 h

B-4 Accuracy of the Listeners’ Estimation of theSpeakers’ Body Weight and Height Based Solely on the Voice Signal—Gordana Kovacic1,Hvoje Domitrovic21University of Amsterdam, Amsterdam, The Netherlands

2University of Zagreb, Zagreb, Croatia

The aim of this paper is to investigate the accuracyof listeners’ estimation of physical characteristicssuch as body weight and height from voice signalsalone. A series of listening tasks was carried out inwhich 20 adult male listeners judged body weightand height from 20 adult male speakers’ voice sam-ples. Additionally, listeners’ perception of thespeakers’ voice pitch was compared to actual voicepitch measured as fundamental frequency (F0) inHz. The results of the Pearson and Spearman cor-relation coefficients calculations indicated that thelisteners’ estimations of body height and weightwere completely wrong, and were actually misledby vocal stereotyping about physical characteristicswhere lower voice pitch was taken as a mark oflarger body weight and height, and vice versa.Convention Paper 5726

11:30 h

B-5 Wave Field Synthesis in the Real World: Part1—In the Living Room—Beate Klehs, ThomasSporer, Fraunhofer AEMT, Ilmenau, Germany

In anechoic rooms the concept of wave field syn-thesis (WFS) has already proven to provide superi-or spatial sound over a large part of the room. Theprogress in microelectronics enables WFS to be-come available in commercial products at a reason-able price. In the future, it will be installed in differ-ent acoustical environments. In anechoic spaceWFS needs a huge number of loudspeakers. Innormal listening conditions simulated and realacoustics interfere with each other making the gen-erated wave field less exact. This paper describeslistening tests conducted to evaluate WFS in com-mon living room conditions. Parameters under testare the number of loudspeakers, the distance be-tween loudspeakers, the position of the simulatedsource, and the position of listeners relative to theloudspeakers.Convention Paper 5727

12:00 h

B-6 Differences in Performance and Preference ofTrained versus Untrained Listeners in Loud-speaker Tests: A Case Study—Sean E. Olive,Harman International Industries, Inc., Northridge,CA, USA. Read by Durand Begault

Listening tests on four different loudspeakers wereconducted over the course of 12 months using 25different groups of listeners. The groups included184 untrained listeners whose occupations fell un-der one of four categories: audio retailer, marketingand sales, professional audio reviewer, and collegestudent. The loudspeaker preferences and perfor-mance of these listeners were compared to those ofa panel of 12 trained listeners. Significant differ-ences in performance, expressed in terms of mag-nitude of loudspeaker F-statistic (FL) were found between the different categories of listeners. Thetrained listeners were the most discriminating andreliable listeners with mean FL values 3 to 27 times

higher than the other four listener categories. Per-formance differences aside, loudspeaker prefer-ences were generally consistent between all cate-gories of listeners, providing evidence that thepreferences of trained listeners can be safely ex-trapolated to a larger population of untrained listen-ers. The highest rated speakers had the flattestmeasured frequency response maintained uniform-ly off-axis. Effects and interactions between train-ing, programs, and loudspeakers are discussed.Convention Paper 5728

Historical PresentationsHISTORICAL CORNERExhibition HallSaturday, March 22, 10:00–18:00 hSunday, March 23, 10:00–18:00 hMonday, March 24, 10:00–18:00 hTuesday, March 25, 10:00–17:00 h

On the Exhibition floor in the Historical Corner a perma-nent display of vintage recording and measuring equip-ment can be found together with a collection of ancientmusical instruments. From time to time interesting pre-sentations will be given, bringing back dear memories.

Workshop 2 Saturday, March 22 11:00–13:00 hRoom A

SAMPLING RATE CONVERTERS

Chair: Bert van der Wolf, NorthStar Recording Services, Haaften, The Netherlands

Panelists: Tony Faulkner, Green Room Productions, Harefield, Middlesex, UK; Emil Berliner Studios, Langenhagen, GermanySimon Heyworth, Super Audio Mastering, UKRonald Prent, Galaxy Studio, Mol, BelgiumMike Story, Data Conversion Systems, Great Chesterford, UKDaniël Weiss, Weiss Digital Audio, Zurich, Switzerland

For many years the compact disc has been the mainconsumer end product for producers of music produc-tions and sound engineers. It carries one standard sam-ple rate and word length. Since the introduction of theversatile DVD disc, many different formats can and willbe used for products such as DVD-video, DVD-audio,and SACD. All have their place in the market and theprofessional engineer is confronted with many new chal-lenges to make optimal use of these new media. Foreach new production, a choice has to be made aboutwhich format to use to start with, and the objective will beto reach an optimal level of quality in all down conver-sions for the different discs that will have to be producedat the end of the chain.

Seminar 2 Saturday, March 22 11:00–13:00 hRoom B

HEARING DAMAGE IN MUSICAL PRACTICE

Chair: Hans Verschuure, Erasmus University, Rotterdam, The Netherlands

Panelists: Eckhard Beste, Hearsafe, Köln, GermanyArja Boasson, Laren, The NetherlandsOlaf van Hees, CommitARBO, Diemen, The NetherlandsJan de Laat, Leiden MC, The Netherlands

The sound levels present at the workplace for performingmusicians, DJs, VJs or sound technicians often exceed


hearing safety levels. Many of these people are aware ofthis fact and actually notice the negative effects of extra-neous sounds like tinnitus and hearing loss. However,they tend to deny these negative effects because theyfear losing their jobs.

In this tutorial seminar panelists will explain the audito-ry system and the damage to it that is caused by extra-neous sounds. The effects of hearing loss will be demon-strated with regard to perceived sound quality andspeech intelligibility in quiet and noisy environments.Possible rehabilitative measures and their limitations willbe presented. We will show that protection is a far bettersolution than hearing rehabilitation.

To assess the protective needs some information is re-quired on sound levels. Information will be presented onactual sound levels during a performance. The problemsrelating to sound level measurement and interpretation inrelation to the effect of aging will be presented. The pro-tective needs can be concluded from this and practicaladvice will be given both with regard to hearing protec-tion and hearing rehabilitation.

11:00 h Room 2Standards Committee Meeting on SC-03-02 TransferTechnologies

Student ActivitiesEDUCATION FAIRSaturday, March 22, 12:00–15:00 hTopaz Lounge

The Education Fair is the perfect opportunity for repre-sentatives of educational institutions to present them-selves to potential new students and to share experi-ences with people from other schools. In this “tabletopsession,” information on each school’s respective pro-gram will be made available through the display of litera-ture and informal conversations with representatives.There is no charge for schools to participate and admis-sion is free and open to everyone.

12:40 h Room C/DTechnical Committee Meeting on Microphones andApplications

12:40 h Room E/FTechnical Committee Meeting on Audio for Telecommunications

13:00 h Room 2Standards Committee Meeting on SC-03-06 DigitalLibrary and Archive Systems

Session C Saturday, March 22 13:30–17:00 hRoom C/D

LOUDSPEAKERS, PART 1

Chair: David Clark, DLC Design, Wixon, MI, USA

13:30 h

C-1 Sensitivity of High Order LoudspeakerCrossover Networks with All-Pass Response—Brandon Cochenour, Carlos Chai, David A. Rich,Lafayette College, Easton, PA, USA

The sensitivity of high-order filter networks to compo-nent-matching tolerances increases with filter order.For an audio loudspeaker’s crossover network that isdesigned to sum to an all-pass network, we demon-strate that the sensitivity to component matching tol-

erances may be dwarfed by sensitivities to other ef-fects. In this paper we examine second-to eighth-or-der Linkwitz-Riley crossovers. The analysis also sub-sumes networks with transmission zeros andoptimized networks where the effects of frequency-re-sponse errors introduced by the driver’s respectivetransfer functions are minimized. We remark oncrossover networks that are least sensitive to thecombined effects of component tolerances, path-de-lay effects, the interaction of filter sections in speakersthat divide the incoming signal into thee or more sub-bands, and driver transfer functions.[Paper not presented at convention, but ConventionPaper 5729 is available.]

14:00 h

C-2 Automated In-Situ Frequency Response Optimization of Active Loudspeakers—AndrewGoldberg, Aki Mäkivirta, Genelec Oy, Iislami, Finland

This paper presents a novel method for robust auto-matic selection of optimal in-situ acoustical frequencyresponse within a discrete-valued set of responsesoffered by room response controls on an active loud-speaker. A frequency response measurement is usedas the input data for the algorithm. The rationale ofthe room response control system is described. Theresponse controls are described for each supportedloudspeaker type. The optimization algorithm is de-scribed. Examples of the optimization process aregiven. The efficiency and performance of the algo-rithm are discussed. The algorithm dramatically im-proves the speed of optimization compared to an ex-haustive search. It improves the acoustical similaritybetween loudspeakers in one space and performs ro-bustly and systematically in widely varying acousticalenvironments. The algorithm is currently in active useby specialists who set up and tune studios and listen-ing rooms.Convention Paper 5730

14:30 h

C-3 Measurement of Loudspeaker Amplitude Modu-lation Distortion—Richard H. Small, Harman/Becker Automotive Systems, Martinsville, IN, USA

Nonlinearities in full-range loudspeakers can resultin amplitude modulation of mid- and high-frequencycontent by strong low frequency components. Ameasurement based on an established two-tone in-termodulation distortion test can assess the amountof distortion produced and provide indication of thedominant nonlinearities causing the distortion. Thepaper discusses measurement and signal process-ing techniques as well as methods of data displayfor interpretation.Convention Paper 5731

15:00 h

C-4 Modeling Room Interaction Effects for Pistonic and Distributed-Mode Loudspeakers in Both the Frequency and Time Domains—Neil Harris1, Malcolm Hawksford21NXT Research Centre, Huntingdon, Cambridgeshire, UK

2University of Essex, Colchester, Essex, UK

In AES Paper 5215 presented at the 109th Conven-tion, both 2-D Finite Element Analysis (FEA) andmeasurements were used to examine the effects of asingle dominant reflection on the radiation of a loud-speaker. This earlier research is extended here by


410 . Audio Eng. Soc., Vol. 51, No. 5, 2003 May

exploiting an analytic 3-D solution to the problem ofan acoustic source located in a nonanechoic room.Unlike the earlier FEA solution, this method is mesh-less, potentially providing output at any point in spaceat any frequency. Application of the inverse Fouriertransform allows the method to model time-domainresults indirectly, thereby providing a complete timeand frequency domain description.Convention Paper 5732

15:30 h

C-5 Nonlinear Modeling of the Heat Transfer in Loudspeakers—Wolfgang Klippel, Klippel GmbH,Dresden, Germany

Traditional modeling describes the heat flow inloudspeakers by an equivalent circuit using integra-tors with different time constants. The parametersof the lumped elements are assumed to be inde-pendent of the amplitude of the signal. This simplemodel fails in describing the air convection coolingwhich becomes an effective cooling mechanism ifthe velocity of the coil and/or the velocity of theforced air in the gap becomes high. This paper pre-sents a large-signal model considering the nonlin-ear interactions between the electro-mechanicaland thermal mechanisms. The model and parame-ters are verified by practical measurements on thedrivers. The dominant paths for the heat flow areidentified and means for increasing the power han-dling capacity are discussed.Convention Paper 5733

16:00 h

C-6 Measurement of Impulsive Distortion, Rub andBuzz, and Other Disturbances—Wolfgang Klippel,Ulf Seidel, Klippel GmbH, Dresden, Germany

The traditional distortion measurement transformsthe time signal into the frequency domain to separatefundamental, harmonic, and intermodulation compo-nents. This technique considers only the mean pow-er in the analyzed interval and neglects the phase in-formation. A new technique for the measurement ofthe signal distortion in time domain is presented thatexploits both amplitude and phase information. It re-veals the fine structure of the distortion and its de-pendency on frequency, displacement or other statevariables. Besides the rms-value of the distortion, thepeak value and the crest factor are important charac-teristics for detection of rub and buzz phenomena.The practical application, the interpretation, and thediagnostics of defects are discussed.Convention Paper 5734

16:30 h

C-7 Effect of Porous Material on the Diffusivity of an Unbaffled DML Panel—Elena Prokofieva, Univer-sity of Bradford, Bradford, UK

The diffusivity of a standard distributed mode loud-speaker (DML) placed in an open base in the vicinityof a rigid board or a porous layer was investigated.The directivity pattern data achieved from the ane-choic chamber’s measurements were filtered of the1 octave and 1/3 octave. The dependence of the dif-fusivity function of the angle and frequency hasbeen investigated for an unbaffled DML panel withand without porous layer. The results show that theattachment of the porous material influences the di-rectivity acoustic properties of a DML panel.Convention Paper 5735

Session D Saturday, March 22 13:30–16:30 hRoom E/F

PSYCHOACOUSTICS, PERCEPTION, AND LISTENINGTESTS, PART 2

Chair: Durand Begault, NASA Ames Research, Mountain View, CA, USA

13:30 h

D-1 Listening Test Methodology for Headphone Evaluation—Toni Hirvonen1, Markus Vaalgamaa2, Juha Backman3, Matti Karjalainen11Helsinki University of Technology, Espoo, Finland2Nokia Mobile Phones, Helsinki, Finland3Nolia Mobile Phones, Espoo, Finland

Two listening tests with six different headphoneswere conducted using speech material. The objec-tives of these tests were to investigate listenersound color preferences using 1) the actual head-phones and 2) dummy head recordings made withthe same devices. The purpose of the recordingswas to simulate the timbres of the actual devices aswell as possible when played back though a pair ofcompensated headphones. The results from thetwo tests were compared and despite some similar-ity, the analysis showed significant differences be-tween the two cases. Additionally, the diffuse-fieldresponses of the headphones were calculated fromfrequency response measurements. The obtainedheadphone preference order cannot fully be explained based on the flatness of the diffuse-fieldresponse as a measure.Convention Paper 5736

14:00 h

D-2 Ideal Point Modeling of the Quality of NoisySpeech in Mobile Communications Based onMultidimensional Scaling—Ville-Veikko Mattila,Nokia Research Center, Tampere, Finland

Multidimensional scaling and preference mappingwere used for the perceptual analysis of the qualityof speech corrupted by car cabin noise in mobilecommunications. Forty-one processing chains, rep-resenting, e.g., transmission of speech over mobilenetworks, were studied. Thirty screened subjectswere used in the quality test and 15 screened andtrained subjects in the MDS test. Based on an exter-nal profiling of the auditory characteristics, the di-mensions appeared to relate to general naturalnessof speech, limitation of the frequency band ofspeech, smoothness of speech, and noisiness ofspeech. The Phase I, ideal point model was used topredict the quality with an average error of about 6percent, to study the interaction between the attribut-es and the linearity of the attributes.Convention Paper 5737

14:30 h

D-3 A Comparison of Speech Quality Judgments in Laboratory and in Real Environment—LaetitiaGros, Noël Chateau, Sylvain Busson, France Tele-com Research and Development, Lannion, France

The question of the validity of listening tests in alaboratory is considered in the case of the speechquality transmitted by mobile phones. First of all,tests are run outside in two places characterized bytwo different environments. Degradations are intro-


duced in the transmitted speech signal heard bysubjects, and sound environments are recorded aswell as transmitted speech signals. In a secondtime, tests are carried out in a laboratory by repro-ducing recorded speech signals though handsetsand recorded sound environments with a Dolby sur-round system. Results show a weak impact of thesound environment on quality judgments and vali-date the use of listening tests in a laboratory.Convention Paper 5738

15:00 h

D-4 Psychophysical Calibration of Sharpness forMultiparameter Distortion Effects Processing—William Martens, Atsushi Marui, University of Aizu,Fukushima-ken, Japan

A physical measure predicting apparent sharpnessfor an arbitrary input was a desired component in thedevelopment of a psychophysical calibration ofsharpness for multiparameter distortion effects pro-cessing. As a first step toward developing such ageneral purpose predictor, a set of listening testswas executed that included sharpness dominancejudgments and dissimilarity ratings for all pairwisecomparisons of nine stimuli. The outputs of thesetypes of distortion effects processors were filtered inorder to produce signals exhibiting thee values ofZwicker Sharpness, which is the conventional predic-tor for apparent sharpness calculated as the weight-ed first moment of the stimulus-specific loudnessfunction. INDSCAL analysis of the dissimilarity rat-ings for these nine stimuli showed that they variedprimarily along two perceptual dimensions. TheZwicker Sharpness values for the nine stimuli exhibit-ed a high correlation with coordinates on the first per-ceptual dimension. Including higher-order momentsof the specific loudness distribution did not improvethe sharpness prediction; however, for the small setof stimuli evaluated in this study, the fourth momentof the stimulus specific loudness, which was termed“Spectral Kurtosis,” was highly correlated with coordi-nates on the second INDSCAL-derived perceptualdimension. This second predictor sensitively mea-sured how peaked stimulus spectra were.Convention Paper 5739

15:30 h

D-5 The Significance of Phase as an Auditory Cue—Koray Ozcan1, Simon C. Busbridge1, Peter A.Fryer2, Gary P. Geaves2, Jon P. Moore11University of Brighton, Brighton, UK2Steyning Research Establishment, B&W Loudspeakers Ltd., Steyning, UK

Auralization results previously presented for interaur-al time and intensity conflict cue experiments are ex-tended by the inclusion of phase for multiple frequen-cy tone bursts and wideband signals. A method ispresented to manipulate the phase of all the compo-nent frequencies in a wideband signal while leavingthe amplitude structure unchanged though the use ofthe Hilbert transform. Therefore, phase and time be-come independent from each other for such signals.The results indicate that localization remains strongin the presence of large phase shifts. Furthermorethe central diffuse sound field that is characteristic ofintensity versus interaural time conflict experimentsis not present when the intensity and phase of wide-band signals are placed in conflict.Convention Paper 5740

16:00 h

D-6 Intelligibility of Reflected Sound Sources: Part1—Karl A. Sagrén1, Björn Hedquist21Bank of Brains, Dept. of R & D, Uppsala., Sweden2The Swedish Institute of CBT, Stockholm, Sweden

A common observation when listening to speech ormusic in acoustically live environments is that nat-ural acoustic sources often sound good and distrib-ute well, while the same sources reinforced and re-generated though a loudspeaker, behaves andsounds quite different in the same environment.While evaluating an experimental coaxial andcoplanar transducer, the Electro-Acoustic Converter(EAC), some interesting observations were made,which we briefly describe in this paper on the be-havior of regenerated complex waveforms in thesemireverberant sound field. One particularly inter-esting application is the possibility to distribute a re-inforced and regenerated signal into a reflectiveacoustic environment, using the natural acousticalproperties of an environment to achieve an evendistribution of the sound in the same way as a di-rect acoustic source behaves.Convention Paper 5722

Workshop 3 Saturday, March 22 13:30–16:30 hRoom L

THE FUTURE OF HIGH RESOLUTION AUDIO

Chair: Malcolm Hawksford, University of Essex, Essex, UK

Panelists: James Angus, University of Salford, Salford, UKTony Faulkner, Green Room Productions, Harefield, Middlesex, UKKevin Halverson, Muse Electronics, Garden Grove, CA, USAErwin Janssen, Philips Research, Eindhoven, The NetherlandsBob Stuart, Meridian Audio, UK

Demonstration Organization:Michael Page, Sony Broadcast and Professional Research Labs, Basingstoke, Hampshire, UKDerk Reefman, Philips Research, Eindhoven, The Netherlands

The technology driving DVD-audio and SACD continuesto evolve in harmony with an expanding awareness with-in both consumer and recording industries. This work-shop will present an update of these technologies span-ning both theoretical and practical issues together withthe so-far neglected relationship between high-resolutionaudio and Blue-Ray DVD. Controversial areas of relativeformat performance, software tools for recordable DVD,and digital interfacing in the consumer environment willbe discussed with a bias toward future evolution. Theworkshop will include a demonstration of streaming high-resolution multichannel audio in DSD format.

Seminar 3 Saturday, March 22 13:30–16:30 hRoom A

STEREO AND SURROUND MICROPHONE TECHNIQUES

Chair: Geoff Martin, Bang & Olufsen A/S, Struer, Denmark

Panelists: Jonathan Allen, Abbey Road Studios, UKFlorian Camerer, Östereichescher Rundfunk (ORF), Austria



Jean-Marie Geijsen, Polyhymnia International, The NetherlandsRon Streicher, Pacific AV Enterprises, CA, USA

This seminar is hosted by leading industry professionalsfrom the areas of classical and film music, as well as ra-dio drama. Issues to be discussed include the character-istics of various microphone configurations, managingstereo and multichannel recordings at the same session,and upward-and-downward compatibility considerations.This session will be of benefit to audio engineers of allbackgrounds, including students.

Exhibitor Seminar 1 Saturday, March 22 13:30–14:30 hRoom P

SYNERGY—AN INTEGRATED BROADCAST CENTER

Clyde BroadcastPresenter: Bas van der Steen

Synergy combines the functionality of a state-of-the-arthard-disc playout system with that of a digital broadcastmixer, suitable for a wide variety of applications. Synergymakes use of converging technology and represents twoyears of active research to identify important functionsand presents these in an elegant and ergonomic man-ner. At the heart of Synergy is a very powerful process-ing engine, based around three state-of-the-art SHARCdigital signal processing devices. As such, this repre-sents a huge amount of processing power, which thecommunications architecture supports, and furthers en-hances the integrated concept.


SURROUND SOUND METERING

DK Audio A/SPresenter: René Moerch

As the industry faces the challenge of mixing increasing-ly in surround sound, the audio engineers need to get agrip of monitoring surround sound signals as these aremuch more complex than in stereo recordings. It is im-portant that not only the level distribution is monitored,but also phase-relations, etc.

An MSD600M will be used to demonstrate the practi-cal use of a surround sound meter to improve standardsof mixing.

The presentation will cover the following topics: Level-meters; Loudness, including Leq(m); SMPTE versusAES standards; Phase-correlation meters; Goniometersand the Jelly-Fish surround meters.

Historical PresentationRECORD CUTTING HISTORY AND DEMOSSaturday, March 22, 15:00–16:00 h

Tim de Wolf will give an overview of record cutting tech-niques in use from back in the twenties until today.

After his presentation he will demonstrate on one ofhis vintage cutting machines how lacquers were cut inthe past. If there is sufficient interest demonstrations maybe repeated at a later time.

15:00 h Room 1Standards Committee Meeting on SC-05-02 AudioConnectors


mLAN: AUDIO AND MIDI OVER FIREWIRE

Yamaha CorporationPresenters: Richard Foss, Jun-ichi Fujimori

This presentation will cover FireWire for the audio andmusic industries, the basic components of FireWire, thetechnical fundamentals of mLan, and available products.FireWire, also known as IEEE-1394, is an advanced seri-al bus offering high-speed asynchronous and isochro-nous data transfer. As such it is ideally suited for real-time multimedia applications and has been widelyadopted, especially for consumer video applications.Yamaha has helped establish a protocol known as mLanfor carrying audio samples and MIDI data over FireWire.

We will discuss how products from companies such asYamaha, PreSonus, Apogee, Korg, and Otari can worktogether seamlessly over mLAN.

Student ActivitiesSTUDENT DELEGATE ASSEMBLY MEETING—PART 1Saturday, March 22, 16:30–18:00 hRoom H

The first Student Delegate Assembly (SDA) meeting isthe official opening of the convention’s student programand a great opportunity to meet with fellow students fromall corners of the world. In this session, which will bechaired by the SDA chair and vice-chair elected at lastyear’s European convention, the activities of the SDAand the student sections will be discussed and the stu-dent program for the convention is presented. Studentsand student sections will be given the opportunity to in-troduce themselves and their activities in order to stimu-late international contacts.

During this session nominations will be made for thenew Europe/International Regions SDA chair and vice-chair. The AES Regional Vice Presidents of the Euro-pean and International regions can each nominate twocandidates from their region. Election results andRecording Competition and Poster Awards will be givenat the Student Delegate Assembly Meeting, Part 2, onTuesday, March 25, at 13:30 h.

16:40 h Room E/F

Technical Council Meeting on Perception andSubjective Evaluation of Audio

16:40 h Room J

Technical Committee Meeting on High ResolutionAudio


MINIATURE MICROPHONES AND SPEECH INTELLIGIBILITY

DPA Microphones A/SPresenters: Eddy B. Brixen, Graig Parrish, Stephen Leth

Moeller

Miniature microphones are widely used in television pro-duction. A microphone may be placed on the persons’chest when used for ENG. In drama production the mi-crophone is often placed beneath clothing or in other hid-den places. The sound therefore suffers from a lack ofspeech intelligibility. Many field mixers do not have theright filtering facilities hence the sound has to be treatedlater—if time is available.


These issues have been studied and the result is DPA4071. The 4071 is acoustically pre-equalized, offering a5 dB presence boost to improve speech intelligibility anddefinition.

Exhibitor Seminar 5 Saturday, March 22 17:00–18:00 hRoom L

GENELEC “LAMINAR SPIRAL ENCLOSURE” (LSE™)TECHNOLOGY

Genelec OyPresenter: Christophe Anet

Genelec, a leader in active studio monitoring, presents atechnical insight into their groundbreaking LSE™ SeriesSubwoofers. A brief company introduction and productrange tour precedes an in-depth review of the principlesof the Laminar Spiral Enclosure design. Special attentionis given to the acoustical aspects, in particular how tur-bulence has been reduced in the vent to yield a loweroverall system distortion. Integrated into the cabinet is aproprietary 6.1 active bass manager, the features ofwhich are presented so that the delegates gain a com-plete understanding of how to accurately monitor mono,stereo, matrix, 5.1, and 6.1 surround sound mixes.

17:00 h Room 2Standards Committee Meeting on SC-04-02Characterization of Acoustical Materials

19:00 h Room 2Standards Committee Meeting on SC-03-12 ForensicAudio

Session E Sunday, March 23 09:00–12:00 hRoom C/D

LOUDSPEAKERS, PART 2

Chair: Juha Backman, Nokia Mobile Phones, Espoo, Finland

Co-chair: Richard Small, Harman/Becker Automotive Systems, Inc., Martinsville, IN, USA

09:00 h

E-1 Detection and Diagnoses of SubharmonicTones Generated in Woofers—John Stewart, Harman/Becker Automotive Systems, Inc., Martinsville, IN, USA

An audible anomaly was heard while evaluating aprototype woofer loudspeaker excited with sinu-soidal wave signals at frequencies near thewoofer’s main resonance. Standard testing did notreveal the nature or cause of this unusual sound.Further investigation showed the loudspeaker wasproducing a frequency component at half the inputfrequency; a subharmonic. Reported cases of sub-harmonic production in woofers have been rare.Most reports have studied subharmonic occur-rences due to diaphragm instability. In this paperthe techniques used to confirm the generation of asubharmonic in this woofer are presented. Themechanisms that combined to generate the sub-harmonic are shown. Further examples found inother woofers that exhibit this sound anomaly onoccasion are also discussed.Convention Paper 5741

09:30 h

E-2 Direct-Radiator Loudspeaker Systems with HighBl—John Vanderkooy1, Paul M. Boers2, Ronald M.Aarts21University of Waterloo, Waterloo, Ontario, Canada 2Philips Research Labs, Eindhoven, The Netherlands

This paper is an extension of AES 113th Paper#5651 which shows additional consequences of adramatic increase in the motor strength Bl of a dri-ver. Not only is the efficiency of the loudspeakerand amplifier greatly increased, but high Bl-valueshave a positive influence on other aspects of loud-speaker systems. Box volume can be significantlyreduced and other parameters can be altered. Aprototype driver unit is studied which performs wellin a small sealed box. Vented or passive radiatorsystems do not benefit as much from high Bl.Convention Paper 5742

10:00 h

E-3 Finite Element Methods and Equivalent Electri-cal Models for Loudspeaker Characterization—Guillaume Pellerin, Jean-Dominique Polack, Jean-Pierre Morkerken, Laboratoire d’AcoustiqueMusicale, Paris, France

Current research into electroacoustics tends to de-termine the global transfer function between an ini-tial electrical signal and the acoustical signal trans-mitted to the ear. Because electrodynamictransducers radiate in a large frequency bandwidth,lumped parameter models such as Thiele andSmall’s are not sufficient to provide a realistic simu-lation of the vibroacoustical behavior of the system.This paper proposes the use of finite element andboundary element methods to compute a complex3-D response of a loudspeaker for each mechanicalmode and then synthesize an equivalent electricalmodel that takes into account acoustical couplingbetween each pair of modes.Convention Paper 5743

10:30 h

E-4 Product Safety—End of Audio Fun?—Erhard E.Werner, Hademstorf, Germany

Protection against hearing impairment is as well anindividual as a general community matter. Regula-tions for occupational noise have existed for a longtime. Proposals for similar protective means for theleisure range resulted in extreme contradictory ar-guments. The contribution is dealing with an Euro-pean attempt to find a compromise between thenatural wish for unlimited individual acoustic funand fundamental consequences of living within acommunity offering social health care. Details of EN50332 and consequences for technical details ofportable audio equipment are presented with refer-ence to the product safety directive.Convention Paper 5744

11:00 h

E-5 Analysis of a Folded Horn—Andrew Bright, NokiaGroup, Helsinki, Finland; Technical University ofDenmark, Lyngby, Denmark

A boundary element model is used to analyze afolded horn. Results from the boundary elementmodel are compared to measurements of the throat



radiation impedance and the far-field acoustic re-sponse. Further analysis shows how one-dimen-sional and lumped parameter models can be de-rived from the boundary element results, and usedto gain insight into the behavior of the folded hornloudspeaker system. It is shown that one type offolded horn behaves more like a vented-box than atraditional horn-loaded loudspeaker system.Convention Paper 5745

11:30 h

E-6 The Full-Sphere Sound Field of ConstantBeamwidth Transducer (CBT) Loudspeaker LineArrays—D. B. (Don) Keele Jr., Harman/Becker Automotive Systems, Martinsville, IN, USA

The full-sphere sound radiation pattern of the CBTcircular-wedge curved-line loudspeaker array ex-hibits a 3-D petal- or eye-shaped sound radiationpattern that stays surprisingly uniform with frequen-cy. Oriented vertically, it not only exhibits the ex-pected uniform control of vertical coverage but alsoprovides significant coverage control horizontally.The horizontal control is provided by a vertical cov-erage that smoothly decreases as a function of thehorizontal off-axis angle and reaches a minimum atright angles to the primary listening axis. This is incontrast to a straight-line array that exhibits a 3-Dsound field that is axially symmetric about its verti-cal axis and exhibits only minimal directivity in thehorizontal plane due to the inherent directionalcharacteristics of each of the sources that make upthe array.Convention Paper 5746

Session F Sunday March 23 09:00–12:00 hRoom E/F

AUTOMOTIVE AUDIO AND INSTRUMENTATION ANDMEASUREMENT

Chair: Richard S. Stroud, Stroud Audio, Inc., Kokomo, IN, USA

09:00 h

F-1 Common-Mode to Differential-Mode Conversionin Shielded Twisted-Pair Cables (Shield-Current-Induced Noise)—Jim Brown1, Bill Whitlock21Audio Systems Group, Inc., Chicago, IL, USA2Jensen Transformers, Inc., Van Nuys, CA, USA

Neil Muncy has shown that audio frequency currentflowing on the shield of twisted-pair audio wiring willbe converted to differential mode voltage by any im-balance in the transfer impedance of the cable. Hehypothesized that the effect is magnified by thepresence of a drain wire and increases linearly withfrequency. Whitlock and others have shown thatconversion also occurs with capacitive imbalance.This paper confirms Muncy’s hypotheses andshows that shield-current-induced noise is signifi-cant well into the MHz range.Convention Paper 5747

09:30 h

F-2 New Vacuum Tube and Output TransformerModels Applied to the Quad II Valve Amplifier—Menno van der Veen1, Pierre Touzelet21Ir Buro Vanderveen, Zwolle, The Netherlands2Sereme, Vélizy, France

In earlier papers new vacuum tube and outputtransformer equivalent models were proposed.These models are now applied to the famous QuadII valve amplifier. Results of models and measure-ments are compared in the frequency-time and am-plitude domains. It is demonstrated that outputtransformers and multigrid vacuum tubes and com-plete vacuum tube amplifiers can be modeled withgreat precision triodes and also to multigrid vacuumtubes, like tetrodes and pentodes.Convention Paper 5748

10:00 h

F-3 System Measurement and Modeling UsingPseudo-Random Filtered Noise and Music Sequences—Malcolm Hawksford, University of Essex, UK

System measurement employing pseudo-randomfiltered noise and music sequences is investigated.An efficient single-pass technique is used to evalu-ate simultaneously transfer function and spectraldomain signal-to-distortion ratio that is applicable toamplifiers, signal processors, digital-to-analog con-verters, perceptual coder performance, and loud-speakers. The technique is extended to determinea power-series model from which nonlinear distor-tion can be estimated for an arbitrary excitationwithout need of remeasurement.Convention Paper 5749

10:30 h

F-4 Improvements of a Horn-Loaded OmnidirectionalSound Source—Christos Goussios, GeorgeKalliris, Charalampos Dimoulas, George Papaniko-laou, Stylianos-Marinos Charalampidis, AristotleUniversity of Thessaloniki, Thessaloniki, Greece

New techniques for the improvement of the fre-quency response, radiation patterns, and maximumsound pressure level of a horn-loaded omnidirec-tional point source are presented. In order to im-prove the response in the region of low frequencies,a ported enclosure system was designed, con-structed, and evaluated. Cone-shaped reflectorswere placed inside the horns to improve the polarpatterns at higher frequencies. Design procedures,measurements, and results are explained. Convention Paper 5750

11:00 h

F-5 The Development of Disc Cutting Heads—SeanDavies, S. W. Davies Ltd., Aylesbury, UK

Proper restoration of historic recordings is assistedby an accurate knowledge of the characteristics ofthe equipment used in the recording process. Thispaper traces the evolution of electromechanical disccutting heads from the early 1920s to the latest mod-els. Use is made of the electrical-mechanical equiva-lent circuits to analyze the performance of movingiron, moving coil, and feedback controlled types.Convention paper 5751

11:30 h

F-6 Automotive Doors as Loudspeaker Enclosures—Roger Shively, Josh King, Harman/Becker Auto-motive Systems, Martinsville, IN, USA

The results of a study of the quality of automotivedoors as loudspeaker enclosures are presented.


Electrical and acoustical measurements of severalautomotive doors are made and compared to linearbox theory. Judgment criteria are considered forquantifying the doors as enclosures and relatingthat to the sound quality using roughness as one ofthe criterion.[Paper not presented at convention, but ConventionPaper 5752 is available.]

Workshop 4 Sunday, March 23 09:00–12:00 hRoom A

MULTICHANNEL/SURROUND SOUND: A Chance for Enhanced Creativity

Chair: Martha de Francisco

Panelists: Florian Camerer, Tonmeister, ORF Austrian Broadcasting Corporation, Vienna, AustriaFriedemann Engelbrecht, Tonmeister/producer Teldex Studio Berlin, Berlin, GermanyJean-Marie Geijsen, Recording Engineer Polyhymnia International, Baarn, The NetherlandsSimon Heyworth, Recording/Mastering Engineer, Superaudio Mastering, UKJames Mallinson, Independent Producer, London, UKRonald Prent, Resident Engineer, Galaxy Studios, Mol, Belgium

With multichannel/surround sound recordings we are en-tering a new era in the transmission of music. This callsfor an increased creativity and new ideas to be applied tomusic recording.

How does the impact of a multichannel recording onthe listener compare to that of the same recording instereo? When recording in surround sound, do we needto follow the spatial rules imposed by the live-concertperformance of classical music and jazz? Are we facinga possible revolution in the recording of pop music?Does the level of technical recording quality have a sig-nificant influence on the way in which the listener per-ceives the performance?

With the help of listening excerpts of a variety of multi-channel recordings, a panel of leading engineers will dis-cuss issues on sound aesthetics and practical aspects ofsurround sound recording.

Seminar 4 Sunday, March 23 09:00–11:00 hRoom B

BASICS OF ROOM ACOUSTICS

Presenter: Diemer de Vries, Delft University of Technology, Delft, The Netherlands

In the chain recording-mixing/mastering-reproduction, theacoustic properties of thee spaces play a role, First, thereis the recording space, which can vary from open air viastudios to cathedrals, all with their specific acoustical con-ditions. Then, there is the control room, often much small-er than the recording space, the acoustics of which shouldnot obstruct a proper assessment of the recording. Finally,there is the reproduction space, often a living room butsometimes a larger space like a cinema. This seminar willexplain how acoustic properties of spaces can be speci-fied by calculations or measurements—acoustics is notblack magic at all!—and how these objective parameters relate to what we hear. Using these parame-ters, we will take a closer look at the spaces mentionedabove and see how we can use our insight in roomacoustics to the benefit of audio practice.

09:00 h Room 2Standards Committee Meeting on SC-06-06 AudioMetadata

Exhibitor Seminar 6 Sunday, March 23 10:30–11:30 hRoom P

DIGASYSTEM IMPLEMENTS THE IDEA OFCONVERGENCE IN NEWSROOM ENVIRONMENTAND JOINT AV CONTENT MANAGEMENT

D.A.V.I.D. GmbHPresenters: Armin Woods, Gerhard Moeller

DigaSystem, the well-introduced tool for radio broadcastand audio applications, is setting sail to transfer its compe-tence to the video/television world. Seeing the needs in modern media content management, regarding theworldwide move toward intermedial integration of audio, textand video content, D.A.V.I.D. is extending the DigaSystem Framework for television applications for ingest, low/high resolution browsing and editing, format con-version, and exchange. Due to the fact that convergence isfar more noticeable in production than in delivery/playout,and D.A.V.I.D. is operating together with a strong grouppartner, we can already show highly interesting applicationsin use at various prestigious customer sites.

Exhibitor Seminar 7 Sunday, March 23 10:30–11:30 hRoom L

QUALITY AUDIO CODING—HOW TO PRESERVEAUDIO QUALITY, SPECIFICALLY RELATED TO DABAND DIGITAL NETWORKS

APT—Audio Processing TechnologyPresenter: Patrick McGrath

With the on-going promotion and deployment of DABnetworks, broadcasters have to ensure that the invest-ment reflects the feature set of digital radio. One specificconcern is audio quality and how to ensure that DAB im-proves upon existing analog services. A major impact onaudio quality in DAB services is multiple MPEG passes.APT will suggest a few simple rules to maintain audioquality.

APT will also highlight the apt-X™ 4:1 compression al-gorithm, where apt-X™ should be used in the audiochain, how it can be implemented, and focus on a num-ber of existing third part companies which offer apt-X™in their product portfolio.

10:30 h Room 1Standards Committee Meeting on SC-02-01 DigitalAudio Measurement Techniques

Session U Sunday, March 23 11:00–13:00 hTopaz Lounge

(POSTERS) ACOUSTICAL AND PERCEPTUAL MODELS

11:00 h

U-1 Parameterizing Human Pinna Shape for the Esti-mation of Head-Related Transfer Functions—Carl Hetherington, Anthony I. Tew, University ofYork, York, UK

This paper describes a method for parameterizingtwo-dimensional cross-sectional contours of the hu-man head and pinnae using the elliptic Fouriertransform. We demonstrate how contours may be


reconstructed from a limited subset of their parame-ters and explore the shape errors that result fromsuch reconstructions. We present the results ofacoustic simulations of contours reconstructed fromparameter sets of varying size in order to explorethe effects of parameter set truncation on pressureresponse. Extensions of the parameterization tothee dimensions are discussed and we suggest howthe parameterization may in the future be integratedwith existing techniques in order to estimate HTFsfrom shape data in an efficient way.Convention Paper 5753

11:00 h

U-2 An Independent Component Analysis Approachto Automatic Music Transcription—Samer A. Abdallah, Mark D. Plumbley, Queen Mary, Universi-ty of London, London, UK

We used Independent Component Analysis (ICA)with sparse coding to analyze music spectral se-quences. We modeled an audio spectrum as an ap-proximate mixture of the spectra of individual notes,using our ICA approach to “unmix” this to find the in-dividual notes and note spectra. Notes are assumedto be approximately independent, and sparse (mostlyoff). Results on synthesized harpsichord music areencouraging, producing an approximate piano-rolltranscription and a passable rendition of the originalmusic when resynthesized. We are currently workingto extend and improve this though the use of tempo-ral information of note activities and to handle morecomplex timbral behavior.Convention Paper 5754

11:00 h

U-3 Efficient Model Performing a Multilevel Struc-ture of Auditory Information Applied to AudioCoding—Enrique Alexandre, Antonio Pena, Universidad de Vigo, Spain

In this paper an efficient model which implements amultilevel structure of auditory information is pre-sented. An exhaustive analysis of the input signal inboth subjective and objective terms is performedwhich makes an ad-hoc coding depending on theparticular characteristics of the input signal. Themodel comprises not only the calculation of themasking threshold but also several techniques andtools designed to reduce the amount of audible arti-facts present in the coded signal.Convention Paper 5755

11:00 h

U-4 BlockCompiler: Efficient Simulation of Acousticand Audio Systems—Matti Karjalainen, HelsinkiUniversity of Technology, Espoo, Finland

An experimental software environment called theBlockCompiler is described that has been devel-oped for flexible, yet efficient, simulation of differentacoustic and audio systems. It is based on compu-tational block objects and their interconnection net-works, and it supports several different modelingparadigms. It is particularly powerful in creatingphysical models where two-directional interactionbetween physical elements has to be represented.High-level model specifications are compiled to effi-cient code, supporting real-time simulation andsound synthesis of relatively complex systems.Simulation examples described include modeling of

musical instruments, speech synthesis, and multidi-mensional acoustical structures. Further cases fitting to the scheme, such as loudspeaker simula-tion, are discussed briefly.Convention Paper 5756

11:00 h

U-5 An Analytical Modeling to Describe the Couplingbetween a Piezoelectric Actuator and a LoadingMedium. Validation of the Method for EngineeringProblems—Pierrick Lotton1, Bertrand Lihoreau1,Michel Bruneau1, Vitali Gusev21Laboratoire d’Acoustique de l’Université du Maine, Le Mans, France

2Laboratoire de Physique de l’Etat Condensé, Le Mans, France

Flexural-mode piezoelectric transducers are exten-sively used as acoustic actuators. The usefulnessof equivalent circuit modeling to characterize thiskind of piezoelectric source has been long recog-nized, even the conventional models are empiricalor even they assume drastic approximations. Theaim of the paper is to improve this kind of equiva-lent network in such a way that it permits linking together the basic physical parameters, avoidingoverly intricate formulation even when it does notassume usual approximations. The equivalent net-work obtained shows a classical structure, the ex-pressions of the parameters of the circuit beinghowever known analytically as simple functions ofthe parameters of the system (expressions primarilygiven in a previous work by the authors). In order tovalid the modeling, an application is given whenloading the piezoelectric source with a resonator.Convention Paper 5757

11:00 h

U-6 Simulation Tools in Electroacoustic Transduc-ers. A Case Study: Different Order Band-PassSystem Design—Juan José Gómez-Alfageme,Beatriz Sánchez-Alonso, Universidad Politécnicade Madrid, Madrid, Spain

The teaching of electroacoustics for an audio engi-neering career, and especially the characterizationof electroacoustic transducers, has always beencomplex from the point of view of the traditional ed-ucational means. The employment of the simulationtools has allowed us to develop applications for thestudy of these transducers based on the employ-ment of equivalent circuits and its analysis both intime and frequency domains. In this paper some ofthese applications are described in the case of de-signing low frequency band-pass radiation systemswith variable geometry, using Mathcad software asa simulation tool. This tool is a simple to programand it allows a great quantity of calculation andgraphic representation possibilities.Convention Paper 5758

11:00 h

U-7 One New Method for Calculating ReverberationTime in a Car Compartment—Fangli Ning1, JuanWei21Northwestern Polytechnical University, Xi’an, Shaanxi, China

2Xi’an Jiao Tong University, Xi’an, Shaanxi, China

In this paper, based on the acoustical finite elementmethod (FEM) model, one new method for calculat-



of the material but also planning and scheduling, mediaasset management as well as different reporting tasks.Just as the companies have got their digital islandsready to operate, they need to be integrated into enter-prise-level systems. This move requires new approachesto system specification, integration of various systems,system level interfaces, workflow, and system mainte-nance procedures. Not to mention the increasing de-mand on high-level user rights management.


WORLDNET SKYLINK—5.1 VIA TCP/IP FOR REMOTEDIRECTORS’ APPROVALS

APT—Audio Processing TechnologyPresenter: Noel McKenna

The 5.1 TCP/IP WorldNet Skylink is the result of collabo-rative works between APT—Audio Processing Technolo-gy and the audio experts at George Lucas’s SkyWalkerRanch. The WorldNet SkyLink can stream 5.1, 6.1 or 7.1multichannel surround sound with SMPTE Time code at16-, 20- or 24-bit word resolution. The WorldNet SkyLinkhas been designed for film directors to remotely approveon-going 5.1 mix work undertaken at a production facility,with particular attention paid on security and encryptionof the content. The WorldNet SkyLink can deliver contentin real time or Store n` Forward depending on the datarates available.

12:10 h Room C/DTechnical Committee Meeting on Loudspeakers andHeadphones

12:10 h Room E/FTechnical Committee Meeting on Audio Recordingand Storage Systems

12:30 h Room 1Standards Committee Meeting on SC-03-04 Storageand Handling of Media

Session G Sunday, March 23 13:30–16:30 hRoom C/D

ROOM ACOUSTICS AND SOUND REINFORCEMENT

Chair: Peter Swarte, P. A. S. Electro-Acoustics, Eindhoven, The Netherlands

13:30 h

G-1 Room Mode Bass Absorption though Combined Diaphragmatic and Helmholtz Resonance Techniques: “The Springzorber”—Evan Reiley,Anthony Grimani, Performance Media IndustriesLtd., Fairfax, CA, USA

This paper documents the research and prototypingof a new form of acoustical bass absorber. Thebass absorber reduces peak and dip frequency re-sponse errors caused by interference from naturallyoccurring axial standing waves in rectangularrooms. The design uses two forms of simple har-monic resonance: pistonic diaphragm resonanceand Helmholtz cavity resonance. The pistonic di-aphragm resonance is achieved by attaching a rigidplanar membrane to metal springs. The Helmholtzcavity resonance is achieved by constructing an en-closed chamber attached to an open cylindrical

ing reverberation time in a car compartment is pre-sented. The new method differs from the Sabineequation in that the diffuse sound field is not ab-solutely necessarily as a condition. As a result, thenew method could give correct reverberation timefor a designed car compartment with any shapeand boundary condition. First, this paper describesthe FEM of a car compartment. Second, the newmethod is presented in detail, and computer pro-gram of it is written. Last, in one model of a carcompartment, the reverberation time is calculatedwith the new method and compared with that givenby experiment; the results show that the method iseffective and feasible.[Paper not presented at convention, but ConventionPaper 5759 is available.]


GROUNDING AND SHIELDING

Chair: Tony Waldron, UK

Panelists: Keith Armstrong, Cherry Clough Consultants, UKJim Brown, Audio Systems Group, UKRick Chinn, USAIan McBurney, Allen & Heath, UKBill Whitlock, Jensen Transformers, USA

The design of high performance professional audio sys-tems was never trivial, but since the development of mi-croprocessor controlled equipment, switch-mode powerconversion, and the expansion of VHF/UHF wirelesspublic broadcasting and cellular telephones, interferenceproblems have escalated by several orders of magni-tude. Fully-shielded wiring between audio components(direct shield bonding at both ends) must now be recon-sidered. We will demonstrate that when current flows inthe shield conductors of audio interconnection cables,hum and/or buzz interference need not be the naturaloutcome.

Historical PresentationOLD TIME MEASURING EQUIPMENTSunday, March 23 11:00–12:00 h

Wim van Barneveld will demonstrate how audio para-meters were measured, using the very first Bruel & Kjaeraudio measuring equipment. A collection of B&K equip-ment will be on show.

11:30 h Room 2Standards Committee Meeting on SC-06-01 Audio-File Transfer and Exchange


CHALLENGES AND SOLUTIONS FOR ENTERPRISE-WIDE RADIO INTEGRATION

Jutel OyPresenter: Jorma Kivelä

Digital audio in radio broadcasting started as a replace-ment of analog devices. Systems were easy to installand configure. Typical issues to be handled in installationprojects were, for example, device noise, heat, wiring,signal levels, and clock distribution. It was easy to adoptexisting system design and maintenance culture, as nomajor changes were needed.

Today the digital systems in broadcasting companiessupport not only the digital production and broadcasting


tube. Coupling these two dissipation devices led toseveral-fold improvement in absorption and totalroom mode attenuation.Convention Paper 5760

14:00 h

G-2 Room Acoustics for Rehearsals and Concerts—The New Festhalle in Landau, Germany—Ernst-Joachim Völker, Wolfgang Teuber, Institute forAcoustics and Building Physics, Oberursel andZweihausen, Germany

The acoustics of concert halls are still a secret. Theymust fulfill many desires and sometimes contrary re-quirements. Nowadays, a concert hall must be a mul-tipurpose hall, suitable for concerts and rehearsalswith and without an audience, for television and radiobroadcasting, for shows and theater with or withoutelectro-acoustical amplification. Usually chairs can beremoved to turn it into a dance floor or a sports arena.The Festhalle in Landau had to reach this goal andhas fulfilled the expectations. But what is the historyof this hall? In 1952 with an audience of 2000, Wil-helm Furtwängler conducted the Berlin Philharmonicand was impressed by the excellent acoustics. Sixyears later a total change from the former Jugendstilinto a modern concert hall “of the 50s” was carriedout. Why? The reason was an another age of acousti-cal knowledge. Today, a reconsciousness takesplace. No wonder that the original building of 1908with its beautiful Jugenstil-architecture should be re-discovered. For almost 100 years many differentevents, architectural properties, and tastes had to beconsidered including the protection of a historicalmonument. The knowledge of acoustics has growntremendously over the past decades. For instance,the “precise structure of sound” or the “distance of pri-vacy” require consequences in acoustical design forboth musicians on the stage or in the pit and for thelistener in the audience. The remodeling of the Fes-thalle started in 1997 and was finished with the open-ing ceremony on January 12, 2002 with an audienceof 1041. During the concert, however, the acousticalwork had to be continued with a recording in order tomeasure the reverberation time with an audience forfinal changes. The paper summarizes the decisionsmade, the progress of the work, and final results.Some comparisons with other halls and requirementsare made.Convention Paper 5761

14:30 h

G-3 What You Specify Is What You Get (Part 1)—Johan van der Werff, Dick de Leeuw, Peutz & Associés, Mook, The Netherlands

The Peutz prediction algorithms for speech intelligi-bility as published in 1971 in the J.A.E.S. Vol. 19,No. 11, are still valid and are found remarkably ac-curate considering their simplicity. However, somerevision is necessary for adaptation to the contem-porary room simulation and sound system designprograms. This paper deals with the prediction ofthe Articulation Loss of consonants (ALcons) basedon data usually available in the design phase of aproject. Special attention is given on how to dealwith multiple sources, nearer and further apart. Forthe attendees to the presentation of this paper therewill be an Excel® spreadsheet available for a quickcalculation according to the proposed method.Convention Paper 5762

15:00 h

G-4 What You Specify Is What You Get (Part 2)—Johan van der Werff, Rob Metkemeijer, Peutz & Associés, Mook, The Netherlands

The Peutz prediction algorithms for the ArticulationLoss of consonants (ALcons) as published in 1988(85th Convention in Los Angeles) did not seem toget the attention they deserved in the acoustical so-ciety. Perhaps this is due to the confusion it mayhave stirred because of the totally different set of al-gorithms compared to the 1971 set, or perhaps dueto the more complicated calculations. But most like-ly it could be due to how and where to get the phys-ical quantities needed for input. This paper will dealwith the underlying principles: how to extract thedata from an impulse response and how to calcu-late the ALcons from that. It is thought that this willbe a valuable addition to the well-known STI mea-surements. The data can be narrow band (one oc-tave wide) and is in the gathering not sensitive forsignal processors in the signal chain or for the typeof filters used in the postprocessing of the data. Forthose attending this paper presentation, there willbe a computer program available which reads a setof measured or calculated impulse responses, ex-tracts the data, calculates the ALcons, and presentsthe results.Convention Paper 5763

15:30 h

G-5 User Interaction and Authoring of 3D SoundScenes in the Carrouso EU Project—RiittaVäänänen, IRCAM, Paris, France

The Carrouso project combines technologies forrecording, transmission, and rendering of 3-Dsound scenes. The rendered virtual scene includesboth the sound content and the spatial and roomacoustic description of the performance space.MPEG-4 tools are utilized for encoding this data,using the general audio coding for compression ofsound streams and the scene description tools forcreating virtual 3-D audio scenes. We describe thecreation of the virtual acoustic space in Carrouso,carried out with the help of room acoustics analysissoftware and an authoring tool. A visual representa-tion of the virtual sound scene is also transmitted tothe renderer, and it acts as a user interface allowingrenderer-side scene modification via the interactionmechanisms provided in MPEG-4.Convention Paper 5764

16:00 h

G-6 The Acoustic and Intelligibility Performance of Assistive Listening and Deaf Aid Loop (AFILS)Systems—Peter Mapp, Peter Mapp Associates,Colchester, UK

Although the electronic performance of inductiveloop and other hard-of-hearing assistive audio sys-tems is well covered by appropriate criteria andcodes of practice, little or no attention appears tohave been given to the acoustic performance andintelligibility requirements of such systems. This pa-per reports on the results of some trial acoustic per-formance testing of these systems. In particular theeffects of system microphone type, distance, andlocation are shown to have a significant effect onthe resultant performance. Test procedures and therequirements for a suitable acoustic test source to



mimic a human talker are also highlighted. It isshown that simply using a head-sized loudspeakerdoes not replicate the sound field of a human talkerand can lead to significant intelligibility measure-ment errors.Convention Paper 5765

Session H Sunday, March 23 13:30–16:30 hRoom E/F

SPATIAL PERCEPTION AND PROCESSING, PART 1


13:30 h

H-1 Study of Sound Source Shape and Wideness inVirtual and Real Auditory Displays—GuillaumePotard1, Jens Spille21University of Wollongong, New South Wales, Australia

2Deitsche Thomson-Brandt GmbH, Hannover, Germany

AudioBIFS, the scene description language ofMPEG-4, can currently describe and present pointsound sources such as a flying insect or a distantsound source, but i t cannot describe soundsources that have certain spatial dimensions suchas a choir, a seashore or rain. Spatial widenessor tonal volume is, however, an important aspectof sound sources in order to improve realism ofvirtual sound scenes. We proposed in July 2002to add sound source wideness and shape to Au-dioBIFS. Subjective experiments then followed.The experiments showed that vertical and hori-zontal wideness could be clearly perceived, how-ever the auditory shapes of sound sources weremore difficult to detect. Therefore the widenessfeature only should be implemented in MPEG.Convention Paper 5766

14:00 h

H-2 A New Approach to Detecting Auditory OnsetsWithin a Binaural Stream—Ben Supper, TimBrookes, Francis Rumsey, University of Surrey,Guildford, UK

The human auditory system is particularly sensitiveto spatial information conveyed in the first two mil-liseconds of an auditory event. Therefore, in order toanalyze a stream of binaural data in a perceptuallyrelevant way, it is important to determine quickly andprecisely the onset of each event within a datastream. This paper details the design of an auditoryonset detector which is intended to assist in the ex-traction of spatial parameters from an arbitrary bin-aural stream. A fast predictive filter architecture isspecified and tested using binaural recordings.These items highlight the strengths, limitations, anddifficulties of computerized onset detection and ofthis approach in particular.Convention Paper 5767

14:30 h

H-3 Techniques and Applications of Wearable Augmented Reality Audio—Aki Härmä, Julia Jakka, Miikka Tikander, Matti Karjalainen, TapioLokki, Heli Nironen, Sampo Vesa, Helsinki University of Technology, Espoo, Finland

The concept of augmented reality audio characterizestechniques where real sound environment is extend-ed with virtual auditory environments and communi-cations scenarios. This paper introduces a frameworkfor wearable augmented reality audio (WARA) basedon a specific headset configuration and a real-timeaudio software system. We will review relevant litera-ture and aim at identifying most potential applicationscenarios for WARA. Listening test results with a pro-totype system will be presented.Convention Paper 5768

15:00 h

H-4 Unidimensional Simulation of the Spatial Attribute “Ensemble Width” for Training Purposes —Tobias Neher1, Francis Rumsey1, Tim Brookes1, Peter Craven21University of Surrey, Guildford, Surrey, UK2Algol Applications, Steyning, West Sussex, UK

This paper reports recent progress toward the de-velopment of a spatial ear trainer. A study into theperceptual construct of “ensemble width” was con-ducted. With the help of a novel surround panner,exemplary stimuli were created. Changes werehighly controlled to enable unidimensional variationof the intended qualitative effect. To assess thesuccess of the simulation, a subjective experimentwas designed based on multidimensional scaling(MDS) techniques and completed by an experi-enced listening panel. Additional verbal and nonver-bal data were collected so as to facilitate analysis ofthe perceptual (MDS) space. Results show that uni-dimensionality was achieved, thus suggesting thestimuli to be suitable for training purposes.Convention Paper 5769

15:30 h

H-5 Algorithms for Moving Sound Images—MitsuoMatsumoto, Mikio Tohyama, Kogakuin University,Hachioji, Japan

A previously introduced algorithm for simulating amoving sound image was evaluated objectively andsubjectively. This algorithm used time-variant con-volution with a method for interpolating binaural im-pulse responses that considers the arrival timeswhen interpolating responses. Thee moving soundimages were used: an actual one recorded using arotating dummy head, one simulated using the con-ventional cross-fading method, and one simulatedusing the algorithm. The objective comparison byspectrograms of the thee images and the subjectiveevaluation showed that the one simulated using thealgorithm was quite close to the actual one.[Paper not presented at convention, but ConventionPaper 5770 is available.]

16:00 h

H-6 Creation and Verification of a Controlled Experi-mental Stimulus for Investigating Selected Perceived Spatial Attributes—Russell Mason,Tim Brookes, Francis Rumsey, University of Surrey,Guildford, Surrey, UK

In order to undertake controlled investigations intoperceptual effects that relate to the interaural cross-correlation coefficient, experiment stimuli that meeta tight set of criteria are required. The requirementsof each stimulus are that it is narrow band, normallyhas a constant cross-correlation coefficient over


time, and can be altered to cover the full range ofvalues of cross-correlation coefficient, includingspecified variations over time if required. Stimulicreated using a technique based on amplitudemodulation are found to meet these criteria, andtheir use in a number of subjective experiments isdescribed. [Paper not presented at convention, but ConventionPaper 5771 is available.]


NEW TECHNOLOGICAL DEVELOPMENTS IN MPEG-4 AUDIO

Chair: Jürgen Herre, Fraunhofer-IIS, Ilmenau, Germany

Panelists: Martin Dietz, Coding Technologies, Nuremberg, GermanyErik Schuijers, Philips Digital Systems Laboratories, Eindhoven, The NetherlandsSchuyler Quackenbush, Audio Research Labs, USA

For more than one decade, MPEG audio standards havebeen defining the state of the art in perceptual coding ofaudio signals. Several phases of standardization(MPEG-1, MPEG-2, MPEG-4) have been pushing bor-ders beyond everybody’s initial expectations. This work-shop reports on thee new extensions to the existingMPEG-4 audio standard which are currently under devel-opment. The work on compatible bandwidth extension ofaudio signals provides an additional performance boostfor AAC coding at very low bit rates, permitting consider-able signal quality at bit rates around 24kbit/s per chan-nel. A second extension augments the existing MPEGparametric audio coder with modes for representinghigh-quality audio signals. Finally, methods for losslessaudio coding are under consideration, extending currentMPEG-4 audio coders toward perfect representation atword lengths and sampling rates typically associatedwith high-definition audio. The workshop will providebackground and demonstrations on these technologies.


MICROPHONE TECHNIQUE THEORY FOR STEREOAND SURROUND

Chair: Geoff Martin, Bang & Olufsen A/S, Struer, Denmark

This seminar will present attendees with the theoretical response characteristics of various microphone configu-rations. In addition, a general understanding of the ef-fects of changes in configurations and microphone selec-tion on the reproduced sound stage will be provided.Simple mathematical analyses of various configurationswill be used to clarify how coincident and spaced micro-phone pairs generate a desired output. Various textbookconfigurations will be discussed, as well as the effects ofreal-world factors such as off-axis frequency responseand early reflections on image position and quality. Thetheoretical aspects will be illustrated with thee-dimen-sional visual plots made available to all attendees.


FFT PROCESSING PROCEDURE FOR VOICE, AUDIO,AND ON AIR SIGNAL PROCESSING

IDT/J&C Intersonic AGPresenters: Marc Straehl, Peter JossProduct: DVP, DEP, VVP

Today’s on-air processing is reviewed in its entirety, fromthe moment of audio signal generation up to the last sig-nal treatment before the transmitter feed to an FM or AMtransmitter. Compared with today’s band-orientated pro-cessing methods FFT is allowing for a more detailed au-dio processing method, thus dividing audio processinginto steps such as voice processing, audio processing,and limiting (0dBr MPX power management). This se-quence of signal treatments leading to a much improvedaudio performance especially toward the compulsoryMPX power recommendations of Norm ERC-5405. Fur-ther, this method will help to avoiding subjective levelchanges between the energy content of compressed mu-sic and the commentators voice.


SUPER AUDIO CD PRODUCTION EXPLAINED

Super Audio CDChair: Bob Charlton, Scribe PRPresenters: Erdo Groot, Polyhymnia International

Ronald Prent, Galaxy StudiosAndreas Neubronner, Tritonus Musikproduction Crispin Murray, Metropolis Mastering, Philips EngineerDavid Walstra, Sony Europe, Bastiaan Kuijt, Sony Europe

Once the decision to make a Super Audio CD is madethe question is: how to continue? The goal of this semi-nar is to give clear insight in the production chain thatleads to the release of a Super Audio CD.

Engineers from the field will approach the path to a(surround) master from three different angles: pure DSD;multitrack for SACD; or bridging high resolution PCM toSACD.

A brief tutorial on the steps from mix to masters forproduction, including practical issues on the way, will befollowed by the story of the mastering of a milestone Su-per Audio CD production.

13:30 h Room 2Standards Committee Meeting on SC-04-03Loudspeaker Modeling and Measurement

Historical PresentationCLASSICAL PRODUCER’S VIEW ON VINTAGE AUDIO RECORDINGSunday, March 23 14:00–15:00 h

Willem Hellweg, a producer combining musical talentand profound knowledge of recording techniques, willgive his view on recording methods used for the classicalPhilips label between 1960 and 1990.


LOW DELAY TRANSFER OF AUDIO SIGNALSTHROUGHOUT TELECOM SYSTEMS

AETA/J&C Intersonic AGPresenters: Marc Straehl / Peter JossProducts: Hifiscoop 5AS, Scoopy



Compared to the internationally standardized MPEG-based audio signal transfer by means of applying LayerII, III or AAC, there are straight forward ways of coveringthe needs for radio broadcasting requirements based onthe 4SB algorithms, which covers a bandwidth of up to15 or 20 kHz within a transfer time of <10 ms from pointto point. This saves additional circuits, as n-1, and at thesame time brings back live aspects in on-air applicationsfor benefit of reporters, journalists, round table discus-sions, and many more applications.

15:00 h Room 1Standards Committee Meeting on SC-03Subcommittee on the Preservation and Restorationof Audio Recording


LAN DELIVERY OF AUDIO AND APPLICATION TODIGITAL LIBRARIES

Chair: Jeremy Cooperstock, McGill University, Quebec, Canada

Panelists: Peter Alyea, Library of Congress, Washington, DC, USAJ. Steven Downie, University of Illinois, USAJohn Grant, Nine Tiles, Cambridge, UKGeorge Massenburg, George Massenburg Labs, Franklin, TN, USAMike Overlin, Yamaha, Buena Park, CA, USAJohn Strawn, S. Systems, Larkspur, CA, USA

While many standards exist for the streaming of digitalcontent over the Internet at large, there are no widelyadopted standards for LAN distribution or for the integra-tion of LAN and WAN distribution systems. A better un-derstanding of the issues and best practices relevant toLAN streaming, including fundamental capabilities, basictransport mechanisms, signal treatment, and infrastruc-ture requirements is needed by the industry. Further-more, consideration of networked media distribution,both though WAN and LAN, as an alternative to “hardmedia” (e.g., CDs) poses important questions to the digi-tal library community. This workshop will explore thesequestions and present a synopsis of approaches beingemployed by leaders in the field.

15:30 h Room 2Standards Committee Meeting on SC-04-07 ListeningTests

15:40 h Room JTechnical Committee Meeting on Coding of AudioSignals

Session V Sunday, March 23 16:00–18:00 hTopaz Lounge

(POSTERS) SIGNAL PROCESSING FOR AUDIO

16:00 h

V-1 High-Resolution Robust Multipitch Analysis ofGuitar Chords—Laurent Bonnet, Roch Lefebvre,University of Sherbrooke, Quebec, Canada

An algorithm capable of extracting multipitch infor-mation from a guitar sound is described. Themethod is based on a two-stage approach. First,the sound is segmented in time, based on the deriv-ative of the signal envelope. This defines the onsetof the sound. Then, an iterative procedure based on

the tracking and subtraction of the strongest har-monics sets is applied to the amplitude spectrum toestimate the candidate fundamental frequencies orharmonics. To improve the frequency resolution ofthe transform used, frequency-bins interpolation isapplied around the detected harmonics in the spec-trum. The system tested with simulated multipitchsignals achieves reliable fundamental frequencydetection. With real guitar chords, the performanceof the algorithm depends on the harmonic complex-ity of the sound.Convention Paper 5772

16:00 h

V-2 Control of Signal Processing Algorithms Usingthe MATRIX Interface—Dan Overholt, University ofCalifornia at Santa Barbara, Santa Barbara, CA, USA

This paper describes the use of the MATRIX inter-face as a controller for several signal processing al-gorithms, including delay, reverb, EQ, and chorus.Although these algorithms are conventional effects,the multidimensional control provided by the MA-TRIX allows users to manipulate audio in a mannerthat was not previously viable in a real-time system.[Paper not presented at convention, but ConventionPaper 5773 is available.]

16:00 h

V-3 Extraction of Weak Background Transients fromAudio Signals—Yves Grenier, Bertrand David,ENST, Paris, France

Extracting a weak audio signal buried below astronger one is a difficult task that may be encoun-tered in forensic applications. Extraction of thebackground signal is made difficult by the signal-to-noise ratio that is clearly negative. Apart from thisdifficulty, the problem would look like the problem ofseparating different components from an audio sig-nal. We investigate the possibility of using separa-tion techniques for this extraction, based upon har-monic models of the stronger signal. We comparetheoretically and practically two approaches: in thefirst one, estimation of harmonic models followed bysubtraction of the harmonic signal; in the secondone, high resolution tracking of damped sinusoidsand synthesis of each component after sorting theindividual patterns.Convention Paper 5774

16:00 h

V-4 Non-Linear Dynamics Processing—JiriSchimmel, Brno University of Technology, Brno,Czech Republic

Dynamics processing is performed by amplifyingdevices where the gain is automatically controlledby the level of the input signal. Nonlinear compo-nents simulating tube amplifiers can be used inthese devices to make the musical signal audiblydense. This paper deals with the simulation of tubeamplifiers using the power polynomial approxima-tion of transfer characteristic and with computationmethod of the power polynomial coefficients ac-cording to the required higher harmonics ratio. Thepaper also shows the influence of nonlinear ampli-fying devices simulating tube amplifiers on the out-put signal spectrum of dynamic effects. Convention Paper 5775


16:00 h

V-5 Error Correction in Class AB Power Amplifiers—Dimitri Danyuk1, Michael Renardson21Kiev, Ukraine2Sunnybank, Hull, UK

An audio power amplifier design is presented whichcan linearize the transfer characteristics of conven-tional class AB output stages in the crossover region.The objective is to offer practical circuits with errorcorrection, that overcome nonlinearities, inherent tothe crossover region of class AB output stages.Convention Paper 5776

16:00 h

V-6 Pitch Detection of the Singing Voice in MusicalAudio—Saurabh Shandilya, Preeti Rao, Indian In-stitute of Technology, Bombay, India

This paper explores the extraction of melodic pitchcontour from the polyphonic soundtrack of a song.The motivation for this work lies in the desire for au-tomatic tools for the melody-based indexing of thedatabase in a music retrieval system. The melody isassumed to be carried by the singer’s voice accom-panied mainly by a percussive instrumental musicbackground. The main challenges raised by thisproblem are presented. A specific aspect of the prob-lem is considered in detail, and the performance oftwo different classes of pitch detection algorithms isinvestigated. A pitch detection method based on aperception model is shown to be a promising ap-proach to the tracking of voice pitch in the presenceof strong percussive background.[Paper not presented at convention, but ConventionPaper 5777 is available.]

16:00 h

V-7 Real-Time Wavelet Packet-Based Low Bit RateAudio Coding on a Dynamic ReconfigurationSystem—Alexey Petrovsky1, Detlef Krahe2,Alexander Petrovsky1,31Belarusian State University of Informatics and Radioelectronics, Minsk, Republic of Belarus

2Wuppertal Technical University, Wuppertal, Germany3Technical University Bialystok, Bialystok, Republic of Poland

Real-time implementation of a psychoacousticallymotivated wavelet packet-based monophonic full-duplex audio coder using a dynamic algorithmtransforms approach is proposed. The principle be-hind the approach is to define the parameter of theinput audio signals (subband entropy) and outputencoded sequences (subband rate) for the givenembedded processor architecture. Adaptivewavelet analysis for audio signal coding purposes isparticularly interesting if the psychoacoustic infor-mation is considered in the WP decompositionscale. The advantages of this approach are betterviewed by considering the wavelet packet growingas a splitting process, i.e., the temporal construc-tion WP tree created for each signal frame presentsan ideal decision for real-time processing imple-mented in a reconfigurable hardware.Convention Paper 5778

16:00 h

V-8 Results for Room Acoustics Equalization Basedon Smoothed Responses—Panagiotis D.

Hatziantoniou, John N. Mourjopoulos, University ofPatras, Patras, Greece

Digital equalization of room acoustics based on in-verse filtering of measured response functions, in-troduces a number of theoretical and practical chal-lenges. To overcome such problems, inversefiltering based on modified measured responses isproposed, derived via their complex transfer func-tion smoothing, so that the processed responsesare more perceptually compliant, of lower order,and less position-sensitive than the original func-tions. The aim of this paper is to evaluate via objec-tive and subjective tests conducted for different-sized rooms and real-time reproduction, the use ofsuch smoothed room responses for the derivationof appropriate room equalization filters, which canimprove the perceived and measured quality of au-dio reproduction in any reverberant environment. Convention Paper 5779

Workshop 7 Sunday, March 23 16:00–18:00 hRoom B

HANDSET AND HEADSET TESTING: Beyond Narrowband

Chair: Andrew Bright, Nokia Corporation, Helsinki, Finland

Panelists: Paul Darlington, Apple Dynamics Ltd., Mold, Flintshire, UKHans Gierlich, Head Acoustics, Herzogenrath, GermanyLars Bierger Nielsen, Brüel & Kjær, Nærum, DenmarkAllen Woo, Plantronics, Santa Cruz, CA, USABob Zurek, Motorola, USA

Wideband speech and music playback are new featuresbeing introduced to mobile telephones. The fundamentalconcepts behind testing handsets and headsets over thenarrow-telephone frequency band are not applicable atthe higher frequencies these new features will repro-duce. New test methods are needed to assess the per-formance of handsets and headsets over these wider fre-quency ranges. An open discussion between testequipment manufacturers, researchers, and telephonemanufacturers is planned to explore possible ways for-ward. Topics that will be discussed are artificial ear de-sign and usage, relation between objective and subjec-tive measurements, and safety and standards.

Student ActivitiesJOB/CAREER SEMINARSunday, March 23, 16:00–17:30 hRoom H

A panel of representatives from the audio industry willaddress the issues of entry-level employment and expec-tations of employers. This panel of experts in the audiofield will discuss the present state of the job market.Broad aspects of the audio industry will be represented,and enthusiastic participation by attendees is anticipated.

Historical PresentationANCIENT ELECTRONIC MUSICSunday, March 23 16:00–17:00 h

In the early days of electronic music development, theproject of the specially composed Poème Electronique atthe 1958 Brussels World Fair was a major referencepoint. A presentation outlining this pioneering project ofcomposers Edgard Varèse and Lannis Xenakis in combi-



nation with a lightshow and decorations designed by LeCorbusier and Philippe Agostini, who were also the ar-chitects of the Philips Pavilion, will be given by expertKees Tazelaar.

16:30 h Room 1Standards Committee Meeting on SC-02Subcommittee on Digital Audio

16:40 h Room C/DTechnical Committee Meeting on Acoustics andSound Reinforcement

16:40 h Room E/FTechnical Committee Meeting on Signal Processing


A SIMPLE WAY TO RECORD SURROUND SOUND

Soundfield Ltd.Presenter: Pieter Schillebeeckx

This seminar gives an overview of the technology behindSoundField microphone systems and compares these toother stereo and surround sound recording methods.Special attention will be paid to the flexibility and benefitsof single point source 3-dimensional sound capture.

The seminar will be an overview interlaced with 5.1surround sound demonstrations. Aspects covered willrange from the physical details of the SoundField tetra-hedral capsule array, the operation of the control units,and an explanation of B-Format, the underlying coreprinciple on which SoundField technology is based.


RIBBON TECHNOLOGY FOR LOUDSPEAKERS

ALCONS AUDIOPresenter: Tom Back

Alcons Audio, the newly-established Amsterdam-basedcompany, will introduce themselves. In this presentationthe company’s vision and its strive for natural sound re-production at low as well as high sound pressure levelswill be described. Alcons’ proprietary high-power profes-sional ribbon technology for loudspeakers will be high-lighted.

Part of the presentation is an interactive demonstrationwhere visitors bringing their favorite CD, can get ac-quainted with Alcons’ professional ribbon technology. Arelaxed, but very informative, way of finishing the day.Don’t forget to bring your own CD!

Special EventOPEN HOUSE OF THE TECHNICAL COUNCIL ANDTHE RICHARD C. HEYSER MEMORIAL LECTURESunday, March 23, 18:10–20:00 hRoom A

Lecturer: Jens Blauert

The Heyser Series is an endowment for lectures by emi-nent individuals with outstanding reputations in audio engi-neering and its related fields. The series is featured twiceannually at both the United States and European AES con-ventions. Established in May 1999, The Richard C. HeyserMemorial Lecture honors the memory of Richard Heyser, ascientist at the Jet Propulsion Laboratory, who was award-ed nine patents in audio and communication techniques

and was widely known for his ability to clearly present newand complex technical ideas. Mr. Heyser was also an AESgovernor and AES Silver Medal recipient.

The Richard C. Heyser distinguished lecturer for the114th AES Convention is Professor Jens Blauert, AESFellow and professional acoustical consultant in the stateof North Rhine Westphalia, where he is also a memberof the Environmental-Protection Council. Prof. Blauert’sfields of interest include binaural technology, models ofbinaural hearing, architectural acoustics, noise engineer-ing, product-sound design, speech technology, virtualenvironments, and telepresence. The author/coauthor ofmore than 140 papers and monographs, Prof. Blauerthas been awarded several patents and is a cofounderand board member of the German Acoustical Society.

Blauert’s lecture, “Communication Acoustics: Audio GoesCognitive!” will explore how the branch of acoustics that re-lates the information technologies has experienced a dra-matic evolution over the past 30 years. Using multimodalsynthesis tools to manufacture auditory scenes, we can pro-vide an astonishing amount of perceptual plausibility, pres-ence, and immersion. These synthesis tools are capable ofsimulating, in addition to sound, the senses of touch, vi-sion, and even the illusion of movement. The synthesissystems are parameter controlled and often interactive.

Many audio engineers are not yet ready to meet thechallenge of designing knowledge-based and multimodalsystems. Cognitive and multimodal phenomena have tobe considered in both audio analysis and synthesis. Con-sequently, future audio systems will often contain knowl-edge-based and multimodal components.

We will increasingly see audio systems being embed-ded in more complex systems and this technologicaltrend will coin the future of communication acoustics,and ultimately, audio engineering. Using computationalauditory scene analysis (CASA) as an example, Prof.Blauert will discuss the development of algorithms for theextraction of parametric representations of real auditoryscenes, and the significant role audio signal processing,symbolic processing, and content processing plays inthis process, as well as how audio engineers can pre-pare for this next wave of the audio future.

Session I Monday, March 24 09:00–12:00 hRoom C/D

AUDIO NETWORKING

Chair: Jeremy Cooperstock, McGill University, Montreal, Quebec, Canada

09:00 h

I-1 Network Audio Recording Environment—MarkGordon1, William Hsu21Soundgordon Consulting, San Francisco, CA, USA2San Francisco State University, San Francisco, CA, USA

Musicians in a recording session are used to collab-orating in real time. Distance makes this difficult,and current technologies that allow distance collab-oration are either prohibitively expensive or rely onnonreal-time, store-and-forward strategies. The net-work audio recording environment (NARE) is an ob-ject-oriented, client-server environment that en-ables real time, collaborative audio production. Thesystem provides full-duplex streaming of data overTCP/IP networks and establishes a custom mes-saging protocol to handle the communication of au-dio and control information between the client andserver. By using the same double buffers for localreading and writing and for network sending and re-


ceiving, NARE allows musicians to enjoy the realtime collaborative interaction that store-and-forwardsystems do not. The success of this project demon-strates that real time, collaborative audio on the Internet is within reach. NARE provides an inexpen-sive solution without the need for proprietary hard-ware or accounts on centralized servers.Convention Paper 5780

09:30 h

I-2 Latency in Audio Ethernet Networks—Nuno Fonseca1, Edmundo Monteiro21Polytechnic Institute of Leiria, Leiria, Portugal2University of Coimbra, Coimbra, Portugal

In a time when several audio Ethernet networkingsolutions are being studied and developed, theanalysis of the latency introduced by theses net-works is fundamental. This analysis is the subject ofthe present paper and is necessary not only to en-able the identification of the factors that can be opti-mized, but also to support the decision about thepossibility or not of the inclusion of in-band syn-chonism signaling.Convention Paper 5781

10:00 h

I-3 On the Performance of Clock SynchonizationAlgorithms for a Distributed Commodity AudioSystem—Men Muheim, Philipp Blum, Swiss FederalInstitute of Technology, Zürich, Switzerland

In this paper we investigate clock synchonization al-gorithms for a distributed audio system built withcommercial off-the-shelf IT components. Ethernettechnology and standard PC hardware offer highperformance at a low price but lack important timeli-ness properties. To allow accurate reproduction ofmultichannel audio, the interfaces have to be syn-chronized within a few microseconds. We have im-plemented thee different types of clock synchoniza-tion algorithms on a Linux based system and haveanalyzed the achieved accuracy. All algorithms canachieve the target accuracy of 10 µs. They differ inthe resilience against variable network load andcommunication requirements. The results show thatEthernet can be used for a distributed audio systemunder professional synchonization requirements.Convention Paper 5782

10:30 h

I-4 Perceptually Motivated Low ComplexityAcoustic Echo Control—Chistof Faller, AgereSystems, Murray Hill, NJ, USA

In hands-free two-way communications systems, theloudspeaker signal feeds back to the microphone, re-sulting in an undesired echo signal component in themicrophone signal. Acoustic echo cancelers modelthe echo path and subtract an estimate of the echosignal from the microphone signal to remove the un-desired echo. We present a novel scheme which es-timates the echo signal in terms of its spectral enve-lope. The time and frequency resolution with whichthis estimation is carried out is chosen according toperceptual criteria. Given the estimated spectral en-velope of the echo signal component, speech en-hancement, and noise suppression algorithms areused to suppress the echo. The presented schemehas low complexity and a high degree of robustness.Convention Paper 5783

11:00 h

I-5 A New Connection Management Architecture forthe Next Generation of mLAN—Jun-ichi Fujimori1,Richard Foss21Yamaha Corporation, Hamamatsu, Japan2Rhodes University, Grahamstown, South Africa

mLAN is a networking technology based on the IEEE1394 standard that allows for the transport of audioand music control data between audio devices. In theoriginal implementation of mLAN, software within eachmLAN node hosted by an audio device contained highlevel plug abstraction and connection managementsoftware. This paper describes a new connectionmanagement architecture that splits the connectionmanagement function between workstation and de-vice. The high level connection management andplug abstraction capability resides on the workstation,while a thin low level connection management capa-bility is left on the device. This approach reduces costand complexity on the device side and ensures thatmLAN systems can be easily upgraded.Convention Paper 5784

11:30 h

I-6 A Comparative Study of mLAN and CobraNetTechnologies and Their Use in the Sound Installation Industry—Bradley Klinkradt, RichardFoss, Rhodes University, Grahamstown, South Africa

This paper highlights the two interconnection tech-nologies of CobraNet and mLAN, and provides acomparative study of these technologies and theirapplicability to the sound installation industry,though a discussion of constraints inherent withinsound installations. Issues such as the adherenceto standards, costs, latency, speed, connectionmanagement, and the control and monitoring of de-vices are explored.Convention Paper 5785

Session J Monday, March 24 09:00–11:30 hRoom E/F

SPATIAL PERCEPTION AND PROCESSING, PART 2

Chair: Gunther Theile, Institut für Rundfunktechnik GmbH, Munich, Germany

9:00 h

J-1 A Novel Method for the Efficient Comparison ofSpatialization Conditions—Michael C. Kelly, Anthony I. Tew, University of York, Heslington York,UK

In this paper we present a novel sound source local-ization method that requires the spatial location oftwo sound sources to be matched by a listener. Themethod is aided by auditory feedback and effectivelyprovides a measure of the minimum audible angle ofthe system under test. The number of front-back re-versals and the time taken to localize each sourceare provided as additional indicators of performance.We demonstrate the application of our method bycomparing localization accuracy for a particularsource with and without a secondary source present.The results demonstrate a small but significant angu-lar increase in the localization error for the dual-source condition and also an increase in the timetaken to perform the localization task.Convention Paper 5786



9:30 h

J-2 A Study on Head Shape Simplification UsingSpherical Harmonics for HRTF Computation atLow Frequencies—Yufei Tao, Anthony I. Tew,Stuart J. Porter, University of York, Heslington,York, UK

Simplified head shapes, such as spheres and ellip-soids have been applied in the research of head-re-lated transfer functions (HRTFs) for their simplicity.However, the effects of the missing head shapefeatures in the simplified head models have notbeen thoroughly examined. In this paper headshapes are represented using spherical harmonics(SHs), which allows the simplification of headshapes to be carried out in a controlled and system-atic way. The KEMAR head shape is low-passed todifferent degrees. The errors in both the headshape and acoustic pressures introduced by thelow-pass filters (LPFs) are studied. Influence on theHRTFs by the head shape features away from theears, e.g., the nose, are discussed.Convention Paper 5787

10:00 h

J-3 Further Investigations of High-Order Ambisonicsand Wavefield Synthesis for Holophonic SoundImaging—Jérôme Daniel, Rozenn Nicol, SébastienMoreau, France Telecom R&D, Lannion, France

Ambisonics and wavefield synthesis are two waysof rendering 3-D audio, which both aim at physicallyreconstructing the sound field. Though they derivefrom distinct theoretical fundamentals, they have al-ready been shown as equivalent under given as-sumptions. This paper generalizes this equivalenceby introducing new results regarding the coding andrendering of finite distance and enclosed sources.An updated view of the current knowledge is firstgiven. A unified analysis of sound pickup and repro-duction by mean of concentric transducer arraysthen provides an insight into the spatial encodingand decoding properties. While merging the analy-sis tools of both techniques and investigating themon a common ground, general compromises arehighlighted in terms of spatial aliasing, error, andnoise amplification. Convention Paper 5788

10:30 h

J-4 Perceptual Evaluation of Binaural Sound Synthesis: The Problem of Reporting LocalizationJudgments—Jean-Marie Pernaux, Marc Emerit,Rozenn Nicol, France Telecom R&D, Lannion, France

This paper relates to work conducted in our labora-tory about perceptual evaluation of 3-D sound ren-dering techniques and, more specifically, binauralsound synthesis. Our aim is to build a perceptualevaluation workbench for the comparison of perfor-mance of various head related transfer functions(HTFs), implementations or models. In a previouspublication, localization tests were performed oneight subjects using individual HTFs. A simple 2-Dgraphical interface using the computer mouse al-lowed subjects to report their localization judg-ments. This interface measured both auditory lo-calization error and reporting error. A localizationtest was performed using a finger pointing tech-nique, under the same experimental conditions toestimate the reporting error and its variations with

the reporting technique. To improve our mouse in-terface, we proposed an enhanced 2-D reportinginterface with a 3-D individual visual feedback. Lo-calization results using the thee interfaces arecompared regarding localization error and disper-sion of the judgments. Data analysis show a signifi-cant reduction of reporting error using both the 2-Dinterface with 3-D visual feedback and the fingerpointing technique. Last, in order to evaluate moreaccurately the reporting error, an original localiza-tion test is proposed replacing the auditory positiondependent stimulus by a tactile stimulus.Convention Paper 5789

11:00 h

J-5 Sound Field Management with the ArkamysSound Process—Jean-Michel Raczinski, JérômeMonceaux, Georges-Claude Vieilledent, Arkamys,Paris, France

This paper presents a new audio process namedArkamys Sound Process. The process is based onthe diffusion of sounds in an acoustic environmentand the capture of these sounds with a specific de-vice. The digital translation of this model results in aspecific database made up of transfer functions thatinclude the acoustic characteristics of the modeland some optimizations. The implementation of thecorresponding filters on a standard general purposeprocessor (1 GHz Pentium III) allows the process-ing of a stereo stream at 48 kHz. The paper focus-es on the main properties of the process: soundfield separation, immersion, and depth; bringing realism and naturalness on binaural, stereo, andmultichannel diffusion devices.Convention Paper 5790

Workshop 8 Monday, March 24 09:00–12:00 hRoom B

ELECTRONIC REVERBERATION FOR CONCERTHALLS


Panelists: David Griesinger, Lares, Belmont, MA, USAEckhard Kahle, Kahle Acoustics, Brussels, BelgiumMendel Kleiner, Chalmers University, Gothenburg, SwedenBen Kok, Dorsserblesgraaf, Eindhoven, The NetherlandsScott Pfeiffer, Kirkegaard Assoc., Chicago, IL, USAMarc Poletti, VRAS Industrial Research, New ZealandJean-Paul Vian, Carmen Cstb, FranceDiemer de Vries, Delft University, Delft, The NetherlandsLeon van Zuylen, SIAP, Uden, The Netherlands

Many new developments have been made in electronicreverberation systems for variable acoustics in auditoria,since the original assisted resonance systems of the1960s (e.g., Royal Festival Hall, London, Grand KremlinTheater, Moscow) were used to selectively increase re-verberation time. The evolution of new systems has beendriven by both modern signal processing technology anda better understanding of the perceptual issues relevantto sound quality for both music and speech. This work-shop was organized around a set of questions that were


developed in advance by the participants to addressmany of the important perceptual issues of variableacoustics. Each panelist will have the opportunity to pre-sent a concise answer to the same questions, allowingfor the audience a comparative view of similar and dis-similar approaches to the same topic. The panelists rep-resent a good cross section of product developers, re-searchers, and acoustical consultants who areconcerned with the perception of reverberation.

Seminar 7 Monday, March 24 09:00–12:00 hRoom A

MIXING AND MASTERING

Chair: Terry Nelson, Studio Equipment, Switzerland

Panelists: Jim AndersonStuart BruceRoland GuillotelGeorge MassenburgAnthony Morris

This seminar includes a panel of leading industry profes-sionals who will discuss various formats for mixing andmastering in surround. Topics to be covered include:

• Mixing and mastering to the various multichannel formats,discrete channels and other solutions

• Mixing for 5.1 vs. 6.0, 7.1 or 10.2• Multichannel mix recording formats in most common

use and compatibility with mastering needs• Mixing and mastering monitor loudspeaker setup/

formats and compatibility with common home theatersetups in the home

• Essential tools for mixing and mastering in multichannel surround

• Mix approaches and preferences: ambient surround vs. direct surround

• Challenges of mastering for DVD-A• Challenges of mastering for SACD• Using stereo material for processing to multichannel

Audio and music examples will be played.

09:00 h Room 1Standards Committee Meeting on SC-04-04Microphone Measurement and Characterization

Student ActivitiesEDUCATION FORUMMonday, March 24, 10:00–12:00 hRoom H

This event is a meeting of the AES Education Commit-tee, teachers, authors, students, and members interest-ed in the issues of primary and continuing education ofthe audio industry. It is an opportunity to discuss the pro-grams of the Education Committee and to provide inputfor future projects of this committee.

Historical PresentationHISTORICAL COMMITTEE MEETINGMonday, March 24 10:00–12:00 hRoom J

Exhibitor Seminar 15 Monday, March 2410:30–11:30 h Room P

A NEW APPROACH TO SOLVING BROADCASTLOUDNESS PROBLEMS: THE DOLBY LM100BROADCAST LOUDNESS METER

Dolby Laboratories Inc.Presenter: John Couling

Inconsistent loudness is a significant problem in broadcastsound, causing viewer annoyance and complaints. Recentstudies have shown that channel-to-channel and program-to-program level discrepancies can be as much as 18 dB or20 dB.This session will discuss how broadcasters and pro-gram makers can start to solve these problems using thenew Dolby LM100 Broadcast Loudness Meter. A noveltechnique will be introduced that enables rapid and accurateassessments of program loudness to be made using thenew “Dialogue Intelligence” feature of the LM100 meter.

Exhibitor Seminar 16 Monday, March 24 10:30–11:30 hRoom L

PRACTICAL APPLICATION OF NEW DIGITALDIRECTIVITY CONTROL CONCEPTS

Duran AudioPresenter: Evert Start

In addition to the digital directivity control (DDC) technol-ogy, implemented in the well-known Axys Intellivox loud-speaker columns, Duran Audio introduces a new andeven more powerful directivity concept: digital directivitysynthesis (DDS).

Using DDS, the radiation pattern of a loudspeaker ar-ray can be molded into any shape to match any desiredsound distribution. The dedicated digital directivity analy-sis (DDA) software enables easy simulation, optimiza-tion, and analysis of complex directivity patterns.

In practice, DDS has already proven to be a useful andeven essential tool for the improvement of the sound(re)production in large, acoustically difficult spaces. Theflexibility of the DDC and DDS concepts will be illustratedby practical design examples.

Session X Monday, March 24 11:00–13:00 hTopaz Lounge

(POSTERS) ACOUSTICS OF RECORDING AND REPRODUCTION

11:00 h

X-1 Digital Stereo Recording of TraditionalMalaysian Musical Instruments—Rick Shriver,Ohio University–Zanesville, Zanesville, OH, USA

This paper offers details of the recording tech-niques and editing processes of the wave samplesof traditional Malaysian musical instruments, whichare used for a multimedia project. The project’s out-come includes a bank of digital samples of the in-digenous instruments for musicians and composersto use with conventional triggering devices such askeyboards, computers or drum triggers. The paperincludes a discussion of Malaysian instrument clas-sification, an examination of Malaysian media musi-cal content, and a brief analysis of contemporaryMalaysian musical culture.[Paper not presented at convention, but ConventionPaper 5791 is available.]

11:00 h

X-2 An Application of Ambiophony for the Enhance-ment of the Reverberant Environment Inside theWalls of a Byzantine Castle—George Papaniko-laou, George Kalliris, Christos Goussios, Charalam-pos Dimoulas, Aristotle University of Thessaloniki,Thessaloniki, Greece

Beethoven’s opera Fidelio was performed inside thewalls of Thessaloniki’s Byzantine castle, known asEptapirgion. The need for the raise of the reverbera-


tion time of this open space was satisfied with an am-biophonic application that used electroacoustical de-vices in order to enhance the already-existing reflec-tions. The whole system was designed to provide areverberant and highly diffusive character. Convention Paper 5792

11:00 h

X-3 A Method of 3-D Sound Image Localization UsingLoudspeaker Arrays—Yasushige Nakayama,Kaoru Watanabe, Setsu Komiyama, Fumio Okano,Yoshinori Izumi, NHK Science and Technical Re-search Laboratories, Kinuta Setagaya-ku, Tokyo,Japan

We are studying 3-D sound image reproduction sys-tems associated with 3-D video images to create ahighly realistic form of 3-D television broadcasting inthe future. This paper describes a method of 3-Dsound image control arbitrarily and continuously us-ing loudspeaker arrays. Two kinds of subjective lis-tening tests were conducted for distance and direc-tional perception. We have confirmed that theposition of the sound image can be controlled in 3-Dspace by using this method.[Paper not presented at convention, but ConventionPaper 5793 is available.]

11:00 h

X-4 The Analysis of Peculiarities of Russian BellsAcoustic Parameters—Irina Aldoshina1, Stanislav Pychkov1, Igor Matcievski2, Alexandr Nicanorov2,Peter Tovstik3, Stepan Cherniav31University of Humanities and Social Sciences, Moscow, Russia

2Russian Institute History of Arts, St. Petersburg, Russia

3State University, St. Petersburg, Russia

The peculiarities of the tuning and acoustical char-acteristics of Russian bells were investigated byvarious scientists for a long time. The results werereviewed in our previous report (108th ConventionAES). In this paper the results of new research arerepresented including: digital recording of the soundof 16 twentieth-century bells made in variousmonasteries and temples of Russia; the computerprocessing and restoration of obtained recordings;the spectral and statistical analysis of soundingsand comparison of bells’ tuning with that of a con-ventional Dutch system; development of mathemati-cal models of bells’ vibration; creation of the soft-ware for the analysis spectral frequencies andmodes of vibration; and synthesis of their geometricform to optimize the structure of spectrum.Convention Paper 5794

11:00 h

X-5 Implementation of a Delay Matrix—Michael S. Pincus, Acentech Incorporated, Cambridge, MA, USA

This paper is a follow-up to “Distributed Sound Reinforcement for Multiple Talker Locations,” pre-sented at the AES 21st International Conference.That paper described a delay matrix used to helplisteners localize several talker locations within aspace. This paper will review that technique and de-scribe its implementation in a recently completedproject. Digital audio files will be presented allowinga subjective comparison between the simulatedmodel and recordings made in the actual space.Convention Paper 5795

11:00 h

X-6 Matrixed Multichannel Extension for AAC Codec—Marek Szczerba, Frans de Bont, Werner Oomen,Leon van de Kerkhof, Philips Digital System Laboratories, Eindhoven, The Netherlands

With the introduction of new standards like DVDand SACD, multichannel audio systems are grow-ing in popularity. In order to maintain compatibilitytoward customers that have not migrated to a multi-channel environment, there is a strong demand forstereo compatibility. Stereo compatibility can beachieved by using thee methods: simulcast, post-mix, and premix. Simulcast requires extra band-width to convey the data. The postmix technique re-quires an overly complex and power consumingstereo decoder and is therefore unsuitable forportable applications. Only the premix or matrixedmultichannel technique allows low complexity, lowpower stereo decoders in combination with efficientuse of bandwidth. In this paper a matrixed multi-channel method based on the MPEG AAC standardis presented. To ensure the highest possible audioquality for both the compatible down-mix and themultichannel signal under all circumstances, sever-al new solutions were introduced. These includedominant center and dominant surround process-ing, as well as common window switching. Subjec-tive listening test results are presented.Convention Paper 5796

11:00 h

X-7 Measurement of Sampling Jitter Using a Musical Signal—Akira Nishimura, Nobuo Koizumi, TokyoUniversity of Information Sciences, Wakaba-ku,Chiba-city, Chiba, Japan

A method to measure sampling jitter which might begenerated by a digital-to-analog converter (DAC) oran analog-to-digital converter (ADC) while repro-ducing or recording musical signals is discussed.We propose a method for estimating sampling jitterwaveforms while reproducing or recording musicalsignals by modification to jitter measurement whichutilizes analytic signals. Computer simulations ofthe measurement technique revealed that the mini-mum detectable jitter amplitude was approximately3 ns. Experimental measurements of sampling jitterin the ADC and DACs were also conducted using amusical signal and a pure tone. The results of themeasurements showed that no measurable jittercomponent was induced by the musical signal.Convention Paper 5797

11:00 h

X-8 Design of Digital Audio Amplifier for Automotive Applications—Alberto Bellini, Antonio De Benedetti,Giovanni Franceschini, University of Parma, Parma,Italy

Switching power amplifiers are becoming quitecommon in audio applications, thanks to semicon-ductor technology advances because of their intrin-sic optimal efficiency. However, usually they areused for low/medium quality application and low fre-quency loudspeaker systems. In this paper the de-sign of a large bandwidth switching audio amplifieris presented. The amplifier is specifically aimed atautomotive applications, where supply voltage,power consumption, and size are peculiar con-straints. A prototype featuring reduced distortion,



114th ConventionPapers and CD-ROMConvention Papers of many of the presentations given at the 114th Convention and a CD-ROM containing the 114thConvention Papers are available from the AudioEngineering Society. Prices follow:

114th CONVENTION PAPERS (single copy)Member: $ 5. (USA) or £3.50 (UK)

Nonmember: $ 10. (USA) or £7.00 (UK)

114th COMPLETE SET (139 papers)Member: $150. (USA) or £ 104. (UK)

Nonmember: $200. (USA) or £ 138. (UK)

114th CD-ROM (139 papers)Member: $110. (USA) or £ 75. (UK)

Nonmember: $160. (USA) or £110. (UK)

114th CONVENTION PAPER SET AND CD-ROM (139 papers, each)

Member: $200. (USA) or £138. (UK)Nonmember: $270. (USA) or £186. (UK)

A complete list of the 114th Convention Papers with a convenient order form is inserted in this issue.


TECHNOLOGY OF ADVANCED AUDIO CODING

Telos SystemsPresenter: Steve Church

Telos Systems will discuss developments in the technolo-gy of spectral band replication (SBR) and advanced audiocoding (AAC), especially as they pertain to codecs.MPEG2-AAC and SBR make a perfect match. AAC is thenewest audio coding method selected by MPEG, and be-came an international standard in April 1997. It is a fullystate-of-the-art audio compression tool kit outperformingany earlier approach. Combining AAC with SBR results ina superset of AAC technology—trade name “aacPlus”—greatly increasing the efficiency of AAC. This presentationwill give you an overview of the subject.

Exhibitor Seminar 18 Monday, March 24 12:00–13:30 hRoom P

DIGITAL AUDIO PRODUCTS INCLUDING DIGITALEQ, A/D, DEESSER/COMPRESSOR, PINGUIN AUDIOMETER

WEISS ENGINEERING LTD.Presenter: Daniel Weiss

In the first part of the seminar the newly introducedWeiss ADC2 A/D Converter will be described in detail. Inaddition the EQ1-DYN dynamic EQ and the DS1-MK2DeEsser/Compressor will be presented.

The second part of the demonstration shows somenew options of the software-based Pinguin Audio Meter:


suitable for large bandwidth loudspeakers was real-ized and successfully tested.Convention Paper 5798

11:00 h

X-9 Emerging and Exotic Auditory Interfaces—Michael Cohen, University of Aizu, Aizu-Wakamat-su, Japan

Anticipating some emerging audio devices and fea-tures, this paper surveys trends in mobile telephony(especially regarding mobile internet in Japan), wear-able/intimate multimedia computing, hand-held/no-midic/portable interfaces, and embedded systemslike multimedia furniture and spatially immersive dis-plays, gleaned from recent press releases, popularmedia, and publications by industrial and academiclaboratories, and the author’s own research group.Such extended and enriched audio interfaces, espe-cially coupled with position tracking systems, encour-age multipresence, the inhabiting by sources andsinks of multiple spaces simultaneously, allowing, forinstance, a user to monitor several aligned spaces atonce (conferences, entertainment, navigation, warn-ings, etc.). Representative instances are cited, andgroup-ware selection functions and their predicatecalculus notation are reviewed.Convention Paper 5819

11:00 h Room 1Standards Committee Meeting on SC-06Subcommittee on Network and File Transfer of Audio

11:40 h Room E/FTechnical Committee Meeting on Audio for Games

114th Convention2003 March 22–March 25Amsterdam

Presented at the 114th Convention2003 March 22–March 25 • Amsterdam


loudness measuring for CD and DVD mastering as wellas for broadcasters. New analyzing methods (IRT) andscales such as the three of the American mastering ex-pert Bob Katz (Digital Domain, Chesky) as well as thenew international loudness standard proposal by Ger-many’s IRT will be presented with audible examples.

12:20 h Room ATechnical Committee Meeting on Studio Practicesand Production

12:10 h Room BTechnical Committee on Archiving, Restoration, andDigital LIbraries

12:10 h Room C/DTechnical Committee Meeting on Network AudioSystems

12:30 h Room 1Standards Committee Meeting on SC-05Subcommittee on Interconnections

Session K Monday, March 24 13:30–16:30 hRoom C/D

MULTICHANNEL SOUND

Chair: Wieslaw Woszczyk, McGill University, Montreal, Quebec, Canada

13:30 h

K-1 Transforming Ambiophonic + Ambisonic 3D Surround Sound to and from ITU 5.1/6.1—RobertE. (Robin) Miller III, FilmakerStudios, Bethlehem, PA,USA

ITU 6.1 with six discrete full-range audio channels,implemented in DVD-A, SACD, and DTS-ES Dis-crete, provide the means to deliver full sphere peri-phonic 3-D surround sound. For compatible distrib-ution, the channels are transformed for 5.1/6.1reproduction, but can still be fully recovered for“PerAmbio” reproduction—an Ambisonic + Ambio-phonic hybrid approach, described in a prior paper,that maximizes 3-D envelopment along with frontstage imaging and spaciousness, while economiz-ing the number of channels and speakers. To clarifythat fewer media channels “r” are required thanspeakers “s” the use of MCN (multichannel number-ing) in the form “r.lfe.s” is proposed. Experimental“PerAmbio 6.1.10” (10 speakers or more plus sub-woofer) recordings test six encoding variations ap-plicable in home theater, virtual reality, and music-only production.Convention Paper 5799

14:00 h

K-2 Improving Speech Intelligibility in Teleconfer-encing by Using Wave Field Synthesis—MarinusM. Boone, Werner P. J. de Bruijn, University ofTechnology, Delft, The Netherlands

Large screen teleconferencing can be enhancedconsiderably with the application of spatial soundrecording, transmission, and reproduction. True

spatial sound reproduction can be obtained withwave field synthesis (WFS) which gives a sound re-production that is independent of the listener posi-tion. Our research has shown that a significant im-provement of speech intelligibility can be obtainedwith WFS as compared with a single loudspeakerreproduction, when there are several interferingspeech signals. The improvement in speech recep-tion threshold (SRT) can be more than 2 dB, mak-ing a change in speech intelligibility from 50 percentto more than 85 percent.Convention Paper 5800

14:30 h

K-3 Application of Wave Field Synthesis in Life-SizeVideoconferencing—Werner P. J. de Bruijn, Marinus M. Boone, Delft University of Technology,Delft, The Netherlands

Spatial reproduction of the voices of conferenceparticipants can greatly enhance the performanceof a life-size videoconferencing system in terms ofqualities such as speech intelligibility, speaker iden-tification, and more generally, the naturalness of aconference. A very suitable technique to implementaccurate spatial sound reproduction including depthis wave field synthesis (WFS). This paper presentsresults of research that has been carried out to in-vestigate the combination of WFS with 2-D videoprojection, including subjective experiments onsound localization, correspondence of perceivedauditory, and visual source directions and speakeridentification in situations with multiple speakers, aswell as speech intelligibility tests, investigations onthe applicability of distributed mode loudspeakers inWFS and coloration artifacts due to discretization ofthe loudspeaker array.Convention Paper 5801

15:00 h

K-4 Comparison of Quality Degradation EffectsCaused by Limitation of Bandwidth and byDown-Mix Algorithms in Consumer MultichannelAudio Delivery Systems—Slawomir K. Zielinski1,Francis Rumsey1, Søren Bech21University of Surrey, Guildford, UK2Bang & Olufsen, Struer, Denmark

The comparative effect on audio quality of con-trolled multichannel audio bandwidth limitation andselected down-mix algorithms was examined. Theinvestigation focused on the standard 5.1 multi-channel audio set-up (Rec. ITU-R BS.775-1) andwas limited to the optimum listening position. Theobtained results indicate that in case of multichan-nel audio systems spatial quality is less importantthan timbral quality for typical program material.Convention Paper 5802

15:30 h

K-5 Temporal Aspects of Listener Envelopment inMultichannel Surround Systems—Gilbert A.Soulodre, Michel C. Lavoie, Scott G. Norcross,Communications Research Centre, Ottawa, Ontario, Canada

It has been shown that listener envelopment (LEV)can be systematically controlled in a multichannel sur-round system by varying the level and angular distrib-ution of the late-arriving sound. While the perceptual



transition point between early and late energy has tra-ditionally been set to 80 ms when predicting LEV, thismatter has not been rigorously investigated. In thepresent study a series of formal subjective tests wereconducted to investigate the perceptual point wherethe early energy ends and the late energy begins. Lis-teners were asked to rate the amount of LEV insound fields where the temporal and spatial distribu-tions of the late energy were varied. The results of thesubjective tests were used to investigate suitable ob-jective measures for predicting LEV.Convention Paper 5803

16:00 h

K-6 Single DSP for Audio Stream Control and Manipulation in Multichannel Receiver—SteveJahnke, Texas Instruments, Tokyo, Japan

Current multichannel audio receivers use a sepa-rate digital signal processor and microcontroller tomanipulate audio stream data and control the audioflow. The biggest challenge in these systems is co-ordinating the interaction between the DSP andMCU to create a high quality audio product. Fur-thermore, with the use of two separate programma-ble devices, software development time, system de-bug time, and total product component costs are increased. This development effort stresses verylimited design resources for the receiver manufac-turer, and allows only for an evolutionary change inproduct offering between generations. Convention Paper 5804

Session L Monday, March 24 13:30–17:30 hRoom E/F

ANALYSIS AND SYNTHESIS OF SOUND

Chair: Mark Sandler, Queen Mary, University of London, London, UK

13:30 h

L-1 Applying Physical Modeling, Set Theory, and Cellular Automata to Create Computer Synthesized Musical Compositions—SigmundRothschild, University of Colorado at Denver, Denver, Colorado, USA

Spectral analysis of a physically modeled audio sam-ple can be related by harmonic spectrum to pitch in-tervals. Pitch class sets, derived from pitch intervals,may be selected that have a significant interval classvector relationship to the harmonic spectrum which aphysically modeled sound displays. A software-based granular synthesizer using cellular automatacomputer modeling techniques has been used toevolve selected pitch class sets over time. Macro-ar-chitectural musical structures are created by combin-ing multiple pitch class set evolutions. Audio andvideo examples of pitch class sets being evolved bycellular automata will be presented. This paper fo-cuses on the application, integration, and implica-tions of these processes, as the mathematical proce-dures are well documented in other sources.Convention Paper 5805

14:00 h

L-2 Automatic Labeling of Unpitched PercussionSounds—Perfecto Herrera1, Amaury Dehamel2,Fabien Gouyon1

1Universitat Pompeu Fabra, Barcelona, Spain2École Nationale Supérieure des Télécommunications, Paris, France

We present a large-scale study on the automaticclassification of sounds from percussion instru-ments. Different subsets of temporal and spectraldescriptors (up to 208) are used as features thatseveral learning systems exploit to learn class parti-tions. More than thirty different classes of acousticand synthetic instruments and almost two thousanddifferent isolated sounds (i.e., not mixed with otherones) have been tested with ten-fold or holdoutcross-validation. The best performance can beachieved with Kernel Density estimation (15 per-cent of errors), although boosted rule systemsyielded similar figures. Multidimensional scaling ofthe classes provides a graphical and conceptualrepresentation of the relationships between soundclasses and facilitates the explanation of sometypes of errors. We also explore several options toexpand the sound descriptors beyond the class la-bel, as for example the manufacturer-model labeland confirm the feasibility of doing that. We finallydiscuss methodological issues regarding the gener-alization capabilities of usual experiments that havebeen done in this area.Convention Paper 5806

14:30 h

L-3 Using a Physiological Ear Model for AutomaticMelody Transcription and Sound Source Recognition—Thorsten Heinz, Andreas Brückmann,Fraunhofer IIS AEMT, Ilmenau, Germany

Recent trends in musical audio signal analysis in-creasingly promote use of perceptually motivatedmodifications of conventional signal processing al-gorithms. A consequent further step consists of theinclusion of the knowledge of the structure of mam-malian auditory periphery. The presented paperuses physiological models in order to mimic activefunctionality of the inner ear including the transduc-tion from mechanical vibrations into neural impuls-es. The main part of the paper describes automatictranscription of melodies from real world musical in-puts. Bottom-up extraction and segmentation ofpitch trajectories based on the outputs of the usedmodels, i.e., concentration of transmitter substanceinside the inner hair cell clefts, are demonstrated.As an example for the wide range of possible fur-ther applications, a sound source recognition ap-proach using woodwind instruments is proposed.Results indicate that the algorithm performs excel-lent compared to traditional methods.Convention Paper 5807

15:00 h

L-4 Prior Subspace Analysis for Drum Transcription—Derry FitzGerald1, Bob Lawlor2, Eugene Coyle11Dublin Institute of Technology, Dublin, Ireland2National University of Ireland, Maynooth, Ireland

This paper introduces the technique of prior sub-space analysis (PSA) as an alternative to indepen-dent subspace analysis (ISA) in cases where priorknowledge about the sources to be separated isavailable. The use of prior knowledge overcomessome of the problems associated with ISA, in par-ticular the problem of estimating the amount of in-formation required for separation. This results in im-


proved robustness for drum transcription purposes.Prior knowledge is incorporated by use of a set ofprior frequency subspaces that characterize fea-tures of the sources to be extracted. The effective-ness and robustness of PSA is demonstrated by itsuse in a simple drum transcription algorithm.Convention Paper 5808

15:30 h

L-5 Real Time Object-Based Coding—Paul M.Brossier, Mark B. Sandler, Mark D. Plumbley,Queen Mary, University of London, London, UK

This paper describes the design of a real-time MP4structured audio codec for monophonic signals. Thecoding of the live input consists of a pitch detectionsystem which returns the MIDI-like data, and an ad-ditive synthesis scheme which creates and modifiesthe current instrument. Both parts are designed tobe fast and scalable. The analysis parameters letthe user choose the computational cost for both theanalysis and the resynthesis. The extracted objectscan then be used in live environments for encodingand/or creation.Convention Paper 5809

16:00 h

L-6 Using and Enhancing the Current MPEG-7 Standard for a Music Content Processing Tool—Emilia Gómez, Fabien Gouyon, Perfecto Herrera,Xavier Amatriain, Universitat Pompeu Fabra,Barcelona, Spain

The aim of this paper is to discuss possible ways ofdescribing some music constructs in a dual context.First, that of the current standard for multimediacontent description: MPEG-7. Second, that of aspecific software application, the Sound Palette (atool for content-based management, content edi-tion, and transformation of simple audio phases).We discuss some MPEG-7 limitations regarding dif-ferent musical layers: melodic (present but under-developed), rhythmic (practically absent), and in-strumental (present though using an exclusiveprocedure). Some proposals for overcoming themare presented in the context of our application.Convention Paper 5810

16:30 h

L-7 Determination of the Meter of Musical AudioSignals: Seeking Recurrences in Beat SegmentDescriptors—Fabien Gouyon, Perfecto Herrera,Universitat Pompeu Fabra, Barcelona, Spain

We address the problem of classifying polyphonicmusical audio signals by their meter: the number ofbeats between regularly recurring accents (ordownbeats). The problem is simplified to a "double"/"triple" decision. Experiments have been conduct-ed on a 70-instance database (20 excerpts frompieces of music without particular genre nor timbrerestriction). Our approach aims to test the hypothe-sis that acoustic evidences for downbeats can bemeasured on signal low-level features, focusing es-pecially on their temporal recurrences. We experi-mented with several approaches to the problem offeature selection and report some interesting re-sults: measurements of a very small set of beat de-scriptors (i.e., 4) and subsequent processing(based on autocorrelation functions) allow about 95percent correct classification. Using only the tempo-

ral centroid, almost 90 percent of correct classifica-tion can be achieved.Convention Paper 5811

17:00 h

L-8 Temporal Segmentation and Pre-Analysis forNonlinear Time-Scaling of Audio—Chis Duxbury,Mark Sandler, Mike Davies, Queen Mary, University of London, UK

We present a method to achieve good segmenta-tion of note events for use with nonlinear time scal-ing algorithms, greatly reducing artifacts due to bothrhythmic distortions and soft note transitions beingtreated as percussive transients. The proposed al-gorithm isolates percussive transients as a subsetof note-onsets, leading to a more meaningful seg-mentation. A subband based hybrid onset detectionalgorithm forms the basis of this segmentationscheme. A new frequency content distance mea-sure, automatic threshold setting, and subband re-sult validation are all key elements of this scheme.At the subband re-combining stage the algorithmdifferentiates between note onsets, which may ap-pear in one or more subbands, from percussivetransients that appear in multiple subbands.Convention Paper 5812

Workshop 9 Monday, March 24 13:30–16:30 hRoom B

CORRELATION BETWEEN SUBJECTIVE AND OBJECTIVE MEASUREMENT FOR AUTOMOTIVESOUND SYSTEMS

Chair: Richard Stroud, Stroud Audio, Inc., Kokomo,IN, USA

Panelists: David L. Clark, DLC Design, Wixom, MI, USARoger Kessler, BMW, GermanyWolfgang Klippel, Harman-Becker, GermanyTom Nousaine, DLC Design, Wixom, MI, USA

Vehicle manufacturers and their audio suppliers areseeking objective measurement as the audio system’sevaluation method of choice. Correlation between provensubjective evaluation methodologies and objective mea-surements must be strong to permit the latter to be suc-cessful. This workshop will discuss current and proposedstates of correlation between subjective evaluation andobjective measurement for automotive sound systems.

Seminar 8 Monday, March 24 13:30–15:30 hRoom A

WORKING WITH MICROPHONES: A Practical Review

Presenter: Ron Streicher, Pacific AV Enterprises, CA, USA

The focus of this seminar will be a hands-on demonstra-tion of many of the practical aspects of using micro-phones: mounting hardware, shock isolation, pop-screens, cables, powering systems, etc. Techniques forrigging or “flying” microphones and arrays also will bepresented.

What will not be discussed is how or where to put amicrophone for the best pickup of [insert your favorite in-strument here.] That is an entirely different tutorial ses-sion. However, once you've chosen the microphone andits location, if you want to know how to get the micro-phone into that position most effectively and obtain opti-mum performance—free from intrusive mechanical nois-es, wind pops or blasts—this seminar is for you.




DVD-AUDIO: EXPLORING THE FORMAT

DVD-AudioPresenters: Craig Anderson, Warner

John Bamford, PioneerBob Michaels, 5.1 EntertainmentDavid Fraser, Dolby Stanley Lipshitz, University of WaterlooJohannes Muller, Mediarte Bob Stuart, Meridian Mark Waldrep, AIX Records

A practical and jargon-busting introduction to all aspectsof this exciting new format. Topics covered will include atechnical overview of the format; the audio coding; copyprotection and watermarking; title production, masteringand authoring; production tools; and a summary of thecurrent status of hardware, software, distribution andmarketing.

If you are considering any aspect of DVD-Audio disc orplayer creation then this seminar will be the perfect placeto learn more. There will be presentations by a numberof leading experts from the US, UK, Germany, andCanada. The seminar will include illustrative demonstra-tions of the audio and video features of DVD-Audio.


BASS CREATOR FOR ACTIVE NOISE CONTROL

AMPCOPresenter: Peter van der Geer

Noise pollution caused by music is an increasing prob-lem in the modern western world. Sound insulation is avery expensive solution and often the results are not sat-isfying. Modern acoustical DSP techniques can offer amore cost-effective way of solving this problem. In thisseminar the principle is explained, the way it is applied,and in which circumstances it can be used. A livedemonstration of the equipment will be given.

14:00 h Room 1Standards Committee Meeting on SC-04-01Acoustics and Sound Source Modeling


LOUDSPEAKERS FOR SOUND MASKING—EXPLORETHE NATURE OF NOISE IN OPEN PLAN, AND LEARNGUIDELINES FOR DEVELOPING AN INTEGRATED DESIGN SOLUTION

AMPCOPresenters: Morgan Sousa, Armstrong Building

Products Peter van der Geer, Ampco/Flashlight Projects

Seventy percent of office space is currently being de-signed as Open Plan Office, despite research showingthat noise can reduce worker productivity. In the USA itis common practice to use sound masking in officespaces. In Europe this principle and its application ishardly known. This seminar will focus on the applicationof sound masking in offices. The concept of improvingspeech privacy by introducing noise will be explained.The advantage of using flat panel loudspeakers instead

of conventional loudspeakers, the effect on buildingcosts, and the improvement of worker productivity will bediscussed.

Historical PresentationHISTORY OF MECHANICAL MUSICMonday, March 24 15:00–16:00 h

Pianolas, music boxes, and other mechanical music arti-facts were the forerunners of recorded music back in thetwenties. Kees Nijssen, an avid collector of these spe-cial automated pianos and other mechanical instruments,will give an overview of their history.

Workshop 10 Monday, March 24 15:30–17:30 hRoom A

AES47 ATM STANDARD

Chair: Mark Yonge, AES Standards Manager

Presenters: David Errock, BBC Technology, London, UKPaul Grant, Glensound Electronics, Maidstone, Kent, UKMike Story, Data Conversion Systems, Great Chesterford, Saffron Walden, UK

Standards are vital to ensure interoperability in systemsusing equipment from multiple vendors. The AES recentlypublished a new standard for connecting professional digi-tal audio over ATM networks, AES47-2002. AES47 allowsconventional structured cabling in buildings to carry liveaudio signals. The network inherently provides powerfulswitching and routing and also supports multiple samplingfrequencies in the same fabric. Unlike conventional com-puter networks, AES47 connections do not incur substan-tial time delays and can be used comfortably in live broad-cast operations, for example. Building a practical systemusing networked audio on a large scale opens up someradical new possibilities. This workshop will explore arange of key issues, such as failsafe circuit protection andflexible sync distribution, using experience gained during acurrent project development for BBC Radio.


WMS 4000—THE VISIONARY WIRELESSMICROPHONE SYSTEM

AKG Acoustics GmbHPresenter: Mario Siokola

The WMS 4000 brings in a new era of wireless technolo-gy. The WMS 4000 is one of the most advanced and in-novative wireless systems for professional use availabletoday. It offers an inspiring combination of sophisticatedengineering, exceptional flexibility, surprising versatility,and maximum ease of use. The application range of theWMS 4000 is almost unlimited, from fixed installations ormobile systems in architecturally difficult environments tomotion picture and broadcast work. In large-scale pro-jects including worship center sound systems, theatricalproductions, or stadium concerts, the WMS 4000 showsoff its professional qualities.

Session Y Monday, March 24 16:00–18:00 hTopaz Lounge

(POSTERS) COMPUTER AUDIO AND NETWORKS


16:00 h

Y-1 Protecting Digital Media with End-to-End Encryption—Jürgen K. Lang, m-sec, c/o ASTORIT Consulting, Gladbach, Germany

A method of protecting digital media from unautho-rized copying is introduced. Comprehensive end-to-end encryption can be applied to digital audio orvideo content of any bandwidth. Protected mediacan be distributed over traditional channels such asphysical media or broadcasting as well as over theInternet. The system is flexible and easy to use andis intended to meet business and market requirements.Convention Paper 5813

16:00 h

Y-2 Testing Challenges in Personal Computer Audio Devices—Wayne Jones1, Michael Wolfe1,Theodore Tanner Jr.2, Daniel Dinu21Audio Precision, Inc., Beaverton, OR, USA2Microsoft Corp., Redmond, WA, USA

The personal computer (PC) audio environment hasevolved over the years to become a significant entitywithin the field of acquisition and rendering of audioinformation. The personal computer is a highly so-phisticated interactive environment that is muchmore complex than a conventional dedicated homeaudio device, leading to new problem areas. Theseinclude, but are not limited to, stochastic interrupts,network accesses, disc I/O, and disparate hardwarequalities. While the environment of a highly matrixedmultitasking concurrent operating system offersmany opportunities to overcome quality issues, thePC, due to the media-rich tools and feature sets, isbecoming the entertainment capture and renderingdevice of choice for future generations. Many of thequality issues have been focused on hardware, suchas converter quality, power supply quality, and com-ponent metrics. We will be focusing on software per-formance metrics which are, by definition, muchmore difficult to ascertain. We will address conven-tional audio measurements such as distortion, fre-quency response and signal-to-noise ratio, but willextend these to new depths and address the uniquedifficulties the PC environment adds to these tests.The tests will also include “glitch verification,” though-put latency, and MIDI latency. Convention Paper 5814

16:00 h

Y-3 Remote Interface for Audio Devices (RIAD)—Nuno Fonseca1, Edmundo Monteiro21Polytechnic Institute of Leiria, Leiria, Portugal2University of Coimbra, Coimbra, Portugal

Although many audio devices have the capabilitiesto be remotely controlled (e.g., MIDI, RS-232, andproprietary systems), usually this control is orientedto automation or computer control and is by manyspecific parameters, which usually don’t support allthe needed control features and normally have userinterface problems, forcing a directed manual inter-vention in the equipment. To resolve these prob-lems we propose RIAD—Remote Interface for Au-dio Devices. RIAD allows remote access to thedevices’ interface (e.g., buttons, displays, LEDs,and knobs) allowing users to have remotely fullcontrol of audio equipment.Convention Paper 5815

16:00 h

Y-4 VOTA Utility: Making the Computer Sing—NunoFonseca, Polytechnic Institute of Leiria, Leiria, Portugal

After several years trying to emulate real musicalinstruments, it is time to emulate a real choir butwith a new challenge—text. This paper presents asoftware solution (based on sampling, not on syn-thesis) to make the computer sing text. You writethe text, play the keyboard (or any MIDI device orsequencer) and hear the choir singing in real-time.Convention Paper 5816

16:00 h

Y-5 Rendering of Advanced 3D Room Models by Enhanced Application Programming Interfaces—Aleksandar Simeonov, Giorgio Zoia, Robert-LluisGarcia, Signal Processing Institute, Lausanne,Switzerland

The geometrical and perceptual room models for 3-D audio production and rendering, proposed by no-table tools and more recently supported by theMPEG-4 standard, are compact and satisfactorydescription schemes for a wide range of audio andmultimedia applications. Support of both approach-es at the same time is a challenging task, especiallywhen the requirements are high quality, precisesynchonization with other media and acceptable la-tency to user interaction, as it often happens instandardized contexts for media integration. In thispaper we present first the results from several ex-periences with different application programming in-terfaces and hardware platforms; suitable exten-sions to support physical and perceptual modelsare then described, and their integration into anMPEG-4 compliant player is presented.Convention Paper 5817

16:00 h

Y-6 Reversible Audio Watermarking—Michiel van der Veen, Arno van Leest, Fons Bruekers, PhilipsResearch, Eindhoven, The Netherlands

In this paper we present a fragile and high capacitywatermarking technique for digital audio signals.The watermark itself is reversible, which in this con-text refers to the ability to restore the original inputsignal in the watermark detector. In summary, theapproach works as follows. In the encoder, the dy-namic range of the input signal is limited, the signalbits are re-ordered, and a part of the unused bitsare deployed for encoding the watermark. Anotherpart of these bits are used to convey information forthe bit-exact reconstruction of signal. It is the pur-pose of the watermark decoder to extract the water-mark and reconstruct the input signal by restoringthe original dynamic range. In this study we exten-sively tested this new algorithm with a variety ofsettings using audio items with different characteris-tics. These experiments showed that for 16-bit PCMaudio sampled at 44.1 kHz, capacities close to44000 bits per second can be achieved, while per-ceptual degradation of the watermarked signal re-mained acceptable.Convention Paper 5818

16:00 h

Y-7 Implementation of MPEG-4 Audio Nodes in anInteractive Virtual 3D Environment—Andreas



Dantele, Ulrich Reiter, Michael Schuldt, HelgeDrumm, Oliver Baum, Technische Universität Ilme-nau, Ilmenau, Germany

We show how to use MPEG-4 audio nodes in an in-teractive virtual 3-D scenery to improve scene real-ism in the auditory domain. Therefore we extendthe virtual reality modeling language (VRML) be-cause of it’s similarity to the MPEG-4 scene de-scription. In addition to the implementation of local-ized sound sources in the scenery, the effect ofacoustic obstruction is discussed. Several possibili-ties for the detection of obstruction are presented.The results demonstrate the capabilities of MPEG-4audio scene description and thus point out the needfor a fully compliant MPEG-4 player to utilize thecomplete functionality of this standard.Convention Paper 5820

16:00 h

Y-8 Robust Real-Time Identification of PCM AudioSources—Frank Kurth, Roman Scherzer, Universityof Bonn, Bonn, Germany

We propose both a framework and a system forcomparing PCM audio streams to multiple candi-date audio streams. This framework allows for areal-time identification of the audio streams w.r.t.several candidate streams. The identification is ro-bust to signal delays of up to several seconds aswell as to signal distortions due to lossy coding, anoisy environment, or analog transmission. An areaof application is the query-by-mobile-phone sce-nario where a user transmits an audio streamrecorded from the radio using his mobile phone asa recording device. The transmitted audio streammay be identified using the proposed system. Forthis, the system matches a compressed version ofthe stream to compressed versions of all possibleradio programs in real-time.Convention Paper 5821

Student ActivitiesRECORDING COMPETITIONMonday, March 24, 16:00–19:00 hRoom L

Finalists selected by an elite panel of judges will givebrief descriptions and play recordings in the different cat-egories. There will be one submission per category perschool/student section. Meritorious awards will be pre-sented at the closing Student Delegate Assembly Meet-ing on Tuesday.

16:40 h Room BTechnical Committee on Automotive Audio

16:40 hRoom C/DTechnical Committee on Multichannel and BinauralAudio Technologies


THE WORLD BEYOND 20 KHZ—EARTHWORKS’MODEL OF HUMAN HEARING

EarthworksPresenter: Eric Blackmer

Earthworks proposes that the old 20 Hz to 20 kHz modelof human hearing does not explain what humans, in fact,

can hear. That is only the range we can hear as tones.We suggest that time domain information is an importantaspect of human hearing, and that human resolution intime, in order of magnitude, is finer than our ability tohear tones. We will show that audio tools with a fastclean time domain response and a flat extended frequen-cy response allows you to capture and reproduce sonicevents with a remarkable degree of realism. We willdemonstrate the wonders of time-accurate playback withour new Sigma 6.3 reference monitors.

17:40 h Room ATechnical Committee Meeting on Transmission andBroadcasting

17:40 h Room E/FTechnical Committee Meeting on Audio Analysis and Representation (Formative Meeting)

Session M Tuesday, March 25 09:00–13:00 hRoom C/D

SIGNAL PROCESSING, PART 1

Chair: Peter Eastty, Sony Broadcast and Professional Research Labs, Oxford, UK

09:00 h

M-1 Controlled Pre-response Antialias Filters forUse at 96 kHz and 192 kHz—Peter G. Craven,Consultant to Meridian Audio Limited, Huntingdon,Cambridgeshire, UK

Antialias and reconstruction filter design at ordinarysample rates has been a compromise—minimumphase designs suffer from group delay variationwhile linear phase designs suffer from pre-respons-es. Increasing the sample rate makes the tradeoffeasier. The paper presents some 96 kHz and 192kHz designs optimized simultaneously for good fre-quency domain performance over the conventionalaudio band 0 to 20 kHz and good time domain per-formance when viewed wideband. We also addressthe question of the cascading of multiple filters in acomplete recording and reproduction chain.Convention Paper 5822

09:30 h

M-2 Design Techniques for High-Performance Discrete A/D Converters—Bruno Putzeys, PhilipsDigital System Labs, Leuven, Belgium

An old 1-bit D/A converter idea using voltageswitching is reintroduced. Some shortcomings ofthe original design are discussed. A new design isproposed that addresses these issues. A one-bitnoise shaping ADC with DSD specifications is builtaround this circuit serving as the feedback D/A.Theoretical noise performance and real perfor-mance are shown and compared. A road map forimproving the present design is outlined. It is con-cluded that this conversion technique is much morepromising than commonly thought.Convention Paper 5823

10:00 h

M-3 Time-Scale Modification of Speech Using a Synchronized and Adaptive Overlap-Add (SAOLA) Algorithm—David Dorran1, Bob Lawlor2,


Eugene Coyle11Dublin Institute of Technology, Dublin, Ireland 2National University of Ireland, Maynooth, Ireland

The synchronized overlap-add (SOLA) algorithm isa commercially popular and considerably re-searched audio time-scale modification technique.It operates in the time domain and uses a correla-tion technique to ensure that synthesis frames over-lap in a synchronous manner. We present a modifi-cation to SOLA that allows the analysis step sizeadapt to the desired time-scale factor. The synchro-nized and adaptive overlap-add (SAOLA) algorithmimproves upon the output quality of SOLA for hightime-scale factors and reduces the computationalrequirements for low time-scale factors. However,the computational requirements for high time-scalefactors are increased.Convention Paper 5824

10:30 h

M-4 Tree-Based Lookahead Sigma Delta Modulators—James A. S. Angus, University of Salford, Salford,Greater Manchester, UK

In a recent paper (“Trellis Noise-Shaping Convert-ers and 1-bit Digital Audio,” presented at the AES112th Convention, Munich, Germany), Kato intro-duced the concept of trellis noise-shaping ∑-∆ mod-ulators. These modulators look forward at k sam-ples of the signal before deciding to output a "one"or a "zero." The Viterbi algorithm is then used tosearch the trellis of all the possibilities that such aprocedure generates. This paper describes an al-ternative algorithm, a tree-based algorithm, whichcan also be used to search the possibilities gener-ated by looking ahead at noise-shaping ∑-∆ modu-lators. Tree-based algorithms are simpler to imple-ment because they do not require any backtrackingthough an array of scores in order to determine thecorrect output value.Convention Paper 5825

11:00 h

M-5 Time-Frequency Analysis, Modeling and Equal-ization of Room Impulse Response Functions—John C. Sarris, George E. Cambourakis, NationalTechnical University of Athens, Athens, Greece

Rooms are modeled as linear time invariant sys-tems, where the room impulse response (RIR) de-scribes the transmission characteristics for a specif-ic source receiver pair. Concepts of time-frequencyanalysis are used to decompose the RIR in thetime-frequency domain, where parametric modelsare employed to model the different subband sig-nals. The evaluated models perform exact modelingof the early reflections, whereas the decay rate ofthe reverberant part is sufficiently approximated.Equalization is performed in the time-frequency do-main, where someone can selectively equalize fre-quency subbands. Two different rooms are studiedas example cases.Convention Paper 5826

11:30 h

M-6 A Class of Sampling Rate Converters with Inter-esting Properties—Sigmar Ries, FachhochschuleSüdwestfalen, Meschede, Germany

A general class of sampling rate converters for con-version between arbitrary sampling rates is present-ed. The performance of these converters can be de-scribed by simple formulas that show how to trade offmemory consumption versus computational complex-ity. Especially, converters with optimum properties inview of computational complexity are presented. Pos-sible applications of these converters are pitch shift-ing or correction as well as sampling rate conversionby digital systems with memory limitation.Convention Paper 5827

12:00 h

M-7 Feedback Strategies in Digitally ControlledClass-D Amplifiers—Rolf Esslinger1, GerhardGruhler1, Robert Stewart21University of Applied Sciences, Heilbronn, Germany2Signal Processing Division, Dept. of Electronic and Electrical Engineering, Glasgow, Scotland, UK

A fundamental problem in digitally controlled class-D power amplifiers is the distortion of the amplifiedpulse signal by imperfections of the power transis-tors and the analog output circuitry. The only way toovercome this is to provide error correction by feed-back. With a pulse-modulated signal (either PWMor sigma-delta-modulation) this correction can beperformed by changing the pulse edge timing, as itis done in some existing solutions. If error correc-tion during the pulse generation inside of the digitalsystem is desired, an analog-to-digital converter isneeded. In this paper the most important solutionsdone so far are reviewed and further ideas basedon feedback into the digital system are introduced.Convention Paper 5828

12:30 h

M-8 Audio Signal Processing Using ComplexWavelets—Patrick J. Wolfe, Simon J. Godsill, University of Cambridge, Cambridge, UK

In this paper we present an overview of complexwavelets and discuss their application to audio sig-nal processing. While traditional audio processinghas involved time-frequency rather than time-scalerepresentations, we demonstrate some of the ad-vantages to be gained though the use of complexwavelets. We focus on two main applications of in-terest to the audio community: noise reduction andsignal compression.Convention Paper 5829

Session N Tuesday, March 25 09:00–11:30 hRoom E/F

LOW BIT-RATE CODING, PART 1

Chair: Marina Bosi, MPEG LA, Denver, CO, USA

09:00 h

N-1 Original File Length (OFL) for mp3, mp3PRO,and Other Audio Codecs—Ernst F. Schroeder, Johannes Boehm, Thomson, Corporate Research,Hannover, Germany

When audio signals are encoded and decoded withtypical "loss" data compression codecs, then it is ex-pected that the decoded audio signals are almostnever identical to the original signals. What is oftenoverlooked is that the decoded audio signals are



also typically no longer time-aligned with the originalsignals. It is shown how these small timing errorsare due to the block structure of data processingand to the look-ahead needed for deciding betweendifferent coding strategies. For audio codecs basedon the ISO/IEC MPEG standards a solution is intro-duced which restores the original timing. This OFLfeature is available in the latest implementations ofthe mp3 and mp3PRO audio codec.Convention Paper 5830

09:30 h

N-2 MDCT Analysis of Sinusoids and Applicationsto Coding Artifacts Reduction—Laurent Daudet1,Mark Sandler21Université Paris, Paris, France2Queen Mary, University of London, London, UK

Due to the non-invariance of all real time-frequencytransforms though time shifts, simple subsequentactions such as coefficients thresholding and quan-tizing act differently on different parts of stationarysignals. In the context of audio coding, this is oneimportant cause of artifacts (called warbling arti-facts or "birdies"), which the human hearing is verysensitive to. In this paper we analyze these time de-pendencies for the modulated discrete cosine trans-form (MDCT) and propose a simple regularizationprocedure that may be used to suppress most ofthe partials that exhibit such artifacts.Convention Paper 5831

10:00 h

N-3 They Exist: Limit Cycles in High Order SigmaDelta Modulators—Joshua D. Reiss, Mark B. Sandler, Queen Mary, University of London, London, UK

A compact form can be used to describe an arbi-trary high order sigma-delta modulator. This pro-vides insight into the structure of limit cycles in sig-ma-delta modulators. We consider modulators ofany order with periodic output. We make no as-sumptions regarding the input and are thus able toprove necessary conditions for limit cycles in theoutput. We show that the input must be periodic butmay have a different period from both integratoroutput and quantized output. We derive what thisimplies regarding limit cycles for sinusoidal inputs.Finally, we give examples where sinusoidal input toa third order modulator results in a limit cycle of adifferent frequency.Convention Paper 5832

10:30 hN-4 Implementation of a DRM Audio Decoder (aacPlus)

on ARM Architecture—Thomas Ziegler1, DanielHomm1, Reinhold Böhm1, Robert Weidner21Coding Technologies, Nürnberg, Germany2DSP Solutions, Regensburg, Germany

Digital radio mondial (DRM) is an ETSI standard forthe digital broadcast in the frequency range below 30MHz. Full bandwidth audio in a CD near stereo quali-ty is maintained by using aacPlus, the combination ofspectral band replication (SBR) technique in conjunc-tion with MPEG-4 aac. The goal of the Eureka projectE!2390-DIAM (Digital AM) is the development of aDRM receiver chipset which will be available by theend of 2003. Within the digital part of that chipset, theaacPlus decoder is currently implemented on a regu-

lar ARM core without any hardware accelerator units.The use of a typical micro controller architecturewidely used in mobile applications such as cellulartelephones, PDAs, and smartphones proves the real-world capability of the SBR technology.Convention Paper 5833

11:00 h

N-5 State-of-the-Art Audio Coding for Broadcastingand Mobile Applications—Andreas Ehret, MartinDietz, Kristofer Kjörling, Coding Technologies,Nürnberg, Germany

This paper will discuss the combination of the newapproach of the Spectral Band Replication (SBR)technology with the leading conventional waveformaudio coder standardized in MPEG, advanced au-dio coding (AAC). With this enhanced audio codingscheme, named aacPlus, it is possible to achievehigh-quality audio in stereo at bit rates as low as 40kbit/s stereo. It is thus especially interesting for ap-plications where highest compression efficiency isdesired for the reasons of cost or physical limita-tions, such as digital broadcasting or mobile appli-cations. An overview on the latest development withrespect to the standardization process of aacPluswithin MPEG-4 and subjective verification resultsare also given.Convention Paper 5834

Workshop 11 Tuesday, March 25 09:00–12:00 hRoom B

LARGE ROOM ACOUSTICS—PROBLEMATIC CASE STUDIES

Chairs: Jan Voetman, DELTA Acoustics and Vibration, Lyngby, Denmark Peter Mapp, Peter Mapp Associates, Colchester, UK

Panelists: Michael Barron, Fleming and Barron, UKDurand Begault, Charles M. Salter Associates Inc.,San Francisco, CA, USAJim Brown, Audio Systems Group, Chicago, IL, USA

The difference between large rooms and small rooms isnot only a matter of volume (no joke!). Acoustically, largerooms are characterized by a high degree of diffuseness,a huge number of closely spaced room modes (Eigen-tones), and often excessive reverberation time. Thisworkshop will open with a tutorial by Michael Barron onbasic issues relating to large room acoustics, followed bypresentations by specialists working in the design ofacoustics and loudspeaker systems in spaces such asmultipurpose halls, sports stadiums, and concert halls.They will discuss research, current thinking, and goodpractice concerning reverberation, early reflections, andsuitable loudspeaker design.

Seminar 9 Tuesday, March 25 09:00–12:00 hRoom A

HOW TO SET UP 5.1 SURROUND

Presenters: Andrew Goldberg, Chistophe Anet, Genelec Oy, Iisalmi, Finland

A modern audio production facility must be able to serveproductions in a large number of different formats. Thechange from mono and stereo to multichannel reproduc-tion has produced many problems, both in converting ex-


isting production facilities to multichannel format and innew installations.

The audio formats that must be handled by a modernproduction facility include currently

— Mono, stereo— Matrix four channel format— Five channels (5.0 systems)— Five channels with a separate Low Frequency

Enhancement channel (5.1 systems)— Advanced multichannel formats such as 6.1, 7.1

and more.This seminar discusses multiple practical questions

about the monitoring loudspeakers, their set-up and pos-sible sources of problems, which should be avoided. Abrief overview of the current multichannel formats and adedicated section on bass management is also included.This seminar does not seek to explain monitoring loud-speaker design and technology.

09:00 h Room 1Standards Committee Meeting on SC-04Subcommittee on Acoustics

Exhibitor Seminar 24 Tuesday, March 25 10:30–11:30 hRoom L

EQUIPMENT FOR DSD CODING

SADiE UKPresenters: Geoff Calbver, Jim Gros

To meet the growing demand for more sophisticated pro-duction tools to create masters for the Super Audio CD(SACD) format, SADiE has introduced the new DSD8editor. DSD8 is capable of the full editing and masteringof DSD program material and incorporates full authoringfacilities for the SACD format to produce both Scarlet-Book cutting masters for SACD and Red-Book mastersfor CD. DST compression tools, necessary for the pro-duction of multichannel SACD masters, are fully integrat-ed, and up to eight channels of DSD material may behandled simultaneously. Today, we will explore the tech-niques involved in successful mastering for SACD.

Session Z Tuesday, March 25 11:00–13:00 hTopaz Lounge

(POSTERS) PSYCHOACOUSTICS AND PERCEPTION

11:00 h

Z-1 AQTtool an Automatic Tool for Design and Synthesis of Psychoacoustic Equalizers—Andrea Azzali1, Alberto Bellini1, Eraldo Carpanoni1,Marco Romagnoli2, Angelo Farina11University of Parma, Parma, Italy2ASK Industries, Reggio Emilia, Italy

Steady-state characterization of acoustic environ-ment is not enough for an efficient compensation ofresonance and distortions. On the other side, thecommon availability of digital systems has spreadthe use of acoustic equalizers at any level of soundreproduction. Therefore, a new concept of equaliza-tion needs to be defined. It relies on dynamic fre-quency response and on articulation to design theequalizer shape. Specifically, the inverse filtershape will be based on the dynamic frequency re-sponse instead of on the steady-state frequency re-sponse. This is obtained by relying on AQT meth-

ods, which use a variable frequency burst as astimuli. In this paper an automatic tool was devel-oped in order to obtain AQT parameters quickly,and to use them to synthesize a nice equalizationfilter shape. Moreover an automatic software toolwas realized, which allows performance of AQTmeasurement with a user friendly GUI, and allowsthe user to synthesize a nice equalizer, fixing a fewdegrees of freedom. The novel dynamic equaliza-tion method was especially tailored for inside a car.The equalizer was experimentally validated in a fewcommercial cars, using a DSP-based board.Convention Paper 5835

11:00 h

Z-2 Speech Intelligibility of the Call Sign AcquisitionTest (CAT) for Army Communication Systems—Mohan Rao1, Tomasz Letowski21Michigan Tech University, MI, USA2U.S. Army Research Laboratory, Aberdeen Proving Grounds, MD, USA

The specific objective of this study was to assessthe effects of noise on the intelligibility of speechelements used in the Callsign Acquisition Test(CAT) developed by the U.S. Army Research Lab-oratory. This study employed thee backgroundnoises (pink noise, white noise, and multitalkerbabble) with one of them (pink noise) presented atseveral levels. A group of 18 listeners between theages of 18 and 25 participated in the study. The re-sults of the study led to the following main conclu-sions. For all thee types of background noises, theintelligibility of the CAT sounds were close to 100percent for speech to noise ratios (SNRs) of zeroor higher. For SNRs below zero, the speech intelli-gibility decreased with decreasing SNR, with pinknoise having the steepest decline followed by mul-titalker babble and white noise. Next, the speechintelligibility did not appear to be affected, to anysignificant degree, by the level of pink noise usedin the study. Additional item analysis revealed thatthe CAT numbers and words yielded higher meanscores when analyzed separately than when ana-lyzed as an alphanumeric unit.Convention Paper 5836

11:00 h

Z-3 Acoustic Personalization of Mobile Phones—Suthikshn Kumar, Larsen & Toubro Infotech Limited, Bangalore, India

Every user of cellular telephones has different hearingrequirements. The individual’s hearing requirementschange over a period of time based on the prolongedexposure to background noise, age-related factors,hearing-related illness, etc. This paper proposes theacoustic personalization of the mobile telephonesbased on the hearing requirements of an individualfor improving speech quality. A smart acoustic vol-ume tuner which uses fuzzy logic to improve thespeech quality based on the audiogram is presented.The mobile telephone is personalized by designingthe fuzzy rule base to suit the individual’s hearing re-quirements. The smart volume tuner continuouslymonitors the background noise levels and classes toadjust the acoustic volume levels using fuzzy logic.Genetic algorithm is used for tuning the input/outputscaling factors, membership functions, and to opti-mize the fuzzy rule-base for personalization.Convention Paper 5837



11:00 h

Z-4 Modeling the Auditory Perception of Size, Shapeand Material: Applications to the Classification ofTransient Sonar Sounds—Simon Tucker, GuyBrown, University of Sheffield, Sheffield, UK

A psychophysical experiment was undertaken to in-vestigate whether human listeners are able to per-ceive the material properties of struck plates whenthey are suspended in air, and also when they areartificially damped by suspension in water. Listen-er’s judgments of the size, shape, and material ofthe plates were found to be less reliable in thedamped condition. A computational model was de-veloped which estimates the material properties ofan impulsive sound by measuring the decay rate ofsignificant acoustic components at the output of anauditory filter bank. The model provides a goodoverall match to the pattern of human responses inthe psychophysical study. The auditory model hasalso been used to classify transient sonar soundswith encouraging results.Convention Paper 5838

11:00 h

Z-5 The Importance of Reflections in a BinauralRoom Impulse Response—Rene E. Jensen1,Todd Welti21Siemens Mobile Phones A/S, Nørresundby, Denmark

2Harman International Industries, Inc., Northidge, CA, USA

A room impulse response constitutes a unique signa-ture of an acoustical space, however much of thefine detail is masked by the direct sound and highestlevel reflections. It is possible that reflections whichare substantially masked may be simplified withoutaudible effects. This simplification could speed upreal time signal processing applications which usethe room impulse response, such as auralization.Masked thresholds of audibility are measured for in-dividual reflections added to a measured binauralroom impulse response (BRIR). Measured thresh-olds are then used as a template to remove low levelreflections from the BRIR, and replace them with asimplified signal. Listening tests show that for voicesignals, altered BRIR’s are virtually indistinguishablefrom the original versions, even when 93 percent ofthe BRIR between 15 ms and 200 ms are replaced.[Paper not presented at convention, but ConventionPaper 5839 is available.]

11:00 h

Z-6 Use of Auditory Temporal Masking in the MPEG Psychoacoustic Model 2—Hossein Najaf-Zadeh,Hassan Lahdidli, Michel Lavoie, Louis Thibault, Advanced Audio Systems, Communications Research Centre Canada, Ottawa, Ontario, Canada

This paper presents a model of auditory temporalmasking for perceptual audio coders. As such, wehave developed a model to incorporate temporalmasking effects into the MPEG psychoacousticmodel 2. The enhanced psychoacoustic model ren-ders more accurate masking thresholds. Since themasking thresholds are used in adaptive bit alloca-tion, a better psychoacoustic model leads to higheraudio quality at a fixed bit rate. Semiformal listeningtests have shown that the incorporation of temporal

masking into the MPEG-1 Layer 2 encoder results ina reduction of 5 to12 percent in the average bit ratefor transparent coding.Convention Paper 5840

11:00 h

Z-7 A Sound Field Control Method Based on an Objective Measure of Spatial Impression—Yoshiki Ohta, Takashi Mitsuhashi, Shinji Koyano,Pioneer Corporation, Tsurugashima-city, Saitama,Japan

A method to control the spatial impression in soundreproduction within a small space has been devel-oped. First, psychological scales of spatial impres-sion were obtained though subjective evaluation ofsounds convolved with the impulse responses ofvarious rooms that differ in volume. Next, the rela-tionship between the psychological scale of spatialimpression and the physical features of the corre-sponding room impulse response was examined.We found that a psychological scale can be repre-sented by a linear combination of energy distribu-tion on a time-frequency plane calculated from animpulse response. We have developed a soundfield control method based on the objective mea-sure and tested the validity of this method thoughexperiments on an actual sound field.[Paper not presented at convention, but ConventionPaper 5841 is available.]

11:00 h

Z-8 Psychoacoustic Investigations on Sound-SourceOcclusion—Hania Farag1,2, Jens Blauert2, OnsyAbdel Alim11University of Alexandria, Alexandria, Egypt2Ruh-Universität Bochum, Germany

Efficient simulation of sound-source occlusion isneeded, e.g., in auditory virtual environments andremains an unsolved problem. In order to overcomethis problem, the changes in psychoacoustical pa-rameters accompanying the perception of sound-source occlusion have to be identified and under-stood. In this paper the impact of occlusion on thelocalization of auditory events is investigated withthe aid of listening tests. Rectangular wood platesof different dimensions are used as occluders. Anoticeable shift in the location of the auditory eventsis observed. The results can be explained on thegrounds of the precedence effect.Convention Paper 5842

Student ActivitiesSTUDENT POSTERSTuesday, March 25, 11:00–13:00 hTopaz Lounge

This event will display the scholarly/research works fromAES student members in the form of a poster presenta-tion. Unlike previous years, the student poster sessionwill now be held in the same space as the professionalposter sessions. This will ensure that the posters willreach a large audience, thus providing a great opportuni-ty to display and discuss the presented work with profes-sionals, educators, and other students.

Historical PresentationHISTORICAL OVERVIEW OF STEREOPHONYTuesday, March 25 11:00–12:00 h


Hans Lauterslager, Senior Recording Engineer, willplay and discuss a series of musical excerpts from aspecially-compiled CD. The excerpts illustrate the differ-ent recording techniques used in the early days ofstereophony and the results obtained.

11:30 h Room 1Standards Committee Meeting AESSC Plenary III

Exhibitor Seminar 25 Tuesday, March 25 12:00–13:00 hRoom L

AUDIO OVER IP

Mayah Communications GmbHPresenter: Joerg Rimkus, Uwe Flatter

The PROXIMA product family has been developed forapplications in the field of “In-House Streaming.” It differsdramatically from existing streaming solutions becausethe quality levels cover an extreme range. Supported areLayer 2, 3, AAC+SBR, mp3PRO, linear audio, apt-X, andothers. With the selectable audio coding algorithms it ispossible to cover completely different requirements andapplications in the broadcasting house simultaneously.The streaming server can send different formats simulta-neously to more than 1000 destinations, e.g., for pre-lis-tening, automation, and for archiving.

PROXIMA consists of a 19-inch, 2U-high streamingserver and audio gateway codec solution CENTAURI, aswell as the remote-controlled ethernet audio convertermodule Ganymed. The complete system can replace orexpand existing audio routing systems.

Session O Tuesday, March 25 13:30–17:00 hRoom C/D

SIGNAL PROCESSING, PART 2

Chair: John Mourjopoulos, University of Patras, Patras, Greece

13:30 h

O-1 A Single Chip AES3 Receiver and Transmitter andAsynchronous Sample Rate Converter Support-ing Sample Rates from 32 kHz to 192 kHz—KevinMcLaughlin, Kumaresh Bathey, Huaijin Chen, AnalogDevices Inc., Wilmington, MA, USA

A Single Chip AES3 receiver and transmitter withan asynchonous sample rate converter is present-ed that supports sample rates from 32 kHz to 192kHz. This single chip has a highly flexible architec-ture that allows the receiver, transmitter, and sam-ple rate converter to be independently intercon-nected among themselves and to two serial inputports and two serial output ports. The AES3 re-ceiver recovers the audio data, channel status,user bits, and clock with jitter less than 150 psfrom an incoming AES3 stream with sample ratesup to 192 kHz while complying with the AES3,AES11, and SMPTE337M standards. The asyn-chonous sample rate converter is capable of up-sampling by 1 to 8 and downsampling by 7.75:1while maintaining a minimum of 120dB THD+N.The AES3 Transmitter can transmit data at samplerates up to 192kHz and supports the AES3 andAES11 standards. The received and transmitted

channel status and user bits are buffered withoutthe use of arbitration logic for fast access though acontrol port.Convention Paper 5843

14:00 h

O-2 Comparison of Modal Equalizer Design Methods—Poju Antsalo1, Matti Karjalainen1, Aki Mäkivirta2,Vesa Välimäki11Helsinki University of Technology, Espoo, Finland2Genelec Oy, Iisalmi, Finland

Modal equalization of low-frequency room modeshas recently been proposed as a method to im-prove sound reproduction in spaces where modaldecay time is too long. Modal equalization isachieved by signal processing reducing the poleradii of problematic modes in the overall transferfunction. In this paper we compare the perfor-mance of two proposed methods for designingmodal equalizers. Comparison includes a prelimi-nary subjective listening test indicating a possiblemarginal improvement by modal equalization overconventional magnitude equalization.Convention Paper 5844

14:30 h

O-3 Efficient Trellis-Type Sigma Delta Modulator—Pieter Harpe1, Derk Reefman2, Erwin Janssen21University of Technology Eindhoven, Eindhoven, The Netherlands

2Philips Research, Eindhoven, The Netherlands

Recently, a new type sigma delta modulator (SDM),a Trellis noise-shaping converter, has been intro-duced as a 1-bit digital audio stream generator.This type has several advantages compared tostandard SDM, including better performance in sta-bility, signal to noise ratio, and linearity. A majordrawback of the Trellis architecture is the largeamount of computations and memory usage com-pared to the standard architecture. An efficient Trel-lis implementation will be introduced resulting in asignificant decrease of the number of computationsand memory with hardly any change in the convert-er’s performance.Convention Paper 5845

15:00 h

O-4 A New Digital-to-Analog Converter Design Technique for HiFi Applications—Derk Reefman1,John van den Homberg1, Ed van Tuijl1, Corné Bastiaansen2, Leon van der Dussen21Philips Research, Eindhoven, The Netherlands2Philips Semiconductors, Eindhoven, The Netherlands

A new digital-to-analog converter (in a .18 micronprocess) will be presented which exhibits excep-tionally large linearity (Total Harmonic Distortion <–115 dB at full scale input) and a very good signal-to-noise (SNR) performance (> 119 dB SNR). Tar-geted as a DAC to meet the Super Audio CD speci-fications, the DAC is designed as an n-bit (n = 4,5,or 6) switched current sigma delta modulator, run-ning at a sample rate of 2.8 or 5.6 MHz. While theSNR is obtained by standard but carefully applieddesign techniques, the linearity of the DAC is ob-tained by a new dynamic element matching (DEM)technique. Contrary to existing DEM approaches,the DEM technique fully removes any error due to



mismatch in the switched current sources, insteadof spectrally shaping the error. The new DEM tech-nique will be explained in detail, and measurementresults will be given.Convention Paper 5846

15:30 h

O-5 Discrete-Time Loudspeaker Modeling—AndrewBright, Nokia Group, Helsinki, Finland; TechnicalUniversity of Denmark, Lyngby, Denmark

Discrete-time models of loudspeaker dynamics arepresented. These have been developed to simplifydigital processing for active loudspeaker control.These discrete-time models differ from classicalloudspeaker models, which are all represented incontinuous time. The discrete-time models simplifydigital implementation of such aspects of activecontrol as parameter identification, equalization,and nonlinear distortion compensation. In this pa-per the fundamental theory of the development ofthe discrete-time models is presented, as well asits application to the problem of loudspeaker para-meter identification.Convention Paper 5847

16:00 h

O-6 Subjective Effects of Regularization on InverseFiltering—Scott G. Norcross, Gilbert A. Soulodre,Michel C. Lavoie, Communications Research Cen-tre, Ottawa, Ontario, Canada

Previous work has shown that inverse filtering of im-pulse responses (IR) can degrade the subjectivequality of audio signals in certain conditions. Theseverity of the degradation depends on both the re-sponse of the system that is being inverted and filterinversion method used to correct this response.Regularization has been proposed as a means toimprove the performance of inverse filtering by limit-ing how much "work" the inverse filter will do to cor-rect the system response. In this paper formal sub-jective tests were conducted to examine thesubjective effects of regularization on inverse filter-ing. The regularization techniques implemented in-clude frequency independent and dependent meth-ods as well as a perceptually-motivated method.The subjective tests were based on the ITU-RMUSHA method.Convention paper 5848

16:30 h

O-7 Audio Processing in the Wavelet Domain—David Darlington, Mark Sandler, Queen Mary, Universityof London, London, UK

Audio signals are often stored or transmitted in acompressed, transform domain representation. Thiscan pose a problem when there is a requirement toperform signal processing, in that it may be neces-sary to convert the signal back to a time domainrepresentation prior to processing, and then re-transform. This is time consuming and computation-ally intensive and may degrade the signal. It is thuspotentially more effective and efficient for the pro-cessing to be applied while the signal remains in thetransform domain. We have implemented a schemewhereby processing may be applied to signals storedin a wavelet domain representation without this im-plicit constraint.Convention Paper 5849

Session P Tuesday, March 25 13:30–15:30 hRoom E/F

LOW BIT-RATE CODING, PART 2

Chair: Jürgen Herre, Fraunhofer Institute for Integrated Circuits, Erlangen, Germany

Co-chair: Erik Larsen, University of Illinois at Urbana-Champaign, Urbana, IL, USA

13:30 h

P-1 Enhancing Audio Coding Efficiency of MPEGLayer-2 with Spectral Band Replication for DigitalRadio (DAB) in a Backwards CompatibleWay—Alexander Gröschel1, Michael Schug1,Michael Beer1, Fredrik Henn21Coding Technologies GmbH, Nürnberg, Germany2Coding Technologies AB, Stockholm, Sweden

Layer-2+SBR is an audio coding scheme which en-hances significantly the coding efficiency of MPEG1/2 Layer-2 especially for broadcasting applicationslike DAB (Digital Audio Broadcasting). Spectralband replication (SBR) is a technique to enhancethe efficiency of perceptual audio codecs. High fre-quency parts of an audio signal are reconstructedon decoder side, so the audio encoder can focus oncoding the low frequency part. Thus, a bit-rate re-duction can be achieved while maintaining subjec-tive audio quality. Besides increasing the coding ef-ficiency, the use of MPEG Layer-2 + SBR insideDAB would maintain backward compatibility: Exist-ing DAB receivers are capable of decoding the(bandwidth limited) Layer-2 part of the bitstream.This paper describes the functionality of SBR andLayer-2+SBR and the achievable increase in cod-ing efficiency. Furthermore, applications and intro-duction scenarios are addressed.Convention Paper 5850

14:00 h

P-2 Perceptual Optimization of the Frequency Selec-tive Switch in Scalable Audio Coding—MiikkaVilermo1, Sebastian Streich1, Mauri Väänänen1,Karsten Linzmeier2, Bernhard Grill2, Ye Wang11Nokia Research Center, Tampere, Finland2Fraunhofer-Gesellschaft IIS, Erlangen, Germany

In a simple scalable audio coding scheme, thereare usually two layers—a base layer and an en-hancement layer. This paper presents a novelscheme with AMR-WB as base layer and AAC asenhancement layer. To optimally code the signal inthe enhancement layer a frequency selective switch(FSS) control algorithm is described. The FSS de-termines whether the original signal or the residualof the original and base layer signals is sent to theenhancement layer in certain frequency bands. Theproposed method introduces some advancedmechanism to the FSS and the quantizationprocess as well as to minimizing the residual toachieve perceptually optimal result in the encodingprocess. These changes do not assume any modifi-cations in the decoder.Convention Paper 5851

14:30 h

P-3 Advances in Parametric Coding for High-Quality Audio—Erik Schuijers1, Werner Oomen1, Bert denBrinker2, Jeroen Breebaart2


1Philips Digital Systems Laboratories, Eindhoven, The Netherlands

2Philips Research Laboratories, Eindhoven, The Netherlands

In the course of the MPEG-4 Extension 2 standard-ization process, a parametric coding scheme is cur-rently under development. This coding scheme isbased on the notion that any audio signal can bedissected into thee objects: transients, sinusoids,and noise. Each of these objects allows for an effi-cient parametric representation. Recently, improve-ments have been made to increase the overall per-formance of the coder, including an improved noisemodel and an efficient parametric representation forthe stereo image.Convention Paper 5852

15:00 h

P-4 Ultra High Quality, Video Frame SynchronousAudio Coding—Michael J. Smithers, Brett G.Crockett, Louis D. Fielder, Dolby Laboratories, SanFrancisco, CA, USA

Two methods of coding and delivering ultra-highquality audio are presented. Both methods arevideo frame synchronous and editable at commonvideo frame rates (23.98, 24, 25, 29.97, and 30frames per second) without the use of sample rateconverters. The first is an ultra-high quality audiocoder that exceeds 4.8 on the ITU-R 5 point audioimpairment scale at a bit-rate of 256 kb/s per chan-nel and at up to thee generations of encoding/de-coding. The second is an enhanced method ofvideo frame synchronous PCM packing. Specificallythe problem of transmitting 48 kHz audio in 29.97Hz frames is examined.Convention paper 5853


THE VALUE OF INFORMATION

Chair: Peter Filleul, European Sound Directors Association ESDA, UK

Panelists: Paul Jessop, Technical Director, IFPI, UKMark Yonge, AES Standards Manager, UKEric van Tijn, Soundwise Studios, The Netherlands

To celebrate its 5th anniversary as a pan-European rep-resentative body, the European Sound Directors Associ-ation is hosting this workshop which will consider re-sponses by studios and producers to the increasing callsfor extended deliverables in metadata, information han-dling, and how standard systems could provide new in-come streams and impact existing studio revenues, stu-dio personnel and practice. Moderated by Peter Filleul(Director of ESDA and Executive Director of APRS in theUK), a panel of studio owners, studio producers, andmetadata experts from Europe debate the issues andpresent some fresh opportunities. This event is beingpresented by the APRS (UK) and co-hosted by the AES.

Seminar 10 Tuesday, March 25 13:30–16:30 hRoom A

BASICS OF SOUND REINFORCEMENT

Presenter: Peter Swarte, PAS, The Netherlands

Sound reinforcement is applied in places where groupsof people are to be informed or entertained. Also, spe-cialized sound systems exist to help ensure the safety ofpersons and their belongings—sound systems for emer-gency purposes.

In sound reinforcement systems microphones andloudspeakers are very often located in the same room.This can lead to highly irritating acoustic feedback (howl-ing). Sometimes systems are driven close to acousticfeedback, resulting in speech unintelligibility and soundcoloration.

This seminar offers an insight into the needed specifi-cation for a stable sound reinforcement system whichalso meets other requirements such as speech intelligi-bility, frequency- and dynamic range. The stability ofsuch systems is strongly dependent on the acoustics ofthe room and a careful choice of the system elements.Using a line-up diagram as a guide, parts of these re-quirements are visualized. The benefits of using such adiagram are explained. The design of a system followsclear rules and has as secondary benefit that it logicallyleads to a measuring protocol.

Mathematics are kept to a minimum: The decibel nota-tion plays the central role.

Student ActivitiesSTUDENT DELEGATE ASSEMBLY MEETING—PART 2Tuesday, March 25, 13:30–15:00 hRoom H

At this meeting the SDA will elect new officers. One votewill be cast by the designated representative from eachrecognized AES student section in the European Re-gions. Judges’ comments and awards will be presentedfor the Recording Competitions and the Student PosterSession. Plans for future student activities at local, re-gional, and international levels will be summarized.


WAVE-FIELD SYNTHESIS APPLICATIONS

Chairs: Thomas Sporer, Fraunhofer Gesellschaft AEMT, GermanyDiemer de Vries, Delft University of Technology, Delft, The Netherlands

Panelists: Rudolf Rabenstein, Erlangen/Nurmberg University, GermanyGuenther Theile, IRT, GermanyRiitta Väänänen, IRCAM, Paris, France

Ten partners—universities, research institutes, and in-dustries—cooperated within the IST 5th framework pro-gram project “CARROUSO.” The project aims at real-t ime recording, transmission, and rendering ofmulti-object sound events, where the acoustic proper-ties of the recording room are also taken into account.The encoding and transmission is done in the MPEG-4format, which supports separate transmission of thesound content and the spatial sound scene description,and interactive modification of the rendered scene. Therendering is performed by wave field synthesis, en-abling the reproduction of sound fields with natural spa-tial properties over the full listening area. During theworkshop several practical applications of WFS will bediscussed.

Throughout the Convention some of these applicationswill be demonstrated in Room R.


The hot topic with regard to sound rein-forcement today is loudspeakers in con-junction with various types of digitalsignal preprocessing. The preprocessingcan be adapted to the needs in severaldifferent ways such as (for example):

1) Beam steering in large arrays withfrequency dependent tapering;

2) Adaptive beam-steering to com-pensate for weather effects on effec-tive directivity;

3) Adaptive beam-steering to com-pensate for differences between individ-ual loudspeaker drivers in arrays;

4) Signal processing to compensatefor nonlinear distortion in loudspeakerdrivers;

5) Signal processing for multiplebeams using one array.

Array technology in general is goingto expand with the availability of inex-pensive multichannel sound cards withdigital signal processors. This appliesboth to the sending end and to the re-ceiving end of sound reinforcement.Array technology is of great interest forremote sound pickup in mobile phonesand other types of hands-free applica-tions, dynamic tracking in sound rein-forcement and reverberation enhance-ment systems, and many otherapplications.

The second emerging area, whichruns in conjunction with the first topic,

is simply in the design and configurationof drivers and waveguides that act tominimize the discrete device spacing is-sue in array performance creating quasi-“continuous” source devices.

The third hot area is the continuingdevelopment of fully digital net-worked systems, where analog-to-dig-ital conversion happens postmicro-phone/preconsole and continues in thedigital domain through consoles, pro-cessing, routing, multinode distribu-tion, and so on, until the amplifiers,which may be internal to the loud-speakers. Typically this runs on itsown stand-alone network but utilizesstandard fiber or UTP cabling.

EMERGING TECHNOLOGY TRENDS IN THE AREAS OF

THE TECHNICAL COMMITTEES OF THE AUDIO ENGINEERING SOCIETY

Acoustics and Sound ReinforcementMendel Kleiner, [email protected]

Dear Colleagues,

Technical Committees are centers of technical expertise within the AES. Coordinated by the AESTechnical Council, the Technical Committees track trends in audio in order to recommend to the So-ciety special papers sessions, workshops, standards, projects, publications, and awards in their fields.Including the new Technical Committee on Semantic Audio Analysis launched at the 114th AESConvention in Amsterdam, we now have 17 such groups of specialists in our organization. Theyconsist of members from diverse backgrounds, countries, companies, and interests. Each committeestrives to foster wide-ranging points of view and approaches to technology.

One of the expected functions of the Technical Council and its committees is to track new, im-portant research and technology trends in audio and report them to the Board of Governors and theSociety’s membership. The information on what is going on in technical areas of audio around theworld may help the governing bodies of the AES to focus on any items that ought to be high prior-ities for the AES. Supplying this information puts our technical expertise to a greater use for theSociety. Technical Committee meetings and informal discussions held during regular conventionsserve to identify the most current and upcoming issues in the specific technical domains concern-ing our Society. The TC meetings are open to all convention registrants.

In order to facilitate this advisory role of the Technical Council, every two years the chairs of allTechnical Committees report to the Technical Council by listing the most current technical and pol-icy matters the committees are aware of. The short reports presented here identify, from the globalperspective, current research trends and interests in various technical fields of audio.

WIESLAW WOSZCZYK

Chair, AES Technical Council


Several trends impacting the field ofaudio archiving and restoration areworth noting. While this report willdetail the few that seem most immedi-ate, it is perhaps worth mentioningthat the current trends of many otherTCs will likely impact those involvedin archiving and restoration activitiesin the future.

1. METADATAMetadata standards are being craftedfor audio on several different fronts.Both the AES Standards Committeeand the EBU are working on similardescriptive metadata sets based on theDublin Core. At the AES 113rd Con-vention the AES-X114 Metadata re-view project discussed an EBU pro-posal to add an XML chunkcontaining descriptive metadata toBroadcast Wave files. The AES Stan-dards Committee is also working ontwo administrative metadata sets, oneto describe the technical details ofanalog and digital audio objects, andthe second to capture informationabout the processing of audio. SMPTEhas also developed a metadata dictio-nary that in part deals with audio,though more specifically from the per-spective of broadcasters. Other stan-dards of potential value to the archiv-ing community include the AES31Standard that has been adopted by anumber of vendors and the MetadataEncoding and Transmission Standard(METS).

2. BORN DIGITALMore and more audio production is tak-ing place in the computer from the start.Born-digital media is a reality that isgoing to have a significant impact onaudio archives and libraries. Decisionswill have to be made about how to mi-grate not only the audio itself but alsothe project-specific information that tiesthe audio together. This will involvemetadata as mentioned above, the issueof normalization of data, or perhaps em-ulation of outdated equipment and soft-ware. A NARAS/AES committee is

working on a recommended deliveryspecification for the use of record com-panies, record producers, recording stu-dios, and related fields.

3. THE PACE OF CHANGEAudio formats come and go with theblink of an eye. This is true for bothanalog and digital formats alike. Forexample, the past year has seen manydigital devices released that operate atsample rates that were unthinkable justa few years ago. As for analog, obtain-ing high-quality analog tape machinesand tape is becoming increasingly diffi-cult. While perhaps not a new trend, itdoes seem as if the pace has quickenedin recent years. This presents severalchallenges to the archiving/restorationcommunity, including the lack of newproduction of premier players for olderformats and a related difficulty in find-ing parts for existing equipment. Thereis also the problem of the dwindlingpool of service personnel trained tomaintain existing equipment in a condi-tion that is suitable for archival use.One can speculate that the availabilityof reproducers for some formats maybecome as much a problem as the fail-ing media itself. This places the audioarchiving community in a Catch-22 sit-uation. Not only do they often desire tomove to newer or better formats, theymust do so due to the kind of obsoles-cence mentioned previously. However,the pace of change is now so fast thatby the time one evaluates a format forlongevity and consumer acceptance, itis often nearly obsolete itself.

4. STORAGEIn the past quarter-inch analog tapewas the standard medium used for thereformatting of audio recordings. Thisis, however, becoming less and lesspractical in light of point 3 above, andto some producers, less desirable dueto point 2. There seems to be no uni-versal agreement at this time as towhich format should replace analogtape as preservation master. However,we believe we can say that there are

two basic camps: those who advocatemass storage repositories and thosewho advocate the choice of a digitalformat, with the understanding thatmigration will be forced more fre-quently than ever before due to point 3above. The advantages of a mass stor-age approach include the ability to au-tomate migration and to perform dataintegrity analysis on a regular basis.However, one major disadvantage iscost. Although the cost of disk-basedstorage continues to fall, it is still asignificant investment to build massstorage systems. At this time the costof one terabyte is anywhere from$10,000 and up. This may place suchsystems out of reach for smaller orga-nizations and individuals for sometime to come. The cost, however, canhardly be quoted per storage space, asthe basic cost, specifically software, isvery expensive. The storage mediathemselves, for example LTO tapes,are inexpensive—less than $1 per gi-gabyte or $2 for one hour of high reso-lution at 96-kHz/24-bit. More andmore small archives are starting by us-ing computer tape (DLT, LTO) digitaltarget media with the intention offeeding them into small jukeboxes at alater stage. Other organizations and in-dividuals unable to commit to massstorage may be forced to choose a for-mat to migrate to, with very little datato rely on as to the life expectancy ofthe media or of the format. As an ex-ample, take the DVD format. Wehave DVD-V, DVD-A, DVD-ROMand so on. Additionally there is theSACD format. Any of these might beattractive formats, as they allow high-resolution audio and, in some cases,other data to be stored. However,from the experience we are havingwith current CD-R format, which fol-lowing the recommendations ofIASA-TC 03 (http://www.iasa-web.org/iasa0013.htm) requires ex-tensive testing for optimalburner/software/blank matching, andsubsequent data-integrity checking ofthe media, one can hardly say that

Archiving Restoration and Digital LibrariesDavid Ackerman, [email protected]

EMERGING TECHNOLOGY TRENDS …


DVD recordable is an appropriatearchival medium without further test-ing. Additionally it is difficult if notimpossible to predict which of thesehigh-resolution audio formats willgain consumer acceptance.

5. TRAININGThere is the problem of providing train-ing to service personnel and operators.Already we have a generation of audioprofessionals working in the field whohave never encountered the splicing

block. At this time there are few peoplewho have practical experience workingwith formats such as wire and cylinderrecordings. We need to reverse thisgeneral trend if we intend to rescue ouraudio heritage in the coming years.

Audio for GamesMartin Wilde, [email protected]

Audio for TelecommunicationsBob Zurek, [email protected]

The emerging technology trends of au-dio for games mirror some of thetrends in other fields, as well as strikeout on their own in many areas.

1) First and foremost is the trend toassess and investigate the quality ofsound in video games from many perspectives:

•A u d i o c o d e c s : g i v e n t heirwidespread use in the games industry,there is no similarly widespread dataon the subjective quality of each ofthem. Manufacturer claims need to beindependently tested and verified uti-lizing the latest subjective testmethodologies.

• 3-D sound, including surround andmultichannel arrangements: the in-creased use of acoustic modeling andsimulation in real-time gaming appli-cations has far outstripped any formal-

ized subjective assessment of the abili-ty these various systems to faithfullyreproduce sounds in three dimensions.

• Game mastering: there exist notechnical recommendations for gameaudio mastering for such things as bit-depth, sampling-rate resolution, or thedynamic output of game audio de-vices. There are also no technical stan-dards for the equipment and proce-dures used to either make or play backgame audio.

• Synthetic sound: there has been noassessment of the quality of the de-vices rendering synthetic sounds suchas physical modeling or those con-tained within the MPEG-4 specifica-tion and physical models.

• Multimodel interaction and coordi-nation: how do the aural and visualcomponents of a game interact, and

how well do they match?2) Interactive audio. Sometimes

called adaptive audio, the nonlinearnature of games demands new audiosystems to facilitate the productionand real-time presentation of audiocontent under highly dynamic circum-stances and in a creatively intelligentmanner.

3) Game audio education. The tech-nical knowledge and skills requiredfor all game professionals is rising.We need to teach and prepare studentsfor careers specifically in game audioby providing information about gameaudio careers, job opportunities, andnecessary skill sets. Students need toknow where to find that informationand the educational institutions pro-viding courses of study in game andmultimedia audio.

The trend in mobile telecommunica-tions has been toward higher qualityaudio. As portable communicationsdevices have moved from novelty tomainstream use, expectations of theaudio quality and the audio featureshave increased.

On the transducer front electrody-namic receivers have replaced piezo-electric devices used in lower-endproducts. This change provides bettersound quality, improved compatibilitywith assistive hearing devices, and al-lows for leak-tolerant designs thatmaintain a consistent system responseover a wide range of device-to-ear sealconditions. Armature ringers, orbuzzers, are being replaced by mi-

croloudspeakers. These microloud-speaker alerts are now being driven bypolyphonic MIDI tones that can bedownloaded to the device either over anetwork or via a PC interface. Thewider range and polyphony allowsthese devices to be used for audio inembedded or downloadable games.The gaming capabilities are becominga more important feature in futureportable communications products.

In addition to the change in trans-ducers, the trend is toward improve-ment in audio connectivity. Monauralheadsets are being replaced with stereoheadsets to allow playback of othersource material over the communica-tions devices such as MIDI, FM radio,

and MP3 audio. In addition to the con-nectivity change, the trend is towardimproved electronics to support thewide frequency range required bythese formats.

New legislation regarding commu-nication use in cars has led to the in-creased desire for hands-free solutionseither as add-on devices or built intothe communication device. This trendis leading toward the development ofbetter hands-free car kits, improvednoise reduction, and improved con-nectivity. Bluetooth systems are beingadded to portable telecom devices al-lowing easy-to-use wireless connec-tion to automotive sound systems.These telematic systems allow the use



of the automobile’s sound system forhands-free speakerphone use. In addi-tion, voice recognition has becomemore prevalent for voice name-dialingusing user-dependent systems, withthe trend leading toward user-independent VR, voice digit-dialing,and voice menu-navigation in futureproducts.

With the introduction of third gener-ation (3G) technologies the demand on

the telecommunication device audiowill only increase. The new systemssuch as the currently implemented I-mode and soon to be implementedWCDMA will allow wider bandwidthMIDI reproduction, two-way videocalling, and streaming audio integratedin the device.

The convergence of all of these tech-nologies and the globalization of wire-less telecommunications has led to myri-

ad standards for telecommunications au-dio. A key topic that has been broughtup at each of the technical committeemeetings over the past two years hasbeen the issue of the large number ofstandards that suppliers are required todesign to. Another topic of interest thathas come up repeatedly is that of chang-ing headset and handset audio testing.Both of these areas will be covered inAES workshops over the next year.

With the advent of high-resolutionaudio, most notably Super Audio CDand DVD-Audio, the requirementsfor the equipment necessary for au-dio recording and audio storage havedramatically increased. Over the lasttwo years we have seen a rapidgrowth in this area, where the initialstart in the high-resolution formatshas been made using existing equip-ment but with a reduction in thenumber of channels in return forhigher sample rate or resolution.While a lower number of channelshas been sufficient for high-endrecordings, with the growing successof the new formats the need for morechannels reappeared.

At the same time the low end, onthe MP3 formats-based market, hascontinued to grow. While recordingfor these formats does not intro-duce specific challenges, storagedoes. Size and power consumptionare exceptionally important particu-larly for consumer applications,which is the main area of applica-tion for MP3.

All these developments call for sig-nificant innovations in several areas.

1. RECORDINGThe dispute between the use of analogversus digital consoles has revived.Where digital consoles are easy towork with, large digital consoles com-pliant with the requirements of high-resolution audio are still very expen-sive. Analog consoles are (almostintrinsically) suitable for the new ap-plications, but lack the ease of use.Very much related to storage is thequestion of backup. Currently requireddata rates are too large for simple tapebackup; disk or RAID backup seemstoo expensive.

2. INTERLINK Interlink will see also significantchanges in the upcoming years. Cur-rently the new, high data rates requirean unacceptable number of AES-EBUcables or a set of expensive MADI orADAT connections. With the ubiqui-tous presence of CAT5-cabling andthe Internet, we expect to see devel-opment of interlinks based on thesecheap media. Optical 1394 is also animportant new development. In thisarea, the situation is still far from apoint of standardization

3. STORAGEStorage is a multifaceted issue becausethere are multiple applications of stor-age, which can categorized as:

• Recording storage;• Backup storage;• Distribution (e.g., CD, DVD, solid

state) storage.For all these applications, the re-

quirements are different. For record-ing, data rate and robustness are themost important issues. For backup,size and price are the most importantissues. This introduces the interestingdiscussion whether tape as archivemedium will be replaced in the nearfuture by hard disk, as HD prices dropfaster compared to tape. For consumerdistribution we face on the one handthe demand of backward compatibili-ty and cheap, small storage, and onthe other the requirement of ever in-creasing storage for improved quality,playing time, and, most interesting,the combination of high-quality videoand audio. Another issue that needs tobe addressed is what content bundlethe consumer actually wants too see—stereo/multichannel; audio/video/stills/Internet links.

Audio Recording and Storage SystemsDerk Reefman, [email protected]

Automotive AudioRichard S. Stroud, [email protected]

The number of suppliers for automo-tive audio systems continues to grow.Various objective and subjective eval-uation methodologies are utilized by

premium audio system suppliers totune and evaluate their systems.

Customers and suppliers of thesepremium systems want to know how to

most accurately judge them, with anincreasing emphasis on objective mea-surement. Evaluation technologieshave been presented at previous



Coding of Audio SignalsJames Johnston, [email protected]

Juergen Herre, [email protected]

AES conventions. The 114th Conven-tion featured a workshop on correla-tion between subjective and objectiveaudio system evaluation.

There is a continuing trend towardthe incorporation of multichannelsound into automotive audio. Powerfuland inexpensive digital signal proces-sors have made this practical, while theincorporation of rear-seat DVD capa-bility drives the application’s need.Multiple, competing formats can bemanaged by these DSP-based systems.

Consumer interest in MP3 has ex-panded due to its reasonable sound qual-ity, easy Internet access, and the avail-

ability of large, inexpensive hard drives. Expected soon is the direct down-

load of MP3 files into vehicles by-home-based r.f. links.

There is a trend toward flatter andmore installation-friendly transducer de-signs. Driven by assembly costs andpackaging tradeoffs, these devices willlikely replace conventional cone driversin many vehicular audio system config-urations. The challenges will be to re-duce costs and maintain performance.

The trend toward higher perfor-mance audio systems is in direct con-flict with recent trends of cost andweight reduction of components in au-

tomobiles. Increased application ofneodymium magnets will help.

The trend toward the incorporationof newer data-compressed digitalsources (MP3, DAB, satellite radio)may disappoint customers expectingCD levels of performance. Subjectiveevaluation results of XM and Siriusbroadcasts may be available at upcom-ing conventions.

It is believed that the automobile hasbecome the primary listening environ-ment for audio without video. Thisadds growing importance to the accep-tance of the vehicle as a serious placeto listen to music.

Over time audio coding has moved be-yond the research stage and into use,and it has emerged as a strong force inthe consumer market. This has a vari-ety of current implications.

• The consumer can choose variousoptions with regard to quality andmeans of delivery.

• The frontiers of compression havebeen pushed further, allowing fullbandwidth signals at very low bit rates(an example is the so-called bandwidthextension technology), to the pointwhere some people consider such sys-tems broadcast quality.

• Medium bit rates have become ex-tremely common on the Internet forboth piracy and sharing of legitimatelyreleased material.

• Even after a considerable numberof previous attempts, an important is-sue still is how to use low- and broad-band-rate Internet connections for pira-cy-free E-commerce. This is part of theproblem of how to use the Internet andhow to move current business modelsinto the Internet, or come up with newbusiness models for those older modelsthat the Internet makes obsolete. Somedemands of this new business form are:

- Watermarking of audio, includingcompressed audio;

- Support for intellectual propertycontrol, licensing, and permissions;

- The trend toward integration ofcompressed audio, video, film,and graphics, and integration ofthe demands of the various mediarequiring sophisticated licensingand permission models.

• Simple synthetic audio is about tomove from the music maker into thegeneral market as a transmissionmethod.

• There is also a trend toward morediverse hybrid/parametric codecs, forinstance by separately coding transients,steady-state, and residual components.

• Object-oriented audio is now be-coming a force in the marketplace,from MPEG-4 and other directions. In-cluded in object-oriented audio are:

- Audio generated by discrete synthesis;

- Audio coded for a position or otherobject-oriented use;

- Detection of auditory objects in anaudio stream, separation into au-ditory (as opposed to acoustic)objects, and storage and manipu-lation as auditory objects.

Solid-state and hard drive-based au-dio is proliferating, and lots of transcod-ing and direct processing in the com-pressed and/or transform domain maybe necessary in the future, once a trendtoward convergence of different media

will be visible on the consumer market.Some methods for coding control,metadata, and transcoding control arealready being developed and more willbecome necessary. In addition totranscoding, the likes of memory sticks,broadband Internet, and flexible net-works will demand things like fine-grain scalability and lossless coding,with lightweight processes to convertdown in rate with reasonable results.

Five years ago the hardware requiredto run a real-time audio codec was ex-pensive and difficult to work with. Now,present audio coders work in a fractionof readily available CPUs, even for mul-tichannel coding, and new research maybe needed to discover how to use the ad-ditional CPU cycles and memory space.Some possibilities are:

• Improved psychoacoustic models;• Sophisticated acoustic scene analy-

sis of many sorts.

Seen overall, the research in audiocoding is moving to the extremes, bothtoward lowest bit rates (very lossycompression with bandwidth exten-sion) and highest bitrates (noiseless/lossless coding for high-resolution au-dio at high sampling rates/resolutions),as well as the more complex (sceneanalysis and soundfield synthesis ofvarious sorts).



1. STATUS OF HIGHRESOLUTION FORMATSHigh resolution audio has seen furtherconsolidation of the two principal opti-cal carrier-based formats of SACD andDVD-Audio, with SACD currently en-joying the higher profile. However,there is a trend toward DVD players of-fering greater compatibility with a rangeof formats spanning MP3 through tohigh-resolution LPCM at 192 kHz/24bit, and if DVD-Audio becomes stan-dard in the next generation of equip-ment it will support a long-term future.Also, there is a trend toward authoringdisks in a hybrid format to strengthenthe range of usefulness and thus broad-en the acceptance of high resolution,with SACD designed already for hy-brids, and members of the DVD forumexamining the potential to include a hy-brid layer in DVDs.

2. DSD SIGNAL PROCESSINGRefinements in SACD coding perfor-mance and theoretical developmentscontinue, including distortion cancel-lation techniques and improved noiseshaping together with better under-standing of dither in terms of subtlelow-level correlated and cyclic distor-tions. It is anticipated this trend willcontinue, although within the contextof the present SACD standard such re-finements will be principally academicin nature except for the most sophisti-cated mastering technology.

The trend toward enhanced signalprocessing paradigms is relevant to therecording engineer in the context of bit-stream processing, especially with theadoption of intermediate ultrahigh reso-lution techniques such as DSD-widebefore conversion to the DSD 1-bit re-lease format. Such approaches may ad-dress criticisms of signal processing at2.8224 Mb/s not being a viable optionif high resolution is to be properly pre-served in the signal path.

The development of professionalrecording and editing tools for DSDhas continued to progress, with exten-

sion of specialized (and proprietary)boards and software to a number ofworkstation platforms and processors,

3. NEW STORAGE ANDDELIVERY MEDIAA major challenge concerns the rela-tionship of high resolution audio to theemerging higher capacity DVD for-mats such as enhanced red laser andblue laser. There exist proposals forHDTV content, possibly exploitinglayered and reverse compatible cod-ing, but no provision or even debatefor enhanced audio content. Hence amajor near-term objective should be tointegrate high resolution audio contentallowing both high resolution videoand audio content to coexist.

An equally major and currently un-explored direction for high resolutionaudio is the emergent direction ofhard-disk storage. Hard-disk storageis central to a number of rapidly de-veloping audio/video trends, includ-ing networked and PC file distribu-tion, streamed audio, music servers,and personal-disk recording forportable players. These directions areprincipally the realm of MP3 and cur-rently home to considerable litigationregarding copy protection, while highresolution is confined to the safer do-main of copy-protected disks. Howev-er, it is not unreasonable to speculatethat a future trend will be the conver-gence between DVD-player/recordersystems and computer file-format sys-tems where, for example, high resolu-tion audio .wav files and video .mpgfiles can be stored and replayed di-rectly. This together with widespreadavailability of authoring software andhigh-capacity hard disks (HD) onportable PCs enables sophisticatedhigh-resolution home recording. Suchtechnology in concert with broadbandInternet access will continue to forgenew and powerful paradigms for dis-tributed and cooperative music pro-duction that will impact both techni-cal and socioeconomic spheres.

4. INDUSTRY SUPPORTA critical area for the smaller special-ized audio company entering the high-resolution marketplace is the enablingtechnologies pivotal to DVD players,advanced signal processing engines,and digital amplification. We observean emerging trend where key compa-nies supply technology for an internalindustry market, thus circumventingfabrication, licensing, and standards-related costs. As an example, the in-troduction of Firewire for high-resolu-tion applications is a key technologythat will be a major requirement in thenext generation of consumer high-endequipment.

5. RESEARCH INMULTICHANNEL AUDIOMultichannel audio gives improvedspatial information over two-channel,two-loudspeaker stereo and as suchcan legitimately be regarded as highresolution even when it does not con-tain word lengths and sampling fre-quencies above CDDA. Consumer ac-ceptance of multichannel audioprimarily for home theater has led tosignificant research efforts in multi-channel surround sound. These in-clude testing and developing ofrecording/coding strategies such asambisonics, ambiophonics, wavefieldsynthesis, and their variants. At thesame time software is becoming avail-able, often as plug-ins for standardprofessional audio workstations andPCs, to enable recording professionalsto research the effects of encoding/de-coding with these methodologies.

Scalability in terms of both individu-al channel performance and spatial per-formance is becoming an importanttopic for research, where both en-hanced spatial coding and its relation-ship toward network audio are envis-aged as a continuing trend. Challengestherefore should embrace the develop-ment of hierarchical coding where aperformance-cost equation can be in-troduced to match individual needs.

High Resolution Audio

Malcolm Omar J. Hawksford, [email protected] R. Melchior, [email protected]



Refinement of the 78-year-old moving-coil loudspeaker transducer isthe primary activity in loudspeakerstoday. On one hand performance is be-ing driven up by refinements in heatremoval and magnetic linearity. Thisin turn allows lower crossover fre-quencies relative to diaphragm size,which in turn allows more choices indirectivity control. On the other handprice and mass are being driven downwhile retaining subjectively acceptableperformance in other industry sectorssuch as automotive and other entertain-ment applications.

In areas such as cell-phone receivers,efficiency improvements are critical toreduce power consumption. In otherapplications such as control monitorsincreased power handling is sought in

order to derive benefit from the ever-declining cost of available amplifierpower.

Just about any loudspeaker tech-nology other than moving-coil isconsidered exotic, not because it isnecessarily new but because it is notwidely applied. These technologiesinclude electrostatic loudspeakers,piezoelectric, and planar magnetic.New models of these technologieswith incremental improvements con-tinue to be introduced. The distribut-ed-mode technology, which is a rela-t ively recent introduction ofengineering radiation by bending-wave propagation continues to makestrong technical and commercialprogress. Automotive trim panelsfunctioning as bending-wave loud-

speakers have been developed.True digital-to-analog converter

loudspeakers continue to be presentedin theoretical papers, but practical im-plementation seems to be some wayoff. The ultrasonic parametric-arrayloudspeaker continues to be developedfor specialty uses requiring ultra-direc-tional sound projection. Papers havebeen given on spherical arrays of smallloudspeakers combined with digitalsignal processing and amplification foreach loudspeaker. There is greatpromise of improved directivity controland active beam steering in practicalpackages.

In general, integration of conven-tional drive units with signal process-ing and amplification seems to be thepath for greatest future development.

Loudspeakers and HeadphonesDavid Clark, [email protected]

Microphones and ApplicationsDavid Josephson, [email protected]

The past year has produced no signifi-cant published advances in micro-phone transducer technology. Incre-mental improvements are being madein manufacturing tolerances and costreduction in traditional designs, andwork continues on relatively new con-cepts such as the analog optical micro-phone (Phon-Or) and differential hot-wire anemometer (Microflown).

Enthusiasm has cooled somewhatfor the performance benefit potentialoffered by microphones using microelectronic mechanical systems(MEMS) technology.

While the theoretical advantages re-main, companies are finding it quitedifficult to achieve good productionyield in silicon microphone technology.

Digital microphones are now on themarket from several vendors, but remainanalog microphones with internal a/dconverters. We are aware of differentapproaches to direct-digital measure-ment of sound but none of these canqualify as emerging just yet.

Another unsolved problem remains

in the design of microphone arrays torecord surround sound. Several groupsare working on adaptations of the vari-ous known stereophonic microphonearrays intended to capture amplitudeand phase differences between chan-nels at the various observation anglesof the array. Work is concentrating onplacing groups of microphones in theair or on the surface of a solid body.As might be expected left-right andfront-back localization can be achievedwith extensions of stereo recording,but source placement at various posi-tions on the side of the soundscape isdifficult just as it is for humans.

In one area of microphone develop-ment there is a potential emergingtrend, but acceptance among micro-phone manufacturers has not beenstrong. Several authors have suggest-ed improved means for describing mi-crophone performance in graphicalterms, which have the potential forgreatly improving the user’s accep-tance and understanding of new mod-els. Microphone companies are held

between competing concerns—theyfind themselves competing with ever-cheaper sources of microphones andfind it difficult to differentiate them-selves in the market using technicaladvantages. This reflects a firmly es-tablished but relatively recent trend inthe audio engineering marketplace,which is the movement of audio-capture activities away from the spe-cialist engineer and increasingly intothe realm of the user or nonspecialist.This has been made possible by thesignificant improvements in dynamicrange and postprocessing broughtabout by digital audio. In the micro-phone arena this has forced reexami-nation of marketing approaches toreach this new kind of Pro Audio con-sumer. The trend here is for micro-phone companies to adjust their mar-keting and sales efforts to reach thisnew audience, and they have general-ly used more subjective tools to do itrather than educating their new audi-ence to be aware of more technicaldetails.



Although multichannel surround soundis seen by many as vital to the future ofthe audio industry, there are still barriersto its widespread adoption. For exam-ple, in broadcasting there is consider-able pressure to use the available band-width of new digital services to delivernumerous programs as opposed to few-er programs in surround sound. In radiobroadcasting there seems as yet to belittle evidence of the widespread trans-mission of digital surround sound. Al-though digital surround sound is begin-ning to be evident in televisionbroadcasting, surround programs are of-ten still transmitted using analog matrixtechniques over two-channel systems.Unlike the situation with consumer diskmedia, data bandwidth is still a relative-ly limited commodity in the broadcast-ing environment. However, consider-able efforts are being made byinterested parties within broadcastingorganizations to promote the impor-tance of surround sound in the face ofother demands on channel capacity.

In the field of music distribution sur-

round sound releases are gradually gain-ing a share of the market, but this is verysmall compared with the sales of moviedisks. This is in spite of the acknowl-edged advantages of surround soundcompared with two-channel stereo,which may be related partly to the de-pressed state of the classical-music mar-ket. It is definitely home cinema that isdriving the introduction of surroundsound into the home, while music repro-duction is following some way behind.The relative scarcity of surround musictitles on consumer media is one contrib-utory factor, as is the somewhat patchymarketing effort for surround sound onbehalf of the music industry. Some alsocite the relative lack of knowledge andtraining among music mixers, accompa-nied by a degree of fear and confusion,as an important barrier to the prolifera-tion of good surround music releases.

In the field of Internet-audio distribu-tion, broadband technology now makespossible the streaming of multichannelaudio to the desktop, and recent imple-mentations of multimedia audio archi-

tectures make provision for surroundsound as a built-in feature. This is seenas important for the future developmentof desktop multimedia. Computergames are also increasingly accompa-nied by digital surround sound, and agrowing number of game platformssupport it.

In relation to new technology there isgrowing interest in the use of wavefieldsynthesis (WFS) for the rendering ofspatial audio. Research is under way thatattempts to identify methods for the effi-cient coding and transmission of WFSsignals, as well as providing better inte-gration and compatibility with conven-tional recording and reproduction tech-niques. Commercial implementations ofthese technologies are likely to be someway off in the mainstream market, partlyowing to the large number of loudspeak-ers required. However the ability to ren-der realistic 3-D audio scenes over awide listening area, allowing listeners tomove freely within the space, is an at-tractive proposition.

Multichannel and Binaural TechnologiesFrancis Rumsey, [email protected]

Network Audio Systems

The committee has identified three im-portant topics related to emergingtechnology this year.

1. AES47 The AES47-2002 Standard for trans-mission of digital audio over asyn-chronous transfer mode (ATM) net-works has now been published and isavailable at www.aes.org/standards to-gether with the companion TechnicalReport AES-R4-2002 which is intend-ed to explain, in plainer language, therationale behind the provisions of thestandard.

ATM is the only network servicethat is widely available from telecom-munications service providers. It offersthe guaranteed quality-of-service re-quired for demanding real-time appli-

cations such as live broadcasting, andalso provides high-speed networking inthe local area.

At least five manufacturers in theUK and two in Germany will haveconformant products in 2003, includ-ing affordable ATM switches designedspecifically for radio broadcasting. Atleast three broadcasters expect to beusing AES47 for live-to-air audio bymid-2003.

Two IEC Projects have been initiat-ed: one to convert AES47 into an IECInternational Standard and the other tospecify a management interface forbroadcasting and other applications,based on the Simple Network Manage-ment Protocol (SNMP, the Internetstandard defined in RFC 1157). AES47specifies a minimal-functionality pro-

tocol for management of audio connec-tions across the network; the latterproject will specify a much morecomprehensive system encompassingthe management of audio signalsall the way from origination to thetransmitter.

2. ISMAThe Internet Streaming Media Alliance(www.isma.tv), comprising membersfrom over 23 companies involved in thenetworked media field, aims “to accel-erate the adoption of open standards forstreaming rich media—video, audio,and associated data over the Internet,”and toward that end is promoting an im-plementation agreement for streamingMPEG-4 video and audio over IP networks.

Jeremy Cooperstock, [email protected] Grant and Brent Harshbarger



These standards currently encompassthe server, client, and file content andhave already been deployed in existingMPEG-4 applications such as Apple’slatest QuickTime Player and demonstra-tion systems developed by Fraunhofer.The next version of the standard isscheduled for release this year; ongoingwork deals with digital rights manage-ment and reliable quality of service.

3. NETWORK MEDIADISTRIBUTION ANDCOMMUNICATION IN THETELEVISION INDUSTRYThe broadcast television industry hasbeen moving seriously into multipointnetworked distribution of video seg-

ments, at least using MPEG-2 overTCP/IP. For its needs the CanadianTelevision Network (CTV) has devel-oped a highly acclaimed system thatsupports encoding, file transfer, remotedecode, and playback with a reason-ably small latency—on the order of afew frames—supported by a databasethat manages search and access.

Others in the broadcast industry areexploring remote video monitoring us-ing networked-media technology, al-lowing studio technicians to observethe transmitted signal from distant lo-cations as a supplement to network di-agnostic output. Miranda has imple-mented a system that scales qualitywith available bandwidth, providing

reasonable quality at reduced framerate, as necessary, and with fairly lowlatency (< 1s) by using codecs fromthe videoconferencing community.

Another direction of interest is thedevelopment of intercommunicationscapabilities between production andon-air staff, encompassing the trans-port of both video and audio over anIP or VoIP platform. One manufactur-er has recently introduced systemsthat allow for integration of theirblack-box units with commodity PCsrunning appropriate software. Issuesinclude support for multiple groups,heterogeneous bandwidth capabili-ties, and integration with third-partyequipment.

Perception and Subjective Evaluation of AudioDurand R. Begault, [email protected]

Signal ProcessingRonald M. Aarts, [email protected]

Newly emerging MPEG audio stan-dards should allow the end usergreater flexibility in controlling andaltering the multichannel signals at thetime of playback. It should be possiblefor the listener to control the acousticand spatial characteristics of individu-al audio objects. For example, a homeprocessing unit may have controls forreverberation, envelopment, and depthalongside traditional bass and treblecontrols. Ongoing work will need toestablish a more complete understand-ing of how these attributes are per-ceived and how they might be con-trolled in an efficient and effectivemanner.

New approaches to the perception

of multimodal stimuli are still anemerging area of development. Au-dio-visual interaction studies that in-dicate the effect of one modality onthe perceived quality of the other oron the overall audio-visual experienceare still relevant and important, andthe influence of tactile-haptic-visceralperceptual input on perceived audioquality is barely understood, despiteits obvious association with low-fre-quency sound.

Related to the need for improvedcontrol over spatial attributes of soundreproduction for consumer systems is aneed for improved control over otherattributes of a musical mix enabled byinexpensive computer software for

amateur musicians.Users of the effects processing pro-

vided in such systems are most oftengiven either only a few nominal se-lections or a set of complicated con-trols only experienced sound engi-neers can understand. There is anopportunity to provide more user-centered control over distortion-basedand modulation-based effects, in-formed by results of ongoing researchin the human perception of the outputof such effects processing in musicalapplications. For example, graphicaluser interfaces based upon psy-chophysical relations make personalchoices more immediately accessibleto a wider range of users.

Due to the rapid growth in processorspeed, memory size and speed, andstorage devices, the ability to use au-dio signal processing, and in particulardigital signal processing (DSP), hasgrown enormously. This is reflected invarious ways.

First, looking at submissions forAES conventions, DSP is among themost popular subjects, even though afew years ago DSP sessions were rare.Further, looking at other conferences

like the ICASSP, DSP is a standard,extremely rich topic.

Second, looking at home as well asindustrial use, there is a rapid growth inthe classic topics such as:

• Growth of the number of discreteaudio channels;

• Increasing audio quality per channel(both width and sampling frequency);

• Increasing quality of buildingblocks, such as sampling rate converters,ADC, and DACs, due to the increasing

availability of consumer-ready DSP.Third, there is an increasing interest

in intelligent user interfaces, like tunequery by humming, searching in largedatabases with abstract queries such as“search Jazz similar to this song,”auto-generated play lists, and the like.

Fourth, there is a direction of qualityimprovement of rendering of soundfrom more or less standard sources likeCD to a higher quality, such as the band-width-extension method, up conversion



of the number of audio channels, andformat conversion; or offering other userbenefits like loudness-level matching incases of one program having a muchhigher or lower loudness than anotherprogram.

Finally, other aspects are topics thatseemed to be oldies, like class D amplifiers,sophisticated echo cancellation, sophisticat-ed headphones SP, real-time processing(such as on PC sound cards), and auraliza-tion, which were known by a few vendors

are now becoming common applications.All of the trends are leading to the de-

mand for much greater computationalpower, memory, word length, and moresophisticated signal processing algo-rithms. This is resulting in demands formore powerful DSPs, but with a lowpower consumption, since much signalprocessing is being done in small,portable, battery-powered devices. On theother hand, due to the increasing capabili-ties of standard PC processors, they are

fast enough to handle a large proportionof the standard studio processing algo-rithms. Furthermore there is a need forsophisticated algorithms on general-pur-pose computers and better design andanalysis tools. In response to these de-mands manufacturers are creating notonly faster processors but processors withhighly parallel architectures and process-ing techniques designed to exploit them,with more regular instruction sets suitablefor high-level programming.

Precious little has changed in studiotechnology in the last year. Most of thenews in Pro Audio has been in the busi-ness and business models.

There are hints that sales of recordedmusic has dropped again last year, al-though it will be some time before fig-ures are forthcoming. Last year on-linemusic retailers were reporting extremelysoft sales. Figures for the drop overallare estimated to be in the 10 percentrange. In an article in The Los AngelesTimes, several companies reported loss-es of both sales and revenue.

European copyright protection isexp i r ing on many impor tan t

recordings from the 1950s. For the most part the industry is con-

tinuing to consolidate (for instance,Mackie, facing bankruptcy, has beentaken over in a stock swap) and down-size (Mars, a large nationwide retailer,has liquidated).

There are continuing increases insales of DVD-A. Notably, sales ofSACD increased dramatically with thererelease/remaster of The Rolling StonesLet It Bleed in SACD hybrid format(combining a Red Book layer with theSACD layer); Sony goes so far as toclaim that this title alone has outsold allDVD-As released since the inception of

the format. There was a hint that theDVD consortium, long resistant to a hy-brid release is now reconsidering addingthis capability to the DVD-A standard.

Increases were reported in both studioproduction of surround music (5.1) andsurround music releases. In 2002 majorlabels started ordering 5.1 mixes alongwith the normal stereo mixes for popmusic releases.

New DSD production equipment wasintroduced by SaDIE, Genex, MergingTechnologies, Mytek, and others.

Neumann released their long-antici-pated Solution-D digitally-interfacedmicrophone.

Studio Practices and ProductionGeorge Massenburg, [email protected]

Transmission and BroadcastingStephen Lyman, [email protected]

Multichannel sound has been unequivo-cally adopted in digital television. Notuniversally by any stretch of the imagi-nation, but high-profile programs suchas sports events, flagship programs, andmovies of the week leave no doubt ofthe future direction. Terrestrial andsatellite services lead the way, cable isright behind. Conventional analogbroadcasters are also capable of provid-ing multichannel sound, albeit in a morerestricted four-channel matrixed format.

There are several factors hinderingmore rapid adoption of multichannelsound. There is the obvious question ofwhat types of locally produced pro-grams can take advantage of the addi-tional flexibility and creative freedomoffered by multichannel audio. Therewill always be a mixture of one, two,and multichannel programs going to air.

Operational practices have been devel-oped around one- and two-channel pro-gramming. The addition of multichan-nel programs makes transitions betweenthese and conventional programs moreobvious; the operational practices need-ed to deal with these transitions have tobe developed.

Digital radio in various forms hasbeen adopted in many nations but is dis-appointingly restricted to one or twochannels of audio. Most systems do,however, allow some form of programassociated data to be sent along with theaudio. This is essentially a parallel pro-gram in the sense that that content has tobe created, have continuity, and can beclosely related to the timing of the audiocontent. It may thus afford a new oppor-tunity to the audio production and post-production communities.

Users of audio for television, and toa lesser extent radio, are beginning todemand that more attention be paid toperceived loudness. The subject recursin the television world regularly, usual-ly stimulated by complaints about loudadvertisements. The usual solution isdynamic range reduction, which doesnot always solve the problem and doesalter the sound. True loudness mea-surements are now being sought to ad-dress the problem directly. Preliminaryresults are expected within the year, af-ter which work on multichannel loud-ness can begin.

USEFUL WEBSITE RELATING TOTHIS ARTICLE

www.aes.org/technical


OF THE

SECTIONSWe appreciate the assistance of thesection secretaries in providing theinformation for the following reports.

NEWS


PNW Hosts MannOn October 28, 60 members andguests of the Pacific Northwest Sec-tion gathered at the Art Institute ofSeattle to hear Hummie Mann talkabout “Music for Film: The Compos-er’s Perspective.” Composer DavidChristensen, an instructor in the Audio Production program at the AIS,opened the meeting by welcomingmembers and guests to the Art Insti-tute of Seattle.

Section vice chair Dave Tosti-Lane,who also chairs the Performance Pro-duction Department at Cornish Collegeof the Arts and is a sound designer, explained that a slide show of the recent AES convention in Los Angeleswas planned, but the computer con-taining the slides decided not to boot.Instead, Jeffrey Bruton, who had

attended the show, was invited to comeforward and give his perspective inwhich he described his impression of ademo of Direct Stream Digital (DSD),noting that it was “digital that soundslike analog.”

Tosti-Lane then introduced HummieMann, who told the group his back-ground. He has been in the film scor-ing business for over 20 years. Hemoved to L.A. from Toronto, wherehe worked with Doc Severinson and“The Tonight Show” band, as well aswith Aretha Franklin. Mann has 40films to his credit as well as hundredsof television shows. He attendedBerklee College of Music in Bostonand majored in composition. Althoughhis specialty is working with orches-tral scores, he does enjoy workingwith different styles of music, includ-ing rock, jazz, and electronic music.

Mann then launched into a detailedoverview of film scoring. Many peo-ple don’t realize that in a typicalrecording session, the musicians sight-read, learn and record a new piece ofmusic in a very short period of time.The musicians are working at anamazing level of proficiency.

Animation music has its own spe-cial requirements. A lot more workgoes into scoring the music to fit theaction of the animation, i.e. Bugs Bun-ny tip-toeing behind Elmer Fudd. Onthe average, Mann can score aboutthree minutes of music a day, but withanimation he is only able to accom-plish about one minute a day.

Mann then moved on to a descrip-tion of the making of “Cyberworld3D” for IMAX. In preparation forplaying a video EPK (Electronic PressKit, used to promote films) of “Cyber-world 3D,” Mann explained that thereare different styles of animation scor-ing; for example, the Warner Bros.style catches every action with themusic. Musicians wear headphones,which contain a click track. There arealso streamers that appear in thescreen that provide visual aids forsync points in the action. He said thatthe basic assignment was to tell thestory with music only.

Mann played the EPK video show-ing the recording of an orchestra forthe score to “Cyberworld 3D,” withMann conducting. It was recorded atManta Sound in Toronto, using an 85-piece orchestra in a day and a half. Heexplained the difference between anorchestrator and a composer. An

Hummie Mann, film composer, addresses Pacific Northwest Section.photo by Rick Smargiassi

and Lifetime, and some have wonawards at film festivals.

He then opened the floor for ques-tions and answers. Mann talked abouthow he got started, how to pick an orchestra and some of the things oneneeds to know when working with afilmmaker. He also touched on moretechnical issues such as microphoneplacement in orchestral settings andmixing. The group thanked Mann fortaking the time out of his busy sched-ule for this presentation, and to the ArtInstitute of Seattle for providing usetheir facility.

Dave Franzwa

Networked Audio With the generous assistance of Digi-gram, a leader in network audio tech-nologies, a seminar on November 12,2002 brought 29 members and guestsof the Singapore Section to the HiltonInternational Hotel on Orchard Roadto hear about “Networked AudioTechnologies.”

Few technologies have raised aslarge a buzz in the audio industry inrecent memory as networked audio.With digital audio compression for-mats now commonplace, audio distri-bution via networks has unleashed afantastic array of new possibilities.These can both inspire and intimidateaudio engineers who have longviewed the IT industry as having littleto do with the professional audio industry.

The options offered by networkedaudio opens up a world of new appli-cations that can extend an audio engi-neer’s palette of working methods.Paul Lee Thiam Seng, sales director ofDigigram, presented a comprehensiveanalysis and explanation of the historyof networking and the possibilities itoffers today. His colleague, Christo-pher Wu, senior support engineer ofDigigram, joined in answering ques-tions from attendees. Prior to the startof the presentation, attendees weregiven close-up demonstrations of sev-eral network audio configurations thathave real-world uses. This involvedvarious Digigram components linkedby simple CAT5 cabling.

Lee started with a brief history of

Yates, and they suggested that theybuy a song from the Irish folk groupthe Chieftans for the film. Mann sug-gested that he might be able to comeup with a few themes over the week-end. He bought all the Chieftansrecordings he could find, familiarizedhimself with the style they were after—very upbeat, with a Scottish/Irish flavor— and worked it success-fully into his score.

Mann then played the musicaltheme as it was written for the endingcredits to illustrate timing issues. Thetheme is in straight time and has eightbar phrases with a consistent beatthroughout. Then, he played the themeas it was arranged to fit the action ofthe scene. It was clear that additionalbeats were added or removed here andthere to fit the music to the action ofthe film.

According to Mann, the process thata composer has to go through consistsof: writing the themes for the film,then making them malleable enoughso they can work in different emotionsand fit the action of the film. Some-times it takes longer to get the tempoand timing right than it does to writethe music. He pointed out that com-posers must treat the score like a char-acter in a movie. The theme shouldchange throughout the film, much likea character changes from beginning toend. For example, he played the themeas it was written for the love scene inthe movie, to demonstrate how itchanged for that emotion.

Mann told the group that he used toteach a film scoring class at UCLA.He started one in Seattle five yearsago called the Pacific Northwest FilmScoring Program, which has been verysuccessful. His classes have scored 21films using orchestras of up to 45pieces. To complete the program, astudent must score a student film (stu-dent project). Graduates of the pro-gram have scored films from WWU,Seattle Film Institute, Bellevue Com-munity College, and Seattle CentralCommunity College. There are three10-week classes in film scoring andorchestration that take place once ayear. According to Mann, some of thefilms from his classes have actuallybeen sold and been shown on HBO

orchestrator chooses which instru-ments play which parts of the alreadycomposed music. A composer writesthe music and decides which instru-ments play the parts. Mann writes byhand, using a 12-line sketch. He findsthat he works faster by hand than byusing a computer with notation soft-ware because software can cause errors that are not always obvious, soit requires another level of proofread-ing to prevent, for example, enhar-monic problems of say, recording anA-flat rather than a G-sharp, or quanti-zation problems, where the notes arenot properly aligned between parts.There can also be problems with copyand paste, where an error is regenerat-ed many times over.

When asked how he writes, Mannsaid he hears the themes in his headand writes them down, then fine tunesthem with the piano. When scoring, acomposer must be conscious of thevoicings and range of the instruments.Electronic scoring (MIDI to score) cancreate notes that do not even exist onsome instruments. A synthesized ver-sion of an orchestration doesn’t neces-sarily sound okay for an orchestra andlive instruments playing together in aroom. The voicings have to be writtendifferently for a live orchestra in aroom.

In talking about the process of filmscoring, Mann said he was hired towrite a score to replace another scorein “The Year of the Comet” (starringPenelope Ann Miller and Tim Daly).To see how the process works, heplayed a section of the film from thework print, which is unfinished. Thisprint has only the dialog. Then heplayed the same section with the othercomposer’s score, and finally the ver-sion with Mann’s score. The workprint version provided a point of refer-ence. He explained that the spottingsession is when the composer sits withthe director to decide where the musicis going and what kind of emotion itshould convey. This occurs after thefilming is finished. Mann’s version isthe finished version. The film is bigger— fits the screen —and is color-cor-rected. When Mann got a call fromRob Reiner to re-score the film, hemet with Reiner and director Peter

OF THE

NEWS

SECTIONS


seminar on the legal aspects of work-ing as an audio professional. SamuelSeow, a leading entertainment and intellectual property lawyer, offered tohost a Christmas Supper at his office and present an overview ofthese laws and how they apply torecording engineers.

With only a week’s notice, the sem-inar “Legally Speaking —The Lawand the Audio Professional” held onDecember 16 drew 18 members andguests to the wood-paneled offices ofPortcullis I. P. Pte. Ltd., where Seowis a director. Joined by fellow solici-tor, Kelvin Sum, Seow distributed aspecially prepared printout to the attendees before the seminar got underway.

Seow covered the handout’s mainpoints, all of which focused on therights likely to affect a sound engi-neer. He answered questions raised after almost every point from many ofthe attendees. His willingness to answer straight away instead of duringthe planned Q&A session at the end ofthe seminar was greatly appreciated.For a majority of the attendees, theramification of the rights available, aswell as those not available, to them assound engineers, was an eye-openerthat will no doubt resonate in theirprofessional careers.

When the evening finally drew to aclose, Seow extended an offer of rea-sonably priced legal advice. Pricing isoften the biggest perceived obstaclefor most engineers to learn more aboutthe application of law to the profes-sional activities of modern-day soundengineers. Seow’s offer will hopefullylead to more engineers overcomingtheir reticence about seeking profes-sional legal advice.

The section thanked Seow andPortcullis I.P. Pte. Ltd. for their time,legal insight, hospitality and use oftheir offices that made this seminarpossible.

Kenneth J. Delbridge

India Visits PALM On September 26-28, members of theIndia Section visited the ProfessionalAudio Live & Musician (PALM)Show & Conference, held at the Hall

each solution could be applied to solvespecific audio distribution require-ments. Lee covered the various draw-backs of each system, such as cablingfailure producing cascading distributionproblems, although he pointed out vari-ous ways to minimize these risks.

Of great interest to all audio engi-neers when considering installation ofaudio distribution equipment is the issue of future proofing. This issuewas addressed, first from the stand-point of the long-term viability of cur-rent encoding formats like MPEG-1Layer III, and the case for AdvancedAudio Coding was put forward, espe-cially how it fits into the triangle ofaudio, control and connectivity for asystem concept that is future proof.Peak Audio’s Cobranet, which distrib-utes uncompressed real-time digitalaudio over a Fast Ethernet network,was also discussed.

With the presentation material cov-ered (supplied in printouts to attendeesas well as a downloadable file via thisWeb site), the question-and-answersession was lively. For many of the attendees, the seminar helped to shedlight on the added options that net-worked audio offers in their day-to-day working lives. Detailed questionson hardware and software capabilitywere also answered.

With the Q&A concluded, the groupgathered for a photo. Many stayed onto ask more questions as well as exam-ine the networked audio demonstra-tions set up in the function room. Thesection thanked Digigram Pte Ltd fortheir generous support in making theseminar possible.

Legally Speaking One of the top requests from Singa-pore Section members has been for a

the underlying technologies and stan-dards of Ethernet networking beforegetting into the substance of his pre-sentation: how network technologyand digital audio have grown closersince their inceptions to the pointwhere the two now offer powerfulcombinations in real-world situations.The main advantages of using net-worked audio, according to Lee, areprogram content, efficiency, controland device monitoring. Ethernet tech-nology now offers audio engineersnew options when working with audio distribution. In a simple exam-ple, Lee explained how cabling costsin a stadium could be reduced byover 95% through the use of inexpen-sive CAT5 cabling instead of tradi-tional audio cables.

Lee went on to discuss networkedaudio solutions for real-world applica-tions, including point-to-point stream-ing of audio, IP point to multipointstreaming, point to multipoint audiodistribution over Ethernet and multi-point to multipoint real-time audio dis-tribution over Ethernet. Each solutionwas demonstrated in real-time, withthe audience invited to ask questionsas the music used for the demonstra-tion started to play out from net-worked playback units.

Digigram equipment was used in thedemonstration and included their NCXnetworked audio management soft-ware, HitPlayer-L networked audio device and Digigram’s patent-pendingEtherSound that enhances establishedtechnologies to provide easy-to-imple-ment, high-quality audio networks.The powerful options included in net-worked audio solutions, like remotevolume control, were previewed.

The example of music playback in ashopping mall was used to show how

OF THE

NEWS

SECTIONS


Singapore Section members and guests gathered at meeting exploring NetworkedAudio Technologies in November. photo by Kenneth J. Delbridge

of the International Region, inaugu-rated the conference. In his address tovisitors, he emphasized the impor-tance of co-operation in the heavilycompetitive pro audio industry. Hemaintained that AES provides the per-fect forum for bringing audio profes-sionals together to help consolidatethe industry in these trying times.“The global recession is showing itseffect on our industry as well. Weshall survive only if we all cooperatewith each other, understand each oth-er’s problems and solve them our-selves together,” he said.

Subhash Gai, noted Indian film director and producer, also gave akeynote address. He said he had highhopes for the Indian film industry andtalked about plans now underway for anew school for film and televisionprofessionals.

The three-day conference programwas hectic and section members wereactive all through the day. There was awide range of workshops on topicssuch as the reproduction and ‘mis-re-production’ of sound, the true mean-ing of stereophonic recording and reproduction, studio monitoring of 5.1surround sound, special techniques ofcinema sound, MIDI techniques, newchallenges in FM radio and more. Onesession that attracted great interest cre-ated a lively exchange of opinions. Itfeatured a panel of musicians andrecording engineers, including: AnilMohile (composer/arranger), UlhasBapat (santoor player), Ronu Mazum-dar (flautist), Ravindra Sathe (vocal-ist), Taufique Qureshi (percussionist),Avinash Oak (studio engineer) andUday Chitre (live engineer). NanduBhende served as moderator.

The conference concluded with ameeting of the India Section. Thieleattended and was able to meet everymember personally and inquire abouthis or her work. Members also reviewed section activities and dis-cussed future plans. The followingday, Thiele and Nandu Bhende, chairof the India Section, proceeded toChennai, to help establish the newSouth India Section. Vijay Modi ofModi Digital and an active sectionmember, was responsible for the for-mation of the South India Section.

Nepal, Sri Lanka, Singapore and HongKong. Studio Systems, a popular audiotrade magazine, along with the IndiaSection hosted the event.

Neville Thiele, AES vice-president

of Harmony, Nehru Centre, in Mum-bai. The PALM show, now in its sec-ond year, is a premier exhibition foraudio professionals in India and neigh-boring countries such as Pakistan,

OF THE

NEWS

SECTIONS


At the PALM Show 2002 (left to right): Mr. and Mrs. Anil Chopra, Neville Thiele,Subhash Ghai, Kiran Shantaram and Nandu Bhende.

Neville Thiele (center) and Vijaysinh V. Merchant, founder of India Section, withother members and guests.

Subhash Ghai (center, foreground) mingles with guests and artists at the ArtGallery of PALM, showing various aspects of the art of sound.

the degree of directivity that can beachieved at audio frequencies from asmall, parametric array is impressive.While the effects of nonlinearity in airare often small enough to be ignored,interesting things can happen if theamplitude of the sound wave is high.Initially, pure sine waves produce har-monics as they propagate. If two sinewaves are transmitted at a high level,then sum and difference frequenciesare produced in addition to harmonics.

For more than 30 years, the U.S.Navy has experimented with differ-ence-frequency generation as a meansof producing a narrow sound beam atlow frequency from the nonlinear interaction of two narrow, high-fre-quency beams. A relatively smallphysical array generates the high-fre-quency beams while the low-frequen-cy beam forms in the water. In effect,the low-frequency beam is producedby a long end-fire array in the water.

More recently, parametric genera-tion has been investigated for audioapplications. While the effect is dra-matic and there are legitimate applica-tions, effective design relies on an understanding of the physical mecha-nisms and their limitations. Gabrielsondemonstrated a small parametric arrayand gave an overview of the mecha-nism of parametric generation ofsound. He also discussed some of thelimitations and design pitfalls that arecommonly ignored when the desire todevelop revolutionary products pre-empts good engineering sense andphysics.

Dan Valente

Acoustics, Part II In October, more than 50 membersand guests of the Vancouver Sectionattended the Acoustics, Part II meetingheld at Dick & Rogers Sound Studio.John Vrtacic provided a history of thebeginning of the pro audio scene atLittle Mountain in Vancouver. He alsohad a few tips on what makes for agreat sounding studio. Afterwards, alively discussion panel on a wide vari-ety of acoustic topics featured RogerMonk, John Vrtacic, Michael Leader,Hamid Bouhioui, and Peter Janis,chair. A special welcome went out to

flow of a typical Pro Tools system, thenbrought it up on his laptop to demosome of the features of the user inter-face. He also talked about high sam-pling rates and proper audio testing.

The spring semester began with avery interesting talk by acoustics fac-ulty member Tom Gabrielson on the“Parametric Generation of Sound inAir.” Gabrielson demonstrated the impressive directivity produced byhis parametric array compared to asingle loudspeaker. He also dis-cussed some of the limitations andpitfalls of parametric arrays, whichinclude low efficiency and difficultyin producing multiple frequencies foraudio applications.

Alexandra Loubeau

ArraysThirty-two members and guests of thePenn State Section gathered on January20, at the Applied Science Building atPenn State University for a talk by TomGabrielson on parametric arrays.

The meeting opened with some gen-eral business, including a brainstorm-ing session on possible guest speakersfor the upcoming semester. In the audience was Jim Anderson, vicepresident of the Eastern Region,USA/Canada. Anderson is a recordingengineer in Brooklyn, NY, and mayreturn later in the semester to give atalk on recording. He pointed out thatAES has restructured the nationalmembership fee system to include a$20 membership for students thatwould allow them to access the AudioEngineering Society Journal online.

Anderson also mentioned that as thenumber of members in the section increase, so does the money given tothe section by the national society.President Dan Valente encouraged active recruitment by all those present,pointing out that undergraduates arewelcome to join as associate members.Alexandra Loubeau, vice chair, alsoinformed the group that she has addeda membership page to the Web site at:www.clubs.psu.edu/aes. The page includes a link to the national mem-bership application.

Gabrielson then took center stage todiscuss parametric generation ofsound in air. According to Gabrielson,

At the same meeting the managingcommittee was also elected. The feel-ing was that there was a strong AESpresence in the region.

Avinash Oak

Penn State RecapThe Penn State Student Section ispleased to report that membership hasincreased over the past severalmonths. A contributing factor in thissuccess has been the high quality ofthe monthly guest presentations. Fol-lowing is a summary of some of themeetings held over the last year.

The first meeting of the Fall 2002 semester was held on September 23 atthe Fred Waring Collection at Pattee Library on the Penn State campus. Alocal from nearby Tyrone, Fred War-ing, a choral conductor and showman,donated his collection of musicalarchives and memorabilia to Penn Stateupon his death. Curator Pete Keifergave a brief history of Waring anddemonstrated how he transfers record-ings from old archival formats such asaluminum records and wire recordingsto digital formats like the CD.

On October 28, acoustics facultymember Victor Sparrow shared hisimpressions of the 2002 AES Conven-tion in Los Angeles. Hot topics at theconvention were loudspeakers andmultichannel sound. Sparrow attendeda Surround Sound Workshop onrecording for the latest multichannelaudio formats. He also went to a virtu-al acoustics demo to hear the latest in auditory spatial illusions.

At the November 11 meeting, MikeDaley, chapter vice chair, discussedand demonstrated his new crossoverdesign for PVC pipe loudspeakers.Using CALSOD, Daley optimized thecrossover and increased the crossoverfrequency. The section held informallistening tests to compare his design tothe original used for the 2001 fresh-man seminar taught by faculty advisorSteven Garrett.

On December 10, Tad Rollow, one ofthe founding members of the Penn StateSection, spoke about the computer-based audio workstation Pro Tools andhis job as a software engineer forDigidesign. He described the signal

OF THE

NEWS

SECTIONS


since the directivity of a sound source depends on the size of the source, tofabricate a highly directional sourcerequires that the source be much largerthan the wavelengths of sound it is reproducing. A gigantic conventionalloudspeaker could be highly direction-al, but also impractical. One solutionis to use a reasonably sized loudspeak-er that generates very small wave-lengths of sound, or ultrasound.

Borrowing from research into underwater sonar techniques datingback to the 1960s, Pompei appliedthese principles to audible propagationthrough air. Researchers as well as potential manufacturers, such as Mat-sushita and Ricoh, had also tried to develop ultrasonic arrays to propagatesound through the air in the 1970s and1980s, but were thwarted by the highcost of large arrays, the excessive dis-tortion (~ 50%), as well as inconsis-tencies in and limited bandwidth ofthe transducers themselves. Othershave also tried “Tartini tone” or beatfrequency methods, where audiblesounds are generated through the interaction of two inaudible sounds; a200-kHz tone interacting with a 201-kHz tone could generate an audible 1-kHz tone. These techniques haveproved to be of limited practicality.

Through mathematical analysis andengineering insight, Pompei overcamethese barriers to develop his audiobeam system. He realized the air mod-ifies and distorts the ultrasonic wavesas they propagate, but in a predictableway. Through characterizing this dis-tortion mechanism, ultrasonic signalscan be calculated and generated suchthat as they propagate through the air,their audible distortion product is thedesired sound. Since this transforma-tion of the ultrasonic signal into theaudible byproduct takes place continu-ously over a distance (of about ten meters for this current system), theplanar array of ultrasonic transducersgenerates a sound source that behavesas a columnar audible array, and ishighly directional.

Pompei’s demonstration systemconsisted of a portable CD player, apowered box about half the size of ashoebox that processes and amplifiesthe source signal, and a panel about

their partnership approach to sounddesign for theater and spoke at lengthand in depth about such topics as using the architecture of the theater toaid in the reinforcement process, thedesign and tuning of loudspeaker sys-tems, the use of wireless microphonesand sound effects design and play-back. Assisting in the technical aspects of the discussion was GaryStocker, director of research and development for Masque Sound.Stocker brought some wireless micro-phone transmitters to show how minia-turized the technology has become. Healso discussed the loss of available fre-quencies for wireless in the theater,HDTV and other commercial uses. Alively question-and-answer periodrounded out the evening.

Swiss MixerOn January 30, more than 30 membersof the Swiss Section gathered at theBasel Theatre for the first meeting ofthe year. After a short introduction byJoël Godel, Hervé De Caro of Innova-Son talked about the latest develop-ments in InnovaSon mixers.

Robert Hermann of the Basel The-atre then discussed the theatre itself,and explained how the sound rein-forcement system, which features sur-round sound, was conceived andinstalled. After a quick look at thehouse mixing desk, Hermann explained how the InnovaSon mixers enhance the theatre’s sound.The meeting concluded with a demon-stration of the mixing desk.

Joël Godel

Loudspeakers Some thirty members and guests of theBoston Section gathered on January14, at the MIT Media Lab to hear JoePompei of Holosonic Research Labtalk about his interest in 3-D audio andhyperdirection loudspeakers.

Pompei began by discussing his interest in 3-D audio and his researchinto using ultrasound as a loudspeakerwhile he was a graduate student atNortheastern University, and subse-quently as a researcher at the MIT Media Lab. According to Pompei,

Jim Manson of Skydance Audio, whoflew in from Kitimat BC just to attendthe event.

At this meeting, members also elect-ed new officers for the 2003 executiveteam. They consist of: Peter Janis,chair; Lucas Truman, vice chair;Gregg Gorrie, treasurer; David Linder,secretary; Gary Osborne, past chair;and Hamid Bouhioui, Mark Gordon,David Kelln, Tom Neudorfl, and KateSimms-Rita, committee members.

Myths ExaminedThe January meeting of the Vancou-ver Section featured Bill Whitlock,president of Jensen Transformers, whospoke to a near capacity crowd at theAi-CDIS campus in Burnaby about thetop ten myths of signal interconnec-tion. Present were installers, engineers,studio technicians and TV people, forwhom signal interconnection plays animportant role in everyday life.

The meeting went by quickly because there was so much informa-tion to cover. However, there was timeat the end for everyone to ask ques-tions. Whitlock began by explainingwhere 600Ω terminations originatedand why with today’s equipment suchterminations are unnecessary and like-ly to cause distortion. He also talkedabout electrical safety and was carefulto point out that even AV equipmentcan cause lethal currents or fires whendeliberately miswired.

By the end of the event, all of the attendees felt that they had come awaywith some new knowledge and a bet-ter understanding of audio intercon-nection issues. The group thanked Ai-CDIS for providing the location forthis event, and Bill Whitlock forspending time with the group.

Broadway Sound DesignThe New York Section met on Febru-ary 11 to hear the members of AcmeSound Partners: Tom Clark, MarkMenard and Nevin Steinberg talkabout Broadway sound design. FormerNew York Section committee memberJohn Kilgore, director of recordingservices for Masque Sound, moderat-ed the meeting.

The three guest speakers discussed

OF THE

NEWS

SECTIONS


WatermarkingThe January meeting of the BritishSection on audio watermarking fea-tured Bob Walker of the BBC’s Research and Development Depart-ment. The event was an introductionand tutorial, rather than detailed analy-ses of particular systems or a discussionof the “rights” or “wrongs” of water-marking. Intended as an overview, themain topics included: What is water-marking? Why do we need it? How dowe do it? How well does it work?

The BBC has been studying hidingsignals in audio streams since 1970.Some of the methods developed at thattime were subsequently used in broad-casting and may still be in use today.Modern technology has made methodsof hiding and subsequently recoveringsignals more practicable than they wereand, of course, vastly more complex.

There is a great deal of interest incopyright management and protection.Methods of embedding signals aim tocreate electronic “gatekeepers” in audio equipment to ensure that onlythose with the right to access it can doso. That way, unauthorized reproduc-tion, and especially unauthorized copy-ing, might be prevented. In addition, anembedded signal can be used to markcontent to enable its provenance to betraced, for example, to identify “leaks”that can occur during the production ofa disk or program. This usually permitslower watermark amplitudes, mainlybecause detection can be spread over alonger interval than for “real-time” con-trol. This significantly eases the prob-lems of finding a compromise betweenthe principal system properties.

The audio watermarking processstill faces myriad technical hurdles,i.e. maintaining audio quality, systemreliability and robustness in the pres-ence of attacks and conflict betweenwatermarking and audio compressionsystems.

Walker began his talk with the originsof steganography and continued with ageneral outline of watermarking systems,especially as they are applied to audioand video signals. He also covered themyriad issues in the recording industrythat revolve around protection and man-agement of intellectual property rights.

14-in across and 1-in thick that con-tained an array of about forty ultrason-ic transducers. A variety of soundsamples (birdcalls, breaking glass,whispers) were played for the audi-ence through the system; the perceivedsound was indeed highly directional,as one could hear the sound clearlywhen the array was pointed directly atthe listener, but not when it was point-ed only a few degrees away. Whenaimed at the ceiling, the perceivedsound source seemed to be the ceilingitself, as the audible sound would reflect off any hard surface. Pompeialso noted that since no reverberantenergy accompanied the sound per-ceived by the listener, the perceivedsound source was quite close.

The audible signal generated by theaudio spotlight has a 12 dB/decadeupward slope, such that the preprocess-ing available includes an adjustablelowpass to reduce the audible high fre-quency content. Increasing the band-width to include more low frequencycontent simply limits overall maxi-mum output of a given system. Thesystem demonstrated to the groupseemed to have audible content downto a couple of hundred cycles per second. Pompei noted that some appli-cations benefit from the use of a conventional subwoofer to augmentthe signal.

In discussing various applicationsfor the audio spotlight, Pompei notedseveral companies have shown interestin the technology. Initial installationswere in amusement parks and muse-ums, and now the field of applicationshas broadened to include automobiles,office environments and wherever lo-calized sound is desired. The meetingconcluded with questions and ademonstration of the system installedat the Media Lab, which consisted ofthree audio spotlights and a directionalvideo screen. A video of a singer withaccompanying musicians varied as onewalked across the room; if one walkedtoward one side, the trumpet was audi-ble and visible. If one walked to theother side, it was possible to both seeand hear the violin.

The Boston Section thanked Pompeifor his illuminating presentation.

Tom Wethern

OF THE

NEWS

SECTIONS


2003 June 23-25 NOISE-CON2003, Cleveland, OH. For in-formation fax: 515-294-3528or e-mail: [email protected].

•2003 June 26-28: AES 24th

International Conference,“Multichannel Audio: The NewReality,“ The Banff Centre,Banff , Alberta, Canada.For more information see:www.banffcentre.ca.

•2003 July 7-10: Tenth Interna-

tional Congress on Sound andVibration, Stockholm, Swe-den. For information e-mail:[email protected].

•2003 September 11-13: Xth

Symposium on Sound Engi-neering and Tonmeistering,Wroclaw University of Tech-nology, Wroclaw, Poland. Forinformation check the Web:zakus.ita.pwr.wroc.pl/isset03or e-mail:[email protected].

•2003 October 10-13: AES 115th

AES Convention, Jacob K.Javits Convention Center,New York, NY, USA. See p. 472 for details.

•2003 October 20-23: NAB

Europe Radio Conference,Prague, Czech Republic. Contact Mark Rebholz (202) 429-3191 or e-mail: [email protected].

•2003 October 30-November 1:

Broadcast India 2003 Exhibi-t ion , World Trade Centre,Mumbai, India. For informa-t ion contact Kavita Meer, director, Saicom Trade Fairs& Exhibitions Pvt. Ltd., tel:+(91-22) 2215 1396, fax:+(91-22) 2215 1269.

•2003 November 10-14: 146th

Meeting of the Acoustical Society of America, Austin,TX. For information contacttel: 516-576-2360, fax: 516-576-2377 or on the Internet:www: asa.aip.org.

Upcoming Meetings

*Acoustics Engineering .....................385www.acoustics-engineering.com

Nissan Technical Center...................459 www.nissanusa.com

*Prism Media Products. Inc. ..............407www.prismsound.com

*SRS Labs. Inc. ...................................391www.srslabs.com

*THAT Corporatioz .............................395www.thatcorp.com

*AES Sustaining Member.

errors. By setting the detection thresh-old, the balance between failure to detect a watermark that is present andwrongly detecting one that is not can beadjusted to suit the application. Mostapplications require the risk of falsealarms to be very low. A false alarmcould result in a consumer being barredfrom using a legitimately acquiredproduct or to someone being wronglyaccused of theft.

The lecture concluded with a briefsummary of how well current (known)watermarking systems work. The prob-lems were subdivided into four cate-gories. The first, entitled “FunctionalInadequacies,” concluded that the per-formances of the best current systemswere probably not adequate for accesscontrol, though they might be for track-ing applications.

The second category, “Patent andIPR Problems,” referred to the water-marking patent and IPR situation. Mostof the watermarking IPR appears to beheld by a very few organizations. Thissituation is likely to lead to only one ortwo systems being available world-wide. In terms of “Security Issues,”those few watermarking systems mightbe attacked and defeated in a very shorttime, especially if there were significantfinancial incentives for doing so. Hav-ing a larger set of potential systemswould reduce the pressure on each fromco-coordinated attacks.

The final set of problems concernthe compatibility with audio compres-sion systems. Compression systemsreduce the data rate by removing inaudible or barely audible compo-nents. A watermarking system worksby adding inaudible or barely audiblecomponents. There can be no doubtthat an ideal coding system will remove a watermark—that is essentially the definition of the per-fect coding system. In practice, withgood coding systems, watermarks willbe severely damaged by significantcompression, unless the watermarkcan be made an integral component ofthe coding system itself. Then itmight have enough information aboutwhat subsequent coding systemsmight do to the host signal to adapt.

The presentation concluded with ageneral discussion.

As a free-to-air broadcaster, onemight not initially think of the BBC ashaving a substantial interest in water-marking. However, according to Walk-er, there are many potential applicationsfor an organization such as the BBC,ranging from the convention protectionof content, through managing water-marks belonging to third parties and using watermarking to help in the pro-gram production process, for example,by directly linking the Metadata recordto the content.

The watermark signal usually startswith a very poor signal-noise ratio, perhaps -80 dB, but at a muchlower data rate than the host audio sig-nal. Detection is possible only withmatched filtering, usually by usingsome form of correlation and averagingprocess. To remain imperceptible, thewatermark must be masked by compo-nents of the host signal.

Masking can be accomplished usingthe inherent background noise level orby exploiting some other property ofhuman hearing, for example, the rela-tive insensitivity to phase and short-term echoes. The watermark might becarried by modulation of the phasecomponents of the host signal or byadding echoes at time delays and ampli-tudes that cause no perceptible changes.Walker also described the principlesand uses of spread-spectrum modula-tion and the benefits of perceptualmasking.

Even with the additional amplitudepermitted by perceptual masking, muchof the required gain in watermark sig-nal-noise ratio has to be provided by thespread spectrum modulation and its cor-relation processes (or by some othersimilar synchronous integrationprocess). Eventually, the detected signalconsists of the sum of the cross-correla-tion of the host audio with the water-mark and the auto-correlation of the watermark. Though its expected valueis (usually) zero, in individual detec-tions the cross-correlation with the hostsignal will be non-zero. If the two can-not be separated, then a detection errormay occur.

All watermarking systems sufferfrom a finite probability that the hostsignal will sometimes resemble the watermark sufficiently to cause such

OF THE

NEWS

SECTIONS


AUDIO SYSTEM DESIGN ENGINEER

Nissan Technical Center North America(NTCNA), Nissan’s North American engineeringresource, has an excellent opportunity availablefor a Audio System Design Engineer at our state-of-the-art Technical Center in Farmington Hills, MI.

This automotive audio engineer will be responsi-ble for design and development of audio systemsfor North American production vehicles. Principleduties will include: design and release of audiosystem components, primarily head units, speak-ers, amplifiers, antennas and related parts. Thiswill include evaluating, designing, preparing speci-f icat ions; packaging and negotiat ing cost and performance with the supplier; interfacing with in-ternal departments and outside vendors; bench-marking competitor vehicle audio systems; andgathering market research/technology trend infor-mation to ensure Nissan's audio systems are aleader in the marketplace.

At a minimum, you must have a BSEE degree oran engineering degree with emphasis in acousticsand 5+ years of automotive audio system designexperience. Experience in evaluating and imple-menting cost reduction proposals, familiarity withsystem packaging and CAD (I-DEAS) experienceare preferred. Must have familiarity with MS Officeapplications and be authorized to work for anyemployer in the United States.

Imaginative new ways of thinking are what Nissanexpects of every employee. To see what you can ex-pect of Nissan and to apply online for this position visit

www.nissanusa.com Be sure to include source code AES031903. EOE

AdvertiserInternetDirectory

LSB (max.), SNR of 89 dB (typicallyat 100 kHz) and THD of –100 dB at100 kHz. The AD7654 is well suitedfor AC servo motor control, frequencyinverters, three-phase power monitor-ing, uninterrupted power supplies,four-channel data acquisition, andautomotive applications. AnalogDevices, Inc., One Technology Way,P.O. Box 9106, Norwood, MA 02062,USA; tel. +1 781-329 4700 or 1 800262 5643 (toll-free); fax +1 781 3268703; Web site www.analog.com.

A E S S U S T A I N I N G M E M B E R

LINE ARRAY SYSTEM is a dual 8-inch, bi-amplified three-way designweighing only 28 kg (62 lb.) andincludes all necessary rigging hard-ware fittings. Model VT4887 arraysare ideally suited for corporate A/Vpresentations, conference, and ball-room venues and a wide range of per-formance-audio applications. TheVT4888 midsize line array element is adual 12-inch, three-way design offer-ing full-range performance in a road-ready package weighing only 108 lb.(including integral suspension hard-ware). The VT4881 compact low fre-quency module is a dual-coil 15-inchsubwoofer array with a frequencyrange of 18 Hz to 160 Hz. GeneralMotors recently employed these sys-tems in a variety of it’s corporatevenues. JBL Professional, 8500 BalboaBoulevard, Northridge, CA 91329,USA; tel. +1 818 894 8850; fax +1 818894 3479; Web site www.jblpro. com.

AND

DEVELOPMENTSProduct information is provided as aservice to our readers. Contact manu-facturers directly for additional infor-mation and please refer to the Journalof the Audio Engineering Society.

NEW PRODUCTS


MEETINGS, CONFERENCES…

NOISE-CON 03, the 2003 NationalConference on Noise Control Engineer-ing, will be held on June 23-25 at theRenaissance Cleveland Hotel in Cleve-land, Ohio. Beth Cooper of the NASAGlenn Research Center in Cleveland isthe general chair of the conference.David K. Holger of the Iowa State Uni-versity is the technical program chair.

The number of abstracts received forthis year’s event is the largest in the his-tory of the NOISE-CON series, whichbegan in 1973. The papers will cover avariety of topics including: NASAGlenn research on aeroacoustics and related jet noise; noise in computer sys-tems; United States noise policy; trans-portation noise; acoustical facilities, design and testing; building acoustics;acoustical materials; classroomacoustics; power plant noise; industrialnoise sources; noise analysis, and soundquality development.

For a complete list of the abstracts,visit the Institute of Noise Control Engi-neering Web site at: www.inceusa.org.Travel planning and hotel informationcan also be found at the above URL.

The Tenth International Sympo-sium on Sound Engineering and Ton-meistering will be held in Wroclaw,Poland, on September 11-13. It is beingorganized by the Institute of Telecom-munications and Acoustics at the Wro-claw University of Technology togetherwith the Polish Section of the Audio Engineering Society.

The conference will cover all branch-es of audio. Contributed papers will bepublished in the Proceedings of the XthISSET 2003, following positive review.The deadline for papers is May 19th. Toregister or for more information, con-tact: inz. Piotr Pruchnicki, Institute ofTelecommunications and Acoustics,Wroclaw University of Technology,Wybrzeze Wyspianskiego 27, 50-370Wroclaw, Poland; e-mail:

[email protected] or on the Internet:

zakus.ita.pwr.wroc.pl/isset03.

TRACK

SOUND

VOCAL/VOICEOVER MICRO-PHONE is a fixed cardioid patternphantom powered FET for users on abudget. Brauner’s Phantom C is a non-tube (FET), large diaphragm micro-phone for vocal use where more prox-imity effect and less sibilance arerequired. The microphone offers 8 dBself-noise, 28 m V/Pa cardioid sensi-tivity, and 142 dB Max SPL at 0.5percent THD. The Phantom C comescomplete with a carrying case andshock mount and retains the qualityand look of the Brauner VM1.Brauner, distributed in the US by theTransamerica Audio Group, 4760West Dewey Drive, Suite # 129, LasVegas, NV 89118, USA; tel. +1 702365 5155; fax +1 702 365 5145; e-mail [email protected]; Website www.transaudiogroup.com.

MULTICHANNEL SAMPLINGADC is a 16-bit, dual 2-channel chargeredistribution simultaneous samplingADC that operates from 5 V. TheAD7654 offers a choice of two sam-pling speeds: 500 kSPS in normalmode and 444 kSPS in impulse modefor low power applications. The unitprovides 16-bit resolution with nomissing codes, INL of plus/minus 3.5

ond, of five chapters, addresses audiocoding standards. Each chapter con-cludes with a summary and an ade-quate list of references. The first tenchapters offer concluding exercises,which makes this portion of the bookappropriate as a student text. Since answers are not provided, and some ofthe exercises are “class projects”rather than simple questions, the textis more suitable for the classroomthan for self study.

Chapter 1 is the obligatory “Intro-duction” and covers the concept oftransformation from the analog domain to the digital, digital audiocoding in general, PCM, the CompactDisc, potential coding errors, and soon. It is brief and to the point.

The second chapter, “Quantization,”also backtracks to what, for most pro-fessionals, is familiar ground: binarynumbers, binary math, uniform andnonuniform quantization, floating-point quantization, companders,round-off and overload errors. At theend of the chapter, we begin to tastethe meat of the book with a section onentropy reduction and Huffman cod-ing. Since these have more to do withdata-rate reduction than quantization,it is not clear what they are doing inthis chapter. Nevertheless one istempted to read further.

Chapter 3 is more mathematical,dealing with the Dirac function, Fouri-er Transform, sampling theorem andthe derivation of some generalizedproperties of audio signals. The chap-ter concludes with a section on “Pre-diction,” that is, a method of reducingthe number of bits required to charac-

terize a signal by quantizing only thedifferences between predicted samplesand actual samples. Before or afterthis might have been a good place toinsert the idea of entropy reductionand Huffman coding.

Predictive coding schemes such asthe one outlined in Chapter 3 arebased in the time domain. Frequency-domain coders have several advan-tages over time-domain coders. To theextent that audio is tonal or redundantin nature, its frequency-domain char-acterization varies slowly with timeand can be quantized with fewer bitsthan would be needed to characterizeit in the time domain. Furthermore, inthe frequency domain, the number ofbits used to encode each componentcan be adapted to the signal. Doing sopermits quantization noise in each fre-quency bin to be adapted to the signaland thus take advantage of the psy-choacoustics of human hearing, e.g.,the frequency-dependent threshold ofhearing, the masking of weakersounds by stronger ones of similar fre-quency, and so on. This is the meatand potatoes of modern perceptual encoding, which is handled in Chap-ters 4 through 7.

Chapters 4 and 5 discuss time-domain to frequency-domain map-ping using two methods. Chapter 4covers transformation using PQMF(pseudo quadrature mirror filter)banks; Chapter 5 deals with theMDCT (modified discrete cosinetransform) method. After discussionof the Z transform, down-samplingand up-sampling in Chapter 4, Bosidevelops the characteristics of

LITERATUREThe opinions expressed are those ofthe individual reviewers and are notnecessarily endorsed by the Editors ofthe Journal.

AVAILABLE

INTRODUCTION TO DIGITALAUDIO CODING AND STAN-DARDS by Marina Bosi and RichardE. Goldberg, Kluwer Academic Pub-lishers, Norwell, MA 02061 USA,2002, 458 pages, $125, ISBN: 1-4020-7357-7.

The primary author of this book,Marina Bosi, needs no introduction tomembers of the Audio EngineeringSociety or to anyone who has fol-lowed the development of perceptualcoding. A consulting professor atStanford University’s Computer Cen-ter for Research in Music and Acous-tics and chief technology officer ofMPEG LA, a firm specializing in thelicensing of multimedia technology,Bosi coauthored numerous articles forthe JAES, was AES president (1998-99), and papers or workshops chair atseveral AES conventions. Now on theBoard of Governors, Bosi was also involved in the development of theMPEG, Dolby and DTS codingschemes and is clearly a well qualifiedauthor.

The credentials of her co-author areless obvious. Richard E. Goldberg islisted as a partner in The BrattleGroup, a management consulting firmspecializing in economics and financeissues. He holds a Ph.D. in physicsfrom Stanford and an A.B. in astro-physics from Princeton. In “About theAuthors,” Goldberg states that “Audiocoding technology and related busi-ness applications have long been areas of interest … .”

The book is divided into two parts:the first, encompassing 10 chapters,covers audio coding methods; the sec-



two-channel perfect reconstructionfilter banks including the QMF(quadrature mirror filter) and CQF(conjugate quadrature filter) solutions. Anticipating the conclu-sions of Chapter 6 that human hearingnaturally divides into 20-30 “criticalbands,” Bosi points out that to createa 30-band filter using a tree of QMFfilter pairs is computationally complexand has a very long impulse response.From there she moves on to discussthe parallel multiband approach vari-ously called a pseudo-QMF (PQMF)or “polyphase quadrature” filter. Thisprovides near-perfect solutions thathave relatively low computationalcomplexity and faster impulse response. These solutions form the basis of Layer I and Layer II coders inMPEG-1 and MPEG-2.

Chapter 5 deals with transformcoders, i.e., coders based upon map-ping discrete blocks of data samplesfrom the time domain to the frequen-cy domain using some variation ofthe Discrete Fourier Transform, pre-dominantly an FFT-implementedMDCT or modified discrete cosinetransform. Early in the chapter, Bosipoints out that the difference betweenthe sub-band (filter-bank) mappingused in MPEG-1 and MPEG-2 Lay-ers I and II, and the transform map-ping used in MPEG AAC is reallymore semantic than real inasmuch asthe two can be shown to be mathe-matically equivalent.

Generally speaking, encoders thatuse relatively few frequency bins areconsidered “sub-band” coders whilethose that employ a larger number ofnarrow frequency bins are considered“transform” coders. Chapter 5 coversfour types of data-block “windowing”prior to transformation: rectangular,sine, Hanning and Kaiser-Bessel. Bosioutlines the pros and cons of each andthe use of block-size and window-shape switching to minimize audibleartifacts.

Chapter 6 is entitled “Introductionof Psychoacoustics,” and that aboutsums it up. Topics include: soundpressure level, loudness, hearingrange, hearing threshold, and masking(both frequency and temporal). Ofgreater importance are sections on

measuring masking curves: narrow-band-noise masking tones, tonesmasking tones, and narrow-band-noise or tones masking narrow-bandnoise. Bosi also introduces the con-cept of “critical bandwidth” (the fre-quency range around the masker overwhich the masking threshold is flat)and the similar-in-concept but differ-ent-in-estimation ERB or equivalentrectangular bandwidth that several authors find more accurate.

In a rather strange twist of affairs, the author chooses to endrather than begin the chapter with asection on how hearing works: outer,middle and inner ear, and so on. It isalso rather strange to find a discussionof the Bark Scale postponed to thenext chapter rather than in conjunctionwith the concept of “critical band-width” and the workings of the basilarmembrane with which it is associated.These are two relatively minor criti-cisms with the organization of thebook.

Chapter 7 (Psychoacoustic Modelsfor Audio Coding) uses the conceptsoutlined in the previous chapter to derive excitation patterns and maskingmodels, models for the spreading ofmasking, and so on. Section 7 discuss-es “Modeling the Effects Of Non-Simultaneous (Temporal) Masking”while Section 8 addresses “PerceptualEntropy.” From these two topics, Bosimoves on to discuss how bits shouldbe allocated to take advantage of themasking effect.

Chapter 8 deals with Bit AllocationStrategies in more depth, movingfrom a simple allocation of bits to optimal bit allocation based upon amathematical approach that mini-mizes the average block error power(without exceeding the permitted datarate) by means of Lagrange multipli-ers. Bosi points out (from the work ofJayant and Noll) that for commonlyused time-to-frequency mappingtechniques, time-domain distortion isequal to frequency-domain distortionso optimizing bit allocation to mini-mize frequency-domain distortionalso ensures minimum distortionwhen data are transformed back tothe time domain.

Chapter 9 takes the reader through

the process of “Building a PerceptualAudio Coder,” beginning with anoverview of the coder building blocks,the computation of masking levels,how to estimate absolute sound pres-sure levels (on which masking basedupon the threshold of hearing depends),and concludes with the need to estab-lish a bitstream format. A slight nodalso is given to “Business Models andCoding Secrets.”

Chapter 10, the concluding chapterof the first section, addresses “QualityMeasurement of Perceptual AudioCodecs.” Most of this is based upon thefive-grade impairment scale establishedby the ITU-R. In this chapter Bosi dis-cusses test methodology includingtraining and grading sessions, selectionof expert listeners and critical material,listening conditions and data analysis.There are also good sections on “Objective Perceptual Measurementsof Audio Quality” and a discussion of“What Are We Listening For?”

The final five chapters address spe-cific perceptual-encoder standards, towit, MPEG-1 Audio, MPEG-2 Audio,MPEG-2 AAC, Dolby AC-3 andMPEG-4 Audio. Those who have readthrough the more tutorial chapters inthe book with due diligence will findSection 2 pleasantly lucid. Surprising-ly, for an author who led technical development for a number of years atDTS, there is no coverage of that sys-tem whatsoever.

The book, published by KluwerAcademic Publishers, has reasonablygood printing and binding, althoughthe layout suggests that the book wasput together piecemeal, without muchplanning for later chapters in the book.Many of the drawings are small, withprinting that is difficult to read. Somelook like they have been copied (poor-ly, and without due credit to theirsource) and shrunk to fit. The book deserved better from its publisher. It isa useful tutorial and reference for thoseinterested in perceptual audio coding.As with any such book, experts in thefield are unlikely to find much that isnew within its pages; technologychanges faster than a book can be writ-ten and published.

Edward J. FosterMarco Island, FL 34145-6723

AVAILABLE

LITERATURE


Section symbols are: Aachen Student Section (AA), Adelaide (ADE), Alberta (AB), All-Russian State Institute of Cinematography(ARSIC), American River College (ARC), American University (AMU), Argentina (RA), Atlanta (AT), Austrian (AU), Ball StateUniversity (BSU), Belarus (BLS), Belgian (BEL), Belmont University (BU), Berklee College of Music (BCM), Berlin Student(BNS), Bosnia-Herzegovina (BA), Boston (BOS), Brazil (BZ), Brigham Young University (BYU), Brisbane (BRI), British (BR),Bulgarian (BG), Cal Poly San Luis Obispo State University (CPSLO), California State University–Chico (CSU), Carnegie MellonUniversity (CMU), Central German (CG), Central Indiana (CI), Chicago (CH), Chile (RCH), Citrus College (CTC), CogswellPolytechnical College (CPC), Colombia (COL), Colorado (CO), Columbia College (CC), Conservatoire de Paris Student (CPS),Conservatory of Recording Arts and Sciences (CRAS), Croatian (HR), Croatian Student (HRS), Czech (CR), Czech RepublicStudent (CRS), Danish (DA), Danish Student (DAS), Darmstadt (DMS), Denver/Student (DEN/S), Detmold Student (DS), Detroit(DET), District of Columbia (DC), Duquesne University (DU), Düsseldorf (DF), Expression Center for New Media (ECNM),Finnish (FIN), Fredonia (FRE), French (FR), Full Sail Real World Education (FS), Graz (GZ), Greek (GR), Hampton University(HPTU), Hong Kong (HK), Hungarian (HU), Ilmenau (IM), India (IND), Institute of Audio Research (IAR), Israel (IS), Italian(IT), Italian Student (ITS), Japan (JA), Kansas City (KC), Korea (RK), Lithuanian (LT), Long Beach/Student (LB/S), Los Angeles(LA), Louis Lumière (LL), Malaysia (MY), McGill University (MGU), Melbourne (MEL), Mexican (MEX), MichiganTechnological University (MTU), Middle Tennessee State University (MTSU), Moscow (MOS), Music Tech (MT), Nashville (NA),Netherlands (NE), Netherlands Student (NES), New Orleans (NO), New York (NY), North German (NG), Northeast CommunityCollege (NCC), Norwegian (NOR), Ohio University (OU), Pacific Northwest (PNW), Peabody Institute of Johns HopkinsUniversity (PI), Pennsylvania State University (PSU), Philadelphia (PHIL), Philippines (RP), Polish (POL), Portland (POR),Portugal (PT), Ridgewater College, Hutchinson Campus (RC), Romanian (ROM), Russian Academy of Music, Moscow (RAM/S),SAE Nashville (SAENA), St. Louis (STL), St. Petersburg (STP), St. Petersburg Student (STPS), San Diego (SD), San Diego StateUniversity (SDSU), San Francisco (SF), San Francisco State University (SFU), Serbia and Montenegro (SAM), Singapore (SGP),Slovakian Republic (SR), Slovenian (SL), South German (SG), Southwest Texas State University (STSU), Spanish (SPA), StanfordUniversity (SU), Swedish (SWE), Swiss (SWI), Sydney (SYD), Taller de Arte Sonoro, Caracas (TAS), Technical University ofGdansk (TUG), The Art Institute of Seattle (TAIS), Toronto (TOR), Turkey (TR), Ukrainian (UKR), University of Arkansas at PineBluff (UAPB), University of Cincinnati (UC), University of Hartford (UH), University of Illinois at Urbana-Champaign (UIUC),University of Luleå-Piteå (ULP), University of Massachusetts–Lowell (UL), University of Miami (UOM), University of NorthCarolina at Asheville (UNCA), University of Southern California (USC), Upper Midwest (UMW), Uruguay (ROU), Utah (UT),Vancouver (BC), Vancouver Student (BCS), Venezuela (VEN), Vienna (VI), West Michigan (WM), William Paterson University(WPU), Worcester Polytechnic Institute (WPI), Wroclaw University of Technology (WUT).

INFORMATION

MEMBERSHIP

Olalekan S. AgbeniyiBlock 22 Flat 5 Festac Extension, Mile 5,Alaba Express, Lagos, Nigeria

Andres AhlenUppsala University Signals & Systems, P.O.Box 528, SE 75120, Uppsala, Sweden (SWI)

Roy Anjan8-6 South End Gardens, Garia, Kolkata, 700084, India (IND)

Paul D. Apolonio1751 W. Citracado Pkwy. #230, Escondido,CA 92029 (LA)

Ward Archer66 Monroe Ave. 1201, Memphis, TN (NA)

Paulo W. AscencaoRua Senhor Matosinjos L023, PT 4400 304,Gaia, Portugal (PT)

Andrey E. BarabanovBotanicheskaya Str. 313 # 41, RU 198504,St. Petersburg, Russia (STP)

Andres I. BarreraGuafo 1968, San Bernardo, Santiago, Chile(RCH)

Eduardo R. BergalloMartin J. Haedo 3336, B1604Buf, BuenosAires, Argentina (RA)

Cesar E. BoschiMatienzo 1761 San Jose, Guaymallen,Mendoza 5519, Argentina (RA)

Rustin J. BullertElectronic Design Co., 4650 Churchill St.,Shoreview, MN 55126 (UMW)

Cranston L. Burks1457 Stacy Dr., Canton, MI 48180 (DET)

Alberto E. CaglieroDoblas 82 70 Piso, Buenos Aires 1424,Argentina (RA)

Christian Charrier4 rue Boileau, FR 92140, Clamart, France(FR)

Paul M. ChavezHarman Professional Systems, 8400 BalboaBlvd., Northridge, CA 91329 (LA)

Myles Cochran13 John St., London, WC1N 2ED, UK (BR)

Blair B. Collins1016 Norwood Ave., Oakland, CA 94610 (SF)

Justo A. ConchaIrarrazaual 4569, Departmento 410, Nunoa,Santiago, Chile (RCH)

Daniel Leon Felipe CorderoGenaro Garcia #51, Col. Jardin Balbuena,Mexico, CP 15900 (MEX)

Russell A. Corte-Real54 Mohammed Mansion, Gokhale Rd. South,Dadar, Mumbai 400028, India (IND)

David Cisneros CortesAlcocatitla #4, Santa Cruz Acalpixca,Xochimilco, Mexico City, DF 16500, Mexico(MEX)

Jena-Marc CoulonLes Heures Claires 39, BE 1400, Nivelles,Belgium (BEL)

John J. Crawford IVAuburn Engineered Audio & Acoustics, 557Old Stage Rd. Unit 1, Auburn, AL 36830(NA)

Jonathan H. Darling8040 Kostner Ave., Skokie, IL 60076 (CH)

Tim J. Davies22 Aneurin Bevans Way, Maesteg, Bridgend,CF34 0SX, UK (BR)

MEMBERS

These listings represent new membership according to grade.

MEMBERSHIP

INFORMATION

Vinayak B. DeoD/203 Veena Sargam Mahavir Nagar, NearDahanukar Wadi, Kandivli (W), Mumbai 400067, India (IND)

Marco A. DiazUraguay 3247 Recoleta, Santiago, Chile(RCH)

Dalibaeva A. DuysenbekovnaPeresovikov St. 25 #32, RU 195279, St.Petersburg, Russia (STP)

Thomas EanthanakunnelSt. Paul Audiovisuals, 58/23rd Rd., T.P.S. III,Bandra (W), Mumbai 400 050, India (IND)

Victor EspinozaAvenida La Compania #1250, Villa Teniente,Rancague, Chile (RCH)

Johnny V. Evans15621 Braddock Rd., Centreville, VA 20120

Kari EythorssonBolstadarhlid 3, IS 105, Reykjavik, Iceland

Sydney C. FereiraA-206 Chancellors Ct., Vallar Nagar Rd.,Opp. Karnataka Bank Borivili (W), Mumbai400103, India (IND)

Mark K. Fosmeon7235 Goldenrod Ct., Brighton, MI 48116(DET)

Brian FranzHarman International, 39001 W. TwelveMile Rd., Farmington Hills, MI 48331 (DET)

Robert J. FriedrichTelarc International Corporation, 23307Commerce Park Rd., Cleveland, OH 44122

Pramod S. GhaisasFlat No. 14 Swapna Bldg., Ramesh Nagar,Amboli, Andheri (W), Mumbai 400 058,India (IND)

Debasish GhoshalG 26 Bikramaditya arni Bidhan Nagar,Durgapur 12, Burdwan, Bengal 713212, India(IND)

Alfredo M. GonzalezMario Fernando Lopez P #309, Col.Escaudron 201, Iztapalpa, CP 09060, Mexico(MEX)

Gerardo L. GozziParaguay St. No. 1621, 4 Fl. Dept. 10,Buenos Aires, Argentina (RA)

Glynne A. Griffith25045 Champlaign Dr., Southfield, MI 48034(DET)

William H. Groves Jr.53 Melrose St., Arlington, MA 02474 (BOS)

Samie K. Gupta11156 Corte Pleno Verano, San Diego, CA92130 (SD)

Sanjeev K. GuptaM’Audeaus Recording Studio, Ca. Div. ofMukta Arts Ltd., Plot A-18 Opp. LaxmiIndustrial Estate, Andheri (W), Mumbai 58,India (IND)

Delphine Hannotin11 rue Pascal, FR 94250, Gentilly, France(FR)

Thomas G. Hays4347 N. Oak Glen St., Calbasas, CA 91302(LA)

Douglas J. Hood11616 St. Rd. 101, Harlan, IN 46743 (CH)

Jon N. Huether882 Gilbert St., Sebastian, FL 32958

Seon Cheol JangMedia Solution Team, Digital Media R&DCenter, Samsung Electronics Co., Ltd., 416Maeton-Dong, Padal-Gu, Suwon, Kyeonggi-Do 442-742, Korea

Vandana JetleyST Microelectronics Private Ltd., Plot No 2& 3, Sector-16A, Institutional Area, Noida201301, Singapore (SGP)

Horacio AcostaTravel Expert Ltda., Av. Churchill 94 #411,Rio de Janeiro, RJ 20020-050, Brazil (BZ)

Scott Bridgewater9609 Hastings Dr., Silver Spring, MD 20901(DC)

Marco Cavallarovia Bonzi, Ripalta Vecchia, IT 26020,Madignano (CR), Italy (IT)

Praveen S. ChitnisMalati Electronics, Shop No. 10, Silverdale,Nasik 422101, India (IND)

Sergio Dell’Olliovia Crispi n. 58, IT 26013, Crema (CR), Italy(IT)

Pradeep V. Deshpande43-3 Shram Safaiza, Linking Rd. (E24),Santa Cruz (W), Mumbai 400054, India(IND)

Harris Dillon2939 Van Ness St. NW #340, WashingtonD.C. 20008 (DC)

Paul C. Fagundes426 Pursima Ave., Sunnyvale, CA 94086 (SF)

Stephen A. FernandesA3/B-4 Triveni Hsg Society, Om NagarSahar Pipe Line Rd., Andheri (E), Mumbai400099, India (IND)

Gregory N. Fernandes42 Dalal House, National Library Rd.,Bandra (W), Mumbai 400050, India (IND)

Colin M. Franey18 Mackeson Rd., London, NW3 LT, UK(BR)

Dharmaraj Ganapathy15 “Sai-Kripa” Nimbjar Coop. Housing.Soc., Malabar Hill Rd., Mulund Colony,Mulund (W), Mumbai 400082, India (IND)

James W. Gangwer942 32nd St., Richmond, CA 94804 (SF)

Carlos L. GauvronAlberto Vignes 1688, Haedo, Buenos Aires1706, Argentina (RA)

Brian C. Gregory6007 Trailside Dr., Springield, VA 22150

Jack P. Holman11407 Crestbrook Dr., Dallas, TX 75230

Michael M.Y. HuiP. O. Box 54081, Irvine, CA 92619 (SD)

Steve Johnson16255 E. Villanova Pl., Aurora, CO 80013(CO)

Aaron P. Kruse12 Weston Court, Cherry Hill, NJ 08003(NY)

Vishesh Likitkar804 20D Mhada Powai ADS Marg, Mumbai76, India (IND)

Chinmay A. Marshe204 Dheeraj Savera Towers, Off WesternExpress Highway, Near Kharau MillCompound, Borivali (E), Mumbai 400 066,India (IND)

Rakesh L. NalavadeN. 37 Ramkutir Plot No. 133, L.J. Road,Mahim, Mumbai 400016, India (IND)

Lou R. NormandeauCanadian Boradcasting Co., P. O. Box 4600,Vancouver, V6B 4A2, British Columbia,Canada (BC)

Setu M. Adams612 Fairview Ave., Murfreesboro, TN 37130(MTSU)

Vinodkumar R. Agarwal602 Satyam 1 Raheja Complex, Malad East,Mumbai 400092, India

Galina Alekseeva7 Lefortovskii val # 39, RU 111116,Moscow, Russia (ARSIC)

Julian M. AlonsoSenillosa 33 3-C, Buenos Aires, CP 1424,Argentina

Eric M. Amendt449 Vista Roma, Newport Beach, CA 92660(CPSLO)

Patrick Andersen594 E. Elizabeth St., Pasadena CA 91104(CTC)

Brian A. Anderson3176 A St., San Diego, CA 92102 (CPSLO)

Maria I. Arango GiraldoCalle 92 Numero 18-32, Apartmento 101,Bogota, DC, Colombia

Jonathan C. Arnold1301 Main St., MTSU Box 2679,Murfreesboro, TN 37132 (MTSU)

Elizabeth V. BarbotPushkin Detskoselki Bulvar 9 # 116, RU189620, St. Petersburg, Russia (STPS)

STUDENTS

ASSOCIATES


In Memoriam

Rex Isom, AES honorary mem-ber, died 11 January 2003, atthe age of 92. He was born in

Mitchell, IN, and earned a bachelor’sdegree from Butler University. Laterstudies at both George WashingtonUniversity and Harvard led to mas-ter’s degrees. He was pursuing a doc-torate when World War II interruptedhis schooling. After military servicehe joined RCA and was engaged inthe development of recording systemsfor film and video as well as telecineproducts, primarily in the New Jersey facility. The method of film pull-down for adapting 24-frame/secondfilm to the 30 frame/second video rateis something that we have all seen —just about every day of our lives —since the 1950s.

When H. E. Roys retired fromRCA Records, Rex was chosen as thenew director of RCA’s Record Engi-neering group in Indianapolis, a position he held from 1966 to his retirement in 1976. Rex held 21patents and was active in internationalstandardization work. He was a mem-ber of Phi Kappa Phi and a fellow ofthe Institute of Electrical and Elec-tronics Engineers, Society of MotionPicture and Television Engineers, andthe Audio Engineering Society. Hewas president of the AES in 1976-77.

It is easy to overlook the era inwhich Rex made his mark in record-ing technology. During the quartercentury after the war, the majorrecord companies worldwide werevertically integrated, complete withtheir own development engineeringgroups. Along with such men asWilliam Bachman of ColumbiaRecords and Ed Uecke of CapitolRecords in the U. S., Rex was respon-sible at RCA Records for develop-mental and experimental work ineverything from plant process controlto studio electronics and consolecomponentry.

One of Rex’s projects was the boldmove at RCA Victor to replace thetraditional 135-gram LP with one

weighing only 80 grams. The notionwas laughed at by many traditional-ists — until Rex demonstrated thatthe smaller profile cross-section actu-ally improved the molding propertiesof vinyl by increasing the particle rateof compound flow during the pressingcycle. The 30% savings in materialmeant that the plants could use premi-um material for all processing, furtherimproving the finished LP product,both standard and CD-4 Quadraphon-ic. The major problem was an initialincrease in disc warpage, which wassolved through new handling process-es on the press floor.

Another thorny problem tackled byRex was replacing the venerable RCA“tape tree,” which had been used forall high-speed duplicating activitiessince the introduction of prerecordedtape in the mid-1950s. That systemhad imposed considerable air drag onthe duplicating master tape and couldnot operate reliably at high speeds.The high duplicating speed ratios,which had been make possible byelectronic and magnetic headimprovements, prompted RCA and

others to move to random tape binsfor buffering the master tapes. Rexand others at the Indianapolis plantcame up with their own high perfor-mance version.

Many AES members will rememberthe monumental 300-page CentennialIssue of the AES Journal…ThePhonograph and Sound Recording After 100 Years (1977 October/November). Rex was the guest editor.He said that it occupied him for theentire year that he retired from RCAVictor. One could always expect ofRex anything from a whimsical bonmot to a sage observation. One day atlunch during an AES Convention inParis, he said it would be a shame tohave come this far and not see themost glorious stained glass in theworld. One hour later, three of us werein a car on our way to Chartres Cathe-dral — returning that evening in timefor the AES Banquet.

Rex is survived by two sons andthree grandchildren. His wife Ruthdied a year ago.

John EargleLos Angeles, CA


Warren Rex Isom 1910-2003


EASTERN REGION,USA/CANADA

Vice President:Jim Anderson12 Garfield PlaceBrooklyn, NY 11215Tel. +1 718 369 7633Fax +1 718 669 7631E-mail [email protected]

UNITED STATES OFAMERICA

CONNECTICUTUniversity of HartfordSection (Student)Howard A. CanistraroFaculty AdvisorAES Student SectionUniversity of HartfordWard College of Technology200 Bloomfield Ave.West Hartford, CT 06117Tel. +1 860 768 5358Fax +1 860 768 5074 E-mail [email protected]

FLORIDAFull Sail Real WorldEducation Section (Student)Bill Smith, Faculty AdvisorAES Student SectionFull Sail Real World Education3300 University Blvd., Suite 160Winter Park, FL 327922Tel. +1 800 679 0100E-mail [email protected]

University of Miami Section(Student)Ken Pohlmann, Faculty AdvisorAES Student SectionUniversity of MiamiSchool of MusicPO Box 248165Coral Gables, FL 33124-7610Tel. +1 305 284 6252Fax +1 305 284 4448E-mail [email protected]

GEORGIA

Atlanta SectionRobert Mason2712 Leslie Dr.Atlanta, GA 30345Home Tel. +1 770 908 1833E-mail [email protected]

MARYLAND

Peabody Institute of JohnsHopkins University Section(Student)

Neil Shade, Faculty AdvisorAES Student SectionPeabody Institute of Johns

Hopkins UniversityRecording Arts & Science Dept.2nd Floor Conservatory Bldg.1 E. Mount Vernon PlaceBaltimore, MD 21202Tel. +1 410 659 8100 ext. 1226E-mail [email protected]

MASSACHUSETTS

Berklee College of MusicSection (Student)Eric Reuter, Faculty AdvisorBerklee College of MusicAudio Engineering Societyc/o Student Activities1140 Boylston St., Box 82Boston, MA 02215Tel. +1 617 747 8251Fax +1 617 747 2179E-mail [email protected]

Boston SectionJ. Nelson Chadderdonc/o Oceanwave Consulting, Inc.21 Old Town Rd.Beverly, MA 01915Tel. +1 978 232 9535 x201Fax +1 978 232 9537E-mail [email protected]

University of Massachusetts–Lowell Section (Student)John Shirley, Faculty AdvisorAES Student ChapterUniversity of Massachusetts–LowellDept. of Music35 Wilder St., Ste. 3Lowell, MA 01854-3083Tel. +1 978 934 3886Fax +1 978 934 3034E-mail [email protected]

Worcester PolytechnicInstitute Section (Student) William MichalsonFaculty AdvisorAES Student SectionWorcester Polytechnic Institute100 Institute Rd.Worcester, MA 01609Tel. +1 508 831 5766E-mail [email protected]

NEW JERSEY

William Paterson UniversitySection (Student)David Kerzner, Faculty AdvisorAES Student SectionWilliam Paterson University300 Pompton Rd.Wayne, NJ 07470-2103

Tel. +1 973 720 3198Fax +1 973 720 2217E-mail [email protected]

NEW YORK

Fredonia Section (Student)Bernd Gottinger, Faculty AdvisorAES Student SectionSUNY–Fredonia1146 Mason HallFredonia, NY 14063Tel. +1 716 673 4634Fax +1 716 673 3154E-mail [email protected]

Institute of Audio ResearchSection (Student)Noel Smith, Faculty AdvisorAES Student SectionInstitute of Audio Research 64 University Pl.New York, NY 10003Tel. +1 212 677 7580Fax +1 212 677 6549E-mail [email protected]

New York SectionRobbin L. GheeslingBroadness, LLC265 Madison Ave., Second FloorNew York, NY 10016Tel. +1 212 818 1313Fax +1 212 818 1330E-mail [email protected]

NORTH CAROLINA

University of North Carolinaat Asheville Section (Student)Wayne J. KirbyFaculty AdvisorAES Student SectionUniversity of North Carolina at

AshevilleDept. of MusicOne University HeightsAsheville, NC 28804Tel. +1 828 251 6487Fax +1 828 253 4573E-mail [email protected]

PENNSYLVANIA

Carnegie Mellon UniversitySection (Student)Thomas SullivanFaculty AdvisorAES Student SectionCarnegie Mellon UniversityUniversity Center Box 122Pittsburg, PA 15213Tel. +1 412 268 3351E-mail [email protected]

Duquesne University Section(Student)Francisco Rodriguez

Faculty AdvisorAES Student SectionDuquesne UniversitySchool of Music600 Forbes Ave.Pittsburgh, PA 15282Tel. +1 412 434 1630Fax +1 412 396 5479E-mail [email protected]

Pennsylvania State UniversitySection (Student)Dan ValenteAES Penn State Student ChapterGraduate Program in Acoustics217 Applied Science Bldg.University Park, PA 16802Home Tel. +1 814 863 8282Fax +1 814 865 3119E-mail [email protected]

Philadelphia SectionRebecca MercuriP.O. Box 1166.Philadelphia, PA 19105Tel. +1 609 895 1375E-mail [email protected]

VIRGINIA

Hampton University Section(Student)Bob Ransom, Faculty AdvisorAES Student SectionHampton UniversityDept. of MusicHampton, VA 23668Office Tel. +1 757 727 5658,

+1 757 727 5404Home Tel. +1 757 826 0092Fax +1 757 727 5084E-mail [email protected]

WASHINGTON, DC

American University Section(Student)Benjamin TomassettiFaculty AdvisorAES Student SectionAmerican UniversityPhysics Dept.4400 Massachusetts Ave., N.W.Washington, DC 20016Tel. +1 202 885 2746Fax +1 202 885 2723E-mail [email protected]

District of Columbia SectionJohn W. ReiserDC AES Section SecretaryP.O. Box 169Mt. Vernon, VA 22121-0169Tel. +1 703 780 4824Fax +1 703 780 4214E-mail [email protected]

DIRECTORY

SECTIONS CONTACTS

The following is the latest information we have available for our sections contacts. If youwish to change the listing for your section, please mail, fax or e-mail the new informationto: Mary Ellen Ilich, AES Publications Office, Audio Engineering Society, Inc., 60 East42nd Street, Suite 2520, New York, NY 10165-2520, USA. Telephone +1 212 661 8528.Fax +1 212 661 7829. E-mail [email protected].

Updated information that is received by the first of the month will be published in thenext month’s Journal. Please help us to keep this information accurate and timely.

CANADAMcGill University Section(Student)John Klepko, Faculty AdvisorAES Student SectionMcGill UniversitySound Recording StudiosStrathcona Music Bldg.555 Sherbrooke St. W.Montreal, Quebec H3A 1E3CanadaTel. +1 514 398 4535 ext. 0454E-mail [email protected]

Toronto SectionAnne Reynolds606-50 Cosburn Ave.Toronto, Ontario M4K 2G8CanadaTel. +1 416 957 6204Fax +1 416 364 1310E-mail [email protected]

CENTRAL REGION,USA/CANADA

Vice President:Jim KaiserMaster Mix1921 Division St.Nashville, TN 37203Tel. +1 615 321 5970Fax +1 615 321 0764E-mail [email protected]


ARKANSAS

University of Arkansas atPine Bluff Section (Student)Robert Elliott, Faculty AdvisorAES Student SectionMusic Dept. Univ. of Arkansasat Pine Bluff1200 N. University DrivePine Bluff, AR 71601Tel. +1 870 575 8916Fax +1 870 543 8108E-mail [email protected]

INDIANA

Ball State University Section(Student)Michael Pounds, Faculty AdvisorAES Student SectionBall State UniversityMET Studios2520 W. BethelMuncie, IN 47306Tel. +1 765 285 5537Fax +1 765 285 8768E-mail [email protected]

Central Indiana SectionJames LattaSound Around6349 Warren Ln.Brownsburg, IN 46112Office Tel. +1 317 852 8379Fax +1 317 858 8105E-mail [email protected]

ILLINOIS

Chicago SectionRobert Zurek

Motorola2001 N. Division St.Harvard, IL 60033Tel. +1 815 884 1361Fax +1 815 884 2519E-mail [email protected]

Columbia College Section(Student)Dominique J. ChéenneFaculty AdvisorAES Student Section676 N. LaSalle, Ste. 300Chicago, IL 60610Tel. +1 312 344 7802Fax +1 312 482 9083

University of Illinois atUrbana-Champaign Section(Student)David S. Petruncio Jr.AES Student SectionUniversity of Illinois, Urbana-

ChampaignUrbana, IL 61801Tel. +1 217 621 7586E-mail [email protected]

KANSAS

Kansas City SectionJim MitchellCustom Distribution Limited12301 Riggs Rd.Overland Park, KS 66209Tel. +1 913 661 0131Fax +1 913 663 5662

LOUISIANA

New Orleans SectionJoseph DohertyFactory Masters4611 Magazine St.New Orleans, LA 70115Tel. +1 504 891 4424Cell +1 504 669 4571Fax +1 504 899 9262

MICHIGAN

Detroit SectionTom ConlinDaimlerChryslerE-mail [email protected]

Michigan TechnologicalUniversity Section (Student)Andre LaRoucheAES Student SectionMichigan Technological

UniversityElectrical Engineering Dept.1400 Townsend Dr.Houghton, MI 49931Home Tel. +1 906 847 9324E-mail [email protected]

West Michigan SectionCarl HordykCalvin College3201 Burton S.E.Grand Rapids, MI 49546Tel. +1 616 957 6279Fax +1 616 957 6469E-mail [email protected]

MINNESOTA

Music Tech College Section

(Student)Michael McKernFaculty AdvisorAES Student SectionMusic Tech College19 Exchange Street EastSaint Paul, MN 55101Tel. +1 651 291 0177Fax +1 651 291 [email protected]

Ridgewater College,Hutchinson Campus Section(Student)Dave Igl, Faculty AdvisorAES Student SectionRidgewater College, Hutchinson

Campus2 Century Ave. S.E.Hutchinson, MN 55350E-mail [email protected]

Upper Midwest SectionGreg ReiersonRare Form Mastering4624 34th Avenue SouthMinneapolis, MN 55406Tel. +1 612 327 8750E-mail [email protected]

MISSOURI

St. Louis SectionJohn Nolan, Jr.693 Green Forest Dr.Fenton, MO 63026Tel./Fax +1 636 343 4765E-mail [email protected]

NEBRASKA

Northeast Community CollegeSection (Student)Anthony D. BeardsleeFaculty AdvisorAES Student SectionNortheast Community CollegeP.O. Box 469Norfolk, NE 68702Tel. +1 402 844 7365Fax +1 209 254 8282E-mail [email protected]

OHIO

Ohio University Section(Student)Erin M. DawesAES Student SectionOhio UniversityRTVC Bldg.9 S. College St.Athens, OH 45701-2979Home Tel. +1 740 597 6608E-mail [email protected]

University of CincinnatiSection (Student)Thomas A. HainesFaculty AdvisorAES Student SectionUniversity of CincinnatiCollege-Conservatory of MusicM.L. 0003Cincinnati, OH 45221

Tel. +1 513 556 9497Fax +1 513 556 0202

TENNESSEE

Belmont University Section(Student)Wesley Bulla, Faculty AdvisorAES Student SectionBelmont UniversityNashville, TN 37212

Middle Tennessee StateUniversity Section (Student)Phil Shullo, Faculty AdvisorAES Student SectionMiddle Tennessee State University301 E. Main St., Box 21Murfreesboro, TN 37132Tel. +1 615 898 2553E-mail [email protected]

Nashville Section Tom EdwardsMTV Networks330 Commerce St.Nashville, TN 37201Tel. +1 615 335 8520Fax +1 615 335 8608E-mail [email protected]

SAE Nashville Section (Student)Larry Sterling, Faculty AdvisorAES Student Section7 Music Circle N.Nashville, TN 37203Tel. +1 615 244 5848Fax +1 615 244 3192E-mail [email protected]

TEXAS

Southwest Texas StateUniversity Section (Student)Mark C. EricksonFaculty AdvisorAES Student Section Southwest Texas State

University224 N. Guadalupe St.San Marcos, TX 78666Tel. +1 512 245 8451Fax +1 512 396 1169E-mail [email protected]

WESTERN REGION,USA/CANADA

Vice President:Bob MosesIsland Digital Media Group,

LLC26510 Vashon Highway S.W.Vashon, WA 98070Tel. +1 206 463 6667Fax +1 810 454 5349E-mail [email protected]


ARIZONA

Conservatory of TheRecording Arts and Sciences

SECTIONS CONTACTSDIRECTORY


Section (Student)Glen O’Hara, Faculty AdvisorAES Student Section Conservatory of The Recording

Arts and Sciences2300 E. Broadway Rd.Tempe, AZ 85282Tel. +1 480 858 9400, 800 562

6383 (toll-free)Fax +1 480 829 [email protected]

CALIFORNIA

American River CollegeSection (Student)Eric Chun, Faculty AdvisorAES Student SectionAmerican River College Chapter4700 College Oak Dr.Sacramento, CA 95841Tel. +1 916 484 8420E-mail [email protected]

Cal Poly San Luis ObispoState University Section(Student)Jerome R. BreitenbachFaculty AdvisorAES Student SectionCalifornia Polytechnic State

UniversityDept. of Electrical EngineeringSan Luis Obispo, CA 93407Tel. +1 805 756 5710Fax +1 805 756 1458E-mail [email protected]

California State University–Chico Section (Student)Keith Seppanen, Faculty AdvisorAES Student SectionCalifornia State University–Chico400 W. 1st St.Chico, CA 95929-0805Tel. +1 530 898 5500E-mail [email protected]

Citrus College Section(Student)Gary Mraz, Faculty AdvisorAES Student SectionCitrus CollegeRecording Arts1000 W. Foothill Blvd.Glendora, CA 91741-1899Fax +1 626 852 8063

Cogswells PolytechnicalCollege Section (Student)Tim Duncan, Faculty SponsorAES Student SectionCogswell Polytechnical CollegeMusic Engineering Technology1175 Bordeaux Dr.Sunnyvale, CA 94089Tel. +1 408 541 0100, ext. 130Fax +1 408 747 0764E-mail [email protected]

Expression Center for NewMedia Section (Student)Scott Theakston, Faculty AdvisorAES Student SectionEx’pression Center for New

Media

6601 Shellmount St.Emeryville, CA 94608Tel. +1 510 654 2934Fax +1 510 658 3414E-mail [email protected]

Long Beach City CollegeSection (Student)Nancy Allen, Faculty AdvisorAES Student SectionLong Beach City College4901 E. Carson St.Long Beach, CA 90808Tel. +1 562 938 4312Fax +1 562 938 4409E-mail [email protected]

Los Angeles SectionAndrew Turner1733 Lucile Ave., #8Los Angeles, CA 90026Tel. +1 323 661 0390E-mail [email protected]

San Diego SectionJ. Russell Lemon2031 Ladera Ct.Carlsbad, CA 92009-8521Home Tel. +1 760 753 2949E-mail [email protected]

San Diego State UniversitySection (Student)John Kennedy, Faculty AdvisorAES Student SectionSan Diego State UniversityElectrical & Computer

Engineering Dept.5500 Campanile Dr.San Diego, CA 92182-1309Tel. +1 619 594 1053Fax +1 619 594 2654E-mail [email protected]

San Francisco SectionBill Orner1513 Meadow LaneMountain View, Ca 94040Tel. +1 650 903 0301Fax +1 650 903 0409E-mail [email protected]

San Francisco StateUniversity Section (Student)John Barsotti, Faculty AdvisorAES Student SectionSan Francisco State UniversityBroadcast and Electronic

Communication Arts Dept.1600 Halloway Ave.San Francisco, CA 94132Tel. +1 415 338 1507E-mail [email protected]

Stanford University Section(Student)Jay Kadis, Faculty AdvisorStanford AES Student SectionStanford UniversityCCRMA/Dept. of MusicStanford, CA 94305-8180Tel. +1 650 723 4971Fax +1 650 723 8468E-mail [email protected]

University of SouthernCalifornia Section(Student)Kenneth LopezFaculty AdvisorAES Student SectionUniversity of Southern California840 W. 34th St.Los Angeles, CA 90089-0851Tel. +1 213 740 3224Fax +1 213 740 3217E-mail [email protected]

COLORADO

Colorado SectionRobert F. MahoneyRobert F. Mahoney &

Associates310 Balsam Ave.Boulder, CO 80304Tel. +1 303 443 2213Fax +1 303 443 6989E-mail [email protected]

Denver Section (Student)Roy Pritts, Faculty AdvisorAES Student SectionUniversity of Colorado at

DenverDept. of Professional StudiesCampus Box 162P.O. Box 173364Denver, CO 80217-3364Tel. +1 303 556 2795Fax +1 303 556 2335E-mail [email protected]

OREGON

Portland SectionTony Dal MolinAudio Precision, Inc.5750 S.W. Arctic Dr.Portland, OR 97005Tel. +1 503 627 0832Fax +1 503 641 8906E-mail [email protected]

UTAH

Brigham Young UniversitySection (Student)Timothy Leishman,

Faculty AdvisorBYU-AES Student SectionDepartment of Physics andAstronomy Brigham Young UniversityProvo, UT 84602Tel. +1 801 422 4612E-mail [email protected]

Utah SectionDeward Timothyc/o Poll Sound4026 S. MainSalt Lake City, UT 84107Tel. +1 801 261 2500Fax +1 801 262 7379

WASHINGTON

Pacific Northwest SectionGary LouieUniversity of Washington

School of Music

PO Box 353450Seattle, WA 98195Office Tel. +1 206 543 1218Fax +1 206 685 9499E-mail [email protected]

The Art Institute of SeattleSection (Student)David G. ChristensenFaculty AdvisorAES Student SectionThe Art Institute of Seattle2323 Elliott Ave.Seattle, WA 98121-1622 Tel. +1 206 448 [email protected]

CANADA

Alberta SectionFrank LockwoodAES Alberta SectionSuite 404815 - 50 Avenue S.W.Calgary, Alberta T2S 1H8CanadaHome Tel. +1 403 703 5277Fax +1 403 762 6665E-mail [email protected]

Vancouver SectionPeter L. JanisC-Tec #114, 1585 BroadwayPort Coquitlam, B.C. V3C 2M7CanadaTel. +1 604 942 1001Fax +1 604 942 1010E-mail [email protected]

Vancouver Student SectionGregg Gorrie, Faculty AdvisorAES Greater Vancouver

Student SectionCentre for Digital Imaging and

Sound3264 Beta Ave.Burnaby, B.C. V5G 4K4, CanadaTel. +1 604 298 [email protected]

NORTHERN REGION,EUROPE

Vice President:Søren BechBang & Olufsen a/sCoreTechPeter Bangs Vej 15DK-7600 Struer, DenmarkTel. +45 96 84 49 62Fax +45 97 85 59 [email protected]

BELGIUM

Belgian SectionHermann A. O. WilmsAES Europe Region OfficeZevenbunderslaan 142, #9BE-1190 Vorst-Brussels, Belgium




Tel. +32 2 345 7971Fax +32 2 345 3419

DENMARK

Danish SectionKnud Bank ChristensenSkovvej 2DK-8550 Ryomgård, DenmarkTel. +45 87 42 71 46Fax +45 87 42 70 10E-mail [email protected]

Danish Student SectionKnud Bank ChristensenSkovvej 2DK-8550 Ryomgård, DenmarkTel. +45 87 42 71 46Fax +45 87 42 70 10E-mail [email protected]

FINLAND

Finnish SectionKalle KoivuniemiNokia Research CenterP.O. Box 100FI-33721 Tampere, FinlandTel. +358 7180 35452Fax +358 7180 35897E-mail [email protected]

NETHERLANDS

Netherlands SectionRinus BooneVoorweg 105ANL-2715 NG ZoetermeerNetherlandsTel. +31 15 278 14 71, +31 62

127 36 51Fax +31 79 352 10 08E-mail [email protected]

Netherlands Student SectionDirk FischerAES Student SectionGroenewegje 143aDen Haag, NetherlandsHome Tel. +31 70 [email protected]

NORWAY

Norwegian SectionJan Erik JensenNøklesvingen 74NO-0689 Oslo, NorwayOffice Tel. +47 22 24 07 52Home Tel. +47 22 26 36 13 Fax +47 22 24 28 06E-mail [email protected]

RUSSIA

All-Russian State Institute ofCinematography Section(Student)Leonid Sheetov, Faculty SponsorAES Student SectionAll-Russian State Institute of

Cinematography (VGIK)W. Pieck St. 3RU-129226 Moscow, RussiaTel. +7 095 181 3868Fax +7 095 187 7174E-mail [email protected]

Moscow SectionMichael LannieResearch Institute for

Television and RadioAcoustic Laboratory12-79 Chernomorsky bulvarRU-113452 Moscow, RussiaTel. +7 095 2502161, +7 095

1929011Fax +7 095 9430006E-mail [email protected]

St. Petersburg SectionIrina A. AldoshinaSt. Petersburg University of

TelecommunicationsGangutskaya St. 16, #31RU-191187 St. Petersburg

RussiaTel. +7 812 272 4405Fax +7 812 316 1559E-mail [email protected]

St. Petersburg Student SectionNatalia V. TyurinaFaculty AdvisorProsvescheniya pr., 41, 185RU-194291 St. Petersburg, RussiaTel. +7 812 595 1730Fax +7 812 316 [email protected]

SWEDEN

Swedish SectionIngemar OhlssonAudio Data Lab ABKatarinavägen 22SE-116 45 Stockholm, SwedenTel. +46 8 644 5865Fax +46 8 641 6791E-mail [email protected]

University of Luleå-PiteåSection (Student)Lars Hallberg, Faculty SponsorAES Student SectionUniversity of Luleå-PiteåSchool of MusicBox 744S-94134 Piteå, SwedenTel. +46 911 726 27Fax +46 911 727 10E-mail [email protected]

UNITED KINGDOM

British SectionHeather LaneAudio Engineering SocietyP.O. Box 645Slough GB-SL1 8BJUnited KingdomTel. +44 1628 663725Fax +44 1628 667002E-mail [email protected]

CENTRAL REGION,EUROPE

Vice President:Markus ErneScopein ResearchSonnmattweg 6CH-5000 Aarau, Switzerland

Tel. +41 62 825 09 19Fax +41 62 825 09 [email protected]

AUSTRIA

Austrian SectionFranz LechleitnerLainergasse 7-19/2/1AT-1230 Vienna, AustriaOffice Tel. +43 1 4277 29602Fax +43 1 4277 9296E-mail [email protected]

Graz Section (Student)Robert Höldrich Faculty SponsorInstitut für Elektronische Musik

und AkustikInffeldgasse 10AT-8010 Graz, AustriaTel. +43 316 389 3172Fax +43 316 389 3171E-mail [email protected]

Vienna Section (Student)Jürg Jecklin, Faculty SponsorVienna Student SectionUniversität für Musik und

Darstellende Kunst WienInstitut für Elektroakustik und

Experimentelle MusikRienösslgasse 12AT-1040 Vienna, AustriaTel. +43 1 587 3478Fax +43 1 587 3478 20E-mail [email protected]

CZECH REPUBLIC

Czech SectionJiri OcenasekDejvicka 36CZ-160 00 Prague 6Czech Republic Home Tel. +420 2 24324556E-mail [email protected]

Czech Republic StudentSectionLibor Husník, Faculty AdvisorAES Student SectionCzech Technical University at

PragueTechnická 2, CZ-116 27 Prague 6Czech RepublicTel. +420 2 2435 2115E-mail [email protected]

GERMANY

Aachen Section (Student)Michael VorländerFaculty AdvisorInstitut für Technische AkustikRWTH AachenTemplergraben 55D-52065 Aachen, GermanyTel. +49 241 807985Fax +49 241 8888214E-mail [email protected]

Berlin Section (Student)Bernhard Güttler Zionskirchstrasse 14DE-10119 Berlin, Germany

Tel. +49 30 4404 72 19Fax +49 30 4405 39 03E-mail [email protected]

Central German SectionErnst-Joachim VölkerInstitut für Akustik und

BauphysikKiesweg 22-24DE-61440 Oberursel, GermanyTel. +49 6171 75031Fax +49 6171 85483E-mail [email protected]

Darmstadt Section (Student)G. M. Sessler, Faculty SponsorAES Student SectionTechnical University of

DarmstadtInstitut für ÜbertragungstechnikMerkstr. 25DE-64283 Darmstadt, GermanyTel. +49 6151 [email protected]

Detmold Section (Student)Andreas Meyer, Faculty SponsorAES Student Sectionc/o Erich Thienhaus InstitutTonmeisterausbildung

Hochschule für Musik Detmold

Neustadt 22, DE-32756Detmold, GermanyTel/Fax +49 5231 975639E-mail [email protected]

Düsseldolf Section (Student)Ludwig KuglerAES Student SectionBilker Allee 126DE-40217 Düsseldorf, GermanyTel. +49 211 3 36 80 [email protected]

Ilmenau Section (Student)Karlheinz BrandenburgFaculty SponsorAES Student SectionInstitut für MedientechnikPF 10 05 65DE-98684 Ilmenau, GermanyTel. +49 3677 69 2676Fax +49 3677 69 [email protected]

North German SectionReinhard O. SahrEickhopskamp 3DE-30938 Burgwedel, GermanyTel. +49 5139 4978Fax +49 5139 5977E-mail [email protected]

South German SectionGerhard E. PicklappLandshuter Allee 162DE-80637 Munich, GermanyTel. +49 89 15 16 17Fax +49 89 157 10 31E-mail [email protected]


HUNGARY

Hungarian SectionIstván MatókRona u. 102. II. 10HU-1149 Budapest, HungaryHome Tel. +36 30 900 1802Fax +36 1 383 24 81E-mail [email protected]

LITHUANIA

Lithuanian SectionVytautas J. StauskisVilnius Gediminas Technical

UniversityTraku 1/26, Room 112LT-2001 Vilnius, LithuaniaTel. +370 5 262 91 78Fax +370 5 261 91 44E-mail [email protected]

POLAND

Polish SectionJan A. AdamczykUniversity of Mining and

MetallurgyDept. of Mechanics and

Vibroacousticsal. Mickiewicza 30PL-30 059 Cracow, PolandTel. +48 12 617 30 55Fax +48 12 633 23 14E-mail [email protected]

Technical University of GdanskSection (Student)Pawel ZwanAES Student Section Technical University of GdanskSound Engineering Dept.ul. Narutowicza 11/12PL-80 952 Gdansk, PolandHome Tel. +48 58 347 23 98Office Tel. +4858 3471301Fax +48 58 3471114E-mail [email protected]

Wroclaw University ofTechnology Section (Student)Andrzej B. DobruckiFaculty SponsorAES Student SectionInstitute of Telecommunications

and AcousticsWroclaw Univ.TechnologyWybrzeze Wyspianskiego 27PL-503 70 Wroclaw, PolandTel. +48 71 320 30 68Fax +48 71 320 31 89E-mail [email protected]

REPUBLIC OF BELARUS

Belarus SectionValery ShalatoninBelarusian State University of

Informatics and Radioelectronics

vul. Petrusya Brouki 6BY-220027 MinskRepublic of BelarusTel. +375 17 239 80 95Fax +375 17 231 09 14E-mail [email protected]

SLOVAK REPUBLIC

Slovakian Republic SectionRichard VarkondaCentron Slovakia Ltd.Podhaj 107SK-841 03 BratislavaSlovak RepublicTel. +421 7 6478 0767Fax. +421 7 6478 [email protected]

SWITZERLAND

Swiss SectionJoël GodelAES Swiss SectionSonnmattweg 6CH-5000 AarauSwitzerlandE-mail [email protected]

UKRAINE

Ukrainian SectionValentin AbakumovNational Technical University

of UkraineKiev Politechnical InstitutePolitechnical St. 16Kiev UA-56, UkraineTel./Fax +38 044 2366093

SOUTHERN REGION,EUROPE

Vice President:Daniel ZalayConservatoire de ParisDept. SonFR-75019 Paris, FranceOffice Tel. +33 1 40 40 46 14Fax +33 1 40 40 47 [email protected]

BOSNIA-HERZEGOVINA

Bosnia-Herzegovina SectionJozo TalajicBulevar Mese Selimovica 12BA-71000 SarajevoBosnia–HerzegovinaTel. +387 33 455 160Fax +387 33 455 163E-mail [email protected]

BULGARIA

Bulgarian SectionKonstantin D. KounovBulgarian National RadioTechnical Dept.4 Dragan Tzankov Blvd. BG-1040 Sofia, BulgariaTel. +359 2 65 93 37, +359 2

9336 6 01Fax +359 2 963 1003E-mail [email protected]

CROATIA

Croatian SectionSilvije StamacHrvatski RadioPrisavlje 3HR-10000 Zagreb, CroatiaTel. +385 1 634 28 81Fax +385 1 611 58 29E-mail [email protected]

Croatian Student SectionHrvoje DomitrovicFaculty AdvisorAES Student SectionFaculty of Electrical

Engineering and ComputingDept. of Electroaocustics (X. Fl.)Unska 3HR-10000 Zagreb, CroatiaTel. +385 1 6129 640Fax +385 1 6129 [email protected]

FRANCE

Conservatoire de ParisSection (Student)Alessandra Galleron36, Ave. ParmentierFR-75011 Paris, FranceTel. +33 1 43 38 15 94

French SectionMichael WilliamsIle du Moulin62 bis Quai de l’Artois FR-94170 Le Perreux sur

Marne, FranceTel. +33 1 48 81 46 32Fax +33 1 47 06 06 48E-mail [email protected]

Louis Lumière Section(Student)Alexandra Carr-BrownAES Student SectionEcole Nationale Supérieure

Louis Lumière7, allée du Promontoire, BP 22FR-93161 Noisy Le Grand

Cedex, FranceTel. +33 6 18 57 84 41E-mail [email protected]

GREECE

Greek SectionVassilis TsakirisCrystal AudioAiantos 3a VrillissiaGR 15235 Athens, GreeceTel. + 30 2 10 6134767Fax + 30 2 10 6137010E-mail [email protected]

ISRAEL

Israel SectionBen Bernfeld Jr.H. M. Acustica Ltd.20G/5 Mashabim St..IL-45201 Hod Hasharon, IsraelTel./Fax +972 9 7444099E-mail [email protected]

ITALY

Italian SectionCarlo Perrettac/o AES Italian SectionPiazza Cantore 10IT-20134 Milan, ItalyTel. +39 338 9108768Fax +39 02 58440640E-mail [email protected]

Italian Student SectionFranco Grossi, Faculty AdvisorAES Student SectionViale San Daniele 29 IT-33100 Udine, ItalyTel. +39 [email protected]

PORTUGAL

Portugal SectionRui Miguel Avelans CoelhoR. Paulo Renato 1, 2APT-2745-147 Linda-a-VelhaPortugalTel. +351 214145827E-mail [email protected]

ROMANIA

Romanian SectionMarcia TaiachinRadio Romania60-62 Grl. Berthelot St.RO-79756 Bucharest, RomaniaTel. +40 1 303 12 07Fax +40 1 222 69 19

SERBIA AND MONTENEGRO

Serbia and Montenegro SectionTomislav StanojevicSava centreM. Popovica 9YU-11070 Belgrade, YugoslaviaTel. +381 11 311 1368Fax +38111 605 [email protected]

SLOVENIA

Slovenian SectionTone SeliskarRTV SlovenijaKolodvorska 2SI-1550 Ljubljana, SloveniaTel. +386 61 175 2708Fax +386 61 175 2710E-mail [email protected]

SPAIN

Spanish SectionJuan Recio MorillasSpanish SectionC/Florencia 14 3oDES-28850 Torrejon de Ardoz

(Madrid), SpainTel. +34 91 540 14 03E-mail [email protected]




TURKEY

Turkish SectionSorgun AkkorSTDGazeteciler Sitesi, Yazarlar

Sok. 19/6Esentepe 80300 Istanbul, TurkeyTel. +90 212 2889825Fax +90 212 2889831E-mail [email protected]

LATIN AMERICAN REGION

Vice President:Mercedes OnoratoTalcahuano 141Buenos Aires, ArgentinaTel./Fax +5411 4 375 [email protected]

ARGENTINA

Argentina SectionHernan Ranucci Talcahuano 141Buenos Aires, ArgentinaTel./Fax +5411 4 375 0116E-mail [email protected]

BRAZIL

Brazil SectionRosalfonso BortoniRua Doutor Jesuíno Maciel,

1584/22Campo BeloSão Paulo, SP, Brazil 04615-004Tel.+55 11 5533-3970Fax +55 21 2421 0112E-mail [email protected]

CHILE

Chile SectionAndres SchmidtHernan Cortes 2768Ñuñoa, Santiago de ChileTel. +56 2 4249583E-mail [email protected]

COLOMBIA

Colombia SectionTony Penarredonda CaraballoCarrera 51 #13-223Medellin, ColombiaTel. +57 4 265 7000Fax +57 4 265 2772E-mail [email protected]

MEXICO

Mexican SectionJavier Posada Div. Del Norte #1008Col. Del ValleMexico, D.F. MX-03100MexicoTel. +52 5 669 48 79Fax +52 5 543 60 [email protected]

URUGUAY

Uruguay SectionRafael AbalSondor S.A.Calle Rio Branco 1530C.P. UY-11100 MontevideoUruguayTel. +598 2 901 26 70,

+598 2 90253 88Fax +598 2 902 52 72E-mail [email protected]

VENEZUELA

Taller de Arte Sonoro,Caracas Section (Student)Carmen Bell-Smythe de LealFaculty AdvisorAES Student SectionTaller de Arte SonoroAve. Rio de Janeiro Qta. Tres PinosChuao, VE-1061 CaracasVenezuelaTel. +58 14 9292552Tel./Fax +58 2 9937296E-mail [email protected]

Venezuela SectionElmar LealAve. Rio de JaneiroQta. Tres PinosChuao, VE-1061 CaracasVenezuelaTel. +58 14 9292552Tel./Fax +58 2 9937296E-mail [email protected]

INTERNATIONAL REGION

Vice President:Neville Thiele10 Wycombe St.Epping, NSW AU-2121,AustraliaTel. +61 2 9876 2407Fax +61 2 9876 2749E-mail [email protected]

AUSTRALIA

Adelaide SectionDavid MurphyKrix Loudspeakers14 Chapman Rd.Hackham AU-5163South AustraliaTel. +618 8 8384 3433Fax +618 8 8384 3419E-mail [email protected]

Brisbane SectionDavid RingroseAES Brisbane SectionP.O. Box 642Roma St. Post OfficeBrisbane, Qld. AU-4003, AustraliaOffice Tel. +61 7 3364 6510E-mail [email protected]

Melbourne SectionGraham J. HaynesP.O. Box 5266

Wantirna South, VictoriaAU-3152, AustraliaTel. +61 3 9887 3765Fax +61 3 9887 [email protected]

Sydney SectionHoward JonesAES Sydney SectionP.O. Box 766Crows Nest, NSW AU-2065AustraliaTel. +61 2 9417 3200Fax +61 2 9417 3714E-mail [email protected]

HONG KONG

Hong Kong SectionHenry Ma Chi FaiHKAPA, School of Film and

Television1 Gloucester Rd. Wanchai, Hong KongTel. +852 2584 8824Fax +852 2588 [email protected]

INDIA

India SectionAvinash OakWestern Outdoor Media

Technologies Ltd.16, Mumbai Samachar MargMumbai 400023, IndiaTel. +91 22 204 6181Fax +91 22 660 8144E-mail [email protected]

JAPAN

Japan SectionKatsuya (Vic) Goh2-15-4 Tenjin-cho, Fujisawa-shiKanagawa-ken 252-0814, JapanTel./Fax +81 466 81 0681E-mail [email protected]

KOREA

Korea SectionSeong-Hoon KangTaejeon Health Science CollegeDept. of Broadcasting

Technology77-3 Gayang-dong Dong-guTaejeon, Korea Tel. +82 42 630 5990Fax +82 42 628 1423E-mail [email protected]

MALAYSIA

Malaysia SectionC. K. Ng King Musical Industries

Sdn BhdLot 5, Jalan 13/2MY-46200 Kuala LumpurMalaysiaTel. +603 7956 1668Fax +603 7955 4926E-mail [email protected]

PHILIPPINES

Philippines SectionDario (Dar) J. Quintos125 Regalia Park TowerP. Tuazon Blvd., CubaoQuezon City, PhilippinesTel./Fax +63 2 4211790, +63 2

4211784E-mail [email protected]

SINGAPORE

Singapore SectionCedric M. M. TioApt. Block 237Bishan Street 22, # 02-174Singapore 570237Republic of SingaporeTel. +65 6887 4382Fax +65 6887 7481E-mail [email protected]

Chair:Dell HarrisHampton University Section(AES)63 Litchfield CloseHampton, VA 23669Tel +1 757 265 1033E-mail [email protected]

Vice Chair:Scott CannonStanford University Section (AES)P.O. Box 15259Stanford, CA 94309Tel. +1 650 346 4556Fax +1 650 723 8468E-mail [email protected]

Chair:Isabella Biedermann European Student SectionAuerhahnweg 13A-9020 Klagenfurt, AustriaTel. +43 664 452 57 22E-mail [email protected]

Vice Chair:Felix Dreher European Student SectionUniversity of Music andPerforming ArtsStreichergasse 3/1 AA-1030 Vienna, AustriaTel. +43 1 920 54 19E-mail [email protected]

EUROPE/INTERNATIONALREGIONS

NORTH/SOUTH AMERICA REGIONS

STUDENT DELEGATEASSEMBLY



AES CONVENTIONS AND CON

23rd International ConferenceCopenhagen, Denmark“Signal Processing in AudioRecording and Reproduction”Date: 2003 May 23–25Location: Marienlyst Hotel,Helsingør, Copenhagen,Denmark

The latest details on the following events are posted on the AES Website: http://www.aes.org

Convention chair:Peter A. SwarteP.A.S. Electro-AcousticsGraaf Adolfstraat 855616 BV EindhovenThe NetherlandsTelephone: +31 40 255 0889Email: [email protected]

Papers chair: Ronald M. AartsVice Chair: Erik Larsen

DSP-Acoustics & SoundReproductionPhilips Research Labs, WY81Prof. Hostlaan 45656 AA Eindhoven, TheNetherlandsTelephone: +31 40 274 3149Fax: +31 40 274 3230Email: [email protected]

Exhibit information:

114th ConventionAmsterdam, The NetherlandsDate: 2003 March 22–25Location: RAI Conference and Exhibition Centre,Amsterdam, The Netherlands

Conference chair:Per RubakAalborg UniversityFredrik Bajers Vej 7 A3-216DK-9220 Aalborg ØDenmarkTelephone: +45 9635 8682Email: [email protected]

Papers cochair: Jan Abildgaard PedersenBang & Olufsen A/SPeter Bangs Vej 15P.O. Box 40,DK-7600 StruerPhone: +45 9684 1122Email: [email protected]

Papers cochair: Lars Gottfried JohansenAalborg University

Papers chair: Geoff MartinEmail: [email protected]

Conference chair:Theresa LeonardThe Banff CentreBanff, CanadaEmail: [email protected]

Conference vice chair:John SorensenThe Banff CentreBanff, CanadaEmail: [email protected]

Fax: +81 3 5494 3219Email: [email protected]

Convention vice chair: Hiroaki SuzukiVictor Company of Japan (JVC)Telephone: +81 45 450 1779Email: [email protected]

Papers chair: Shinji KoyanoPioneer Corporation

Telephone: +81 49 279 2627Fax: +81 49 279 1513Email:[email protected]

Workshops chair: Toru KamekawaTokyo National University of Fine Art& MusicTelephone: +81 3 297 73 8663Fax: +81 297 73 8670

11th Regional ConventionTokyo, JapanDate: 2003 July 7–9Location: Science Museum,Chiyoda, Tokyo, JapanConvention chair:Kimio HamasakiNHK Science & Technical ResearchLaboratoriesTelephone: +81 3 5494 3208

24th International ConferenceBanff, Canada“Multichannel Audio:The New Reality”Date: 2003 June 26–28Location: The Banff Centre,Banff, Alberta, Canada

Exhibit information:Thierry BergmansTelephone: +32 2 345 7971Fax: +32 2 345 3419Email: [email protected]

116th ConventionBerlin, GermanyDate: 2004 May 8–11Location: Messe BerlinBerlin, Germany

Papers chair:James D. JohnstonMicrosoft CorporationTelephone: + 1 425 703 6380Email: [email protected]

Convention chair:Zoe ThrallThe Hit Factory421 West 54th StreetNew York, NY 10019, USATelephone: + 1 212 664 1000Fax: + 1 212 307 6129Email: [email protected]

115th ConventionNew York, NY, USADate: 2003 October 10–13Location: Jacob K. JavitsConvention Center, NewYork, New York, USA

Banff2003

New York

2003

2003

Am

ster

dam

Berlin, Germany2004

2003

Thierry BergmansTelephone: +32 2 345 7971Fax: +32 2 345 3419Email: [email protected]

Call for papers: Vol. 50, No. 6,p. 535 (2002 June)

Convention preview: Vol. 51, No. 1/2,pp. 76–92 (2003 January/February)

Convention report: This issue,pp. 386–441 (2003 May)

FERENCESPresentationManuscripts submitted should betypewritten on one side of ISO size A4(210 x 297 mm) or 216-mm x 280-mm(8.5-inch x 11-inch) paper with 40-mm(1.5-inch) margins. All copies includingabstract, text, references, figure captions,and tables should be double-spaced.Pages should be numbered consecutively.Authors should submit an original plustwo copies of text and illustrations.ReviewManuscripts are reviewed anonymouslyby members of the review board. After thereviewers’ analysis and recommendationto the editors, the author is advised ofeither acceptance or rejection. On thebasis of the reviewers’ comments, theeditor may request that the author makecertain revisions which will allow thepaper to be accepted for publication.ContentTechnical acrticles should be informativeand well organized. They should citeoriginal work or review previous work,giving proper credit. Results of actualexperiments or research should beincluded. The Journal cannot acceptunsubstantiated or commercial statements.OrganizationAn informative and self-containedabstract of about 60 words must beprovided. The manuscript should developthe main point, beginning with anintroduction and ending with a summaryor conclusion. Illustrations must haveinformative captions and must be referredto in the text.

References should be cited numerically inbrackets in order of appearance in thetext. Footnotes should be avoided, whenpossible, by making parentheticalremarks in the text.

Mathematical symbols, abbreviations,acronyms, etc., which may not be familiarto readers must be spelled out or definedthe first time they are cited in the text.

Subheads are appropriate and should beinserted where necessary. Paragraphdivision numbers should be of the form 0(only for introduction), 1, 1.1, 1.1.1, 2, 2.1,2.1.1, etc.

References should be typed on amanuscript page at the end of the text inorder of appearance. References toperiodicals should include the authors’names, title of article, periodical title,volume, page numbers, year and monthof publication. Book references shouldcontain the names of the authors, title ofbook, edition (if other than first), nameand location of publisher, publication year,and page numbers. References to AESconvention preprints should be replacedwith Journal publication citations if thepreprint has been published.IllustrationsFigure captions should be typed on aseparate sheet following the references.Captions should be concise. All figures

should be labeled with author’s name andfigure number.Photographs should be black and white prints without a halftone screen,preferably 200 mm x 250 mm (8 inch by10 inch).Line drawings (graphs or sketches) can beoriginal drawings on white paper, or high-quality photographic reproductions.The size of illustrations when printed in theJournal is usually 82 mm (3.25 inches)wide, although 170 mm (6.75 inches) widecan be used if required. Letters on originalillustrations (before reduction) must be largeenough so that the smallest letters are atleast 1.5 mm (1/16 inch) high when theillustrations are reduced to one of the abovewidths. If possible, letters on all originalillustrations should be the same size.Units and SymbolsMetric units according to the System ofInternational Units (SI) should be used.For more details, see G. F. Montgomery,“Metric Review,” JAES, Vol. 32, No. 11,pp. 890–893 (1984 Nov.) and J. G.McKnight, “Quantities, Units, LetterSymbols, and Abbreviations,” JAES, Vol.24, No. 1, pp. 40, 42, 44 (1976 Jan./Feb.).Following are some frequently used SIunits and their symbols, some non-SI unitsthat may be used with SI units (), andsome non-SI units that are deprecated ( ).

Unit Name Unit Symbolampere Abit or bits spell outbytes spell outdecibel dBdegree (plane angle) () °farad Fgauss ( ) Gsgram ghenry Hhertz Hzhour () hinch ( ) injoule Jkelvin Kkilohertz kHzkilohm kΩliter () l, Lmegahertz MHzmeter mmicrofarad µFmicrometer µmmicrosecond µsmilliampere mAmillihenry mHmillimeter mmmillivolt mVminute (time) () minminute (plane angle) () ’nanosecond nsoersted ( ) Oeohm Ωpascal Papicofarad pFsecond (time) ssecond (plane angle) () ”siemens Stesla Tvolt Vwatt Wweber Wb

INFORMATION FOR AUTHORS

Niels Jernes Vej 14, 4DK-9220 Aalborg ØPhone: +45 9635 9828Email: [email protected]

Call for papers: Vol. 50, No. 9,p. 737 (2002 September)

Conference preview: Vol. 51, No. 3,pp. 170–179 (2003 March)

Call for contributions: Vol. 50, No. 10,pp. 851–852 (2002 October)

Conference preview: Vol. 51, No. 4,pp. 258–270 (2003 April)

Exhibit information:Chris PlunkettTelephone: +1 212 661 8528Fax: +1 212 682 0477Email: [email protected]

Call for papers: Vol. 51, No. 1/2,pp. 112 (2003 January/February)

Email: [email protected]

Exhibit chair: Tadahiko NakaokiPioneer Business Systems DivisionTelephone: +81 3 3763 9445Fax : +81 3 3763 3138Email: [email protected]

Section contact: Vic GohEmail: [email protected]

Call for papers: Vol. 50, No. 12,pp. 1124 (2002 December)

sustainingmemberorganizations AESAES

VO

LU

ME

51,NO

.5JO

UR

NA

L O

F T

HE

AU

DIO

EN

GIN

EE

RIN

G S

OC

IET

Y2003 M

AY

JOURNAL OF THE AUDIO ENGINEERING SOCIETYAUDIO / ACOUSTICS / APPLICATIONSVolume 51 Number 5 2003 May

The Audio Engineering Society recognizes with gratitude the financialsupport given by its sustaining members, which enables the work ofthe Society to be extended. Addresses and brief descriptions of thebusiness activities of the sustaining members appear in the Octoberissue of the Journal.

The Society invites applications for sustaining membership. Informa-tion may be obtained from the Chair, Sustaining Memberships Committee, Audio Engineering Society, 60 East 42nd St., Room2520, New York, New York 10165-2520, USA, tel: 212-661-8528.Fax: 212-682-0477.

ACO Pacific, Inc.Air Studios Ltd.AKG Acoustics GmbHAKM Semiconductor, Inc.Amber Technology LimitedAMS Neve plcATC Loudspeaker Technology Ltd.Audio LimitedAudiomatica S.r.l.Audio Media/IMAS Publishing Ltd.Audio PartnershipAudio Precision, Inc.AudioScience, Inc.Audio-Technica U.S., Inc.AudioTrack CorporationAutograph Sound Recording Ltd.B & W Loudspeakers LimitedBMP RecordingBritish Broadcasting CorporationBSS Audio Cadac Electronics PLCCalrec AudioCanford Audio plcCEDAR Audio Ltd.Celestion International LimitedCerwin-Vega, IncorporatedClearOne Communications Corp.Community Professional Loudspeakers, Inc.Crystal Audio Products/Cirrus Logic Inc.D.A.S. Audio, S.A.D.A.T. Ltd.dCS Ltd.Deltron Emcon LimitedDigidesignDigigramDigital Audio Disc CorporationDolby Laboratories, Inc.DRA LaboratoriesDTS, Inc.DYNACORD, EVI Audio GmbHEastern Acoustic Works, Inc.Eminence Speaker LLC

Event Electronics, LLCFerrotec (USA) CorporationFocusrite Audio Engineering Ltd.Fostex America, a division of Foster Electric

U.S.A., Inc.Fraunhofer IIS-AFreeSystems Private LimitedFTG Sandar TeleCast ASHarman BeckerHHB Communications Ltd.Innova SONInnovative Electronic Designs (IED), Inc.International Federation of the Phonographic

IndustryJBL ProfessionalJensen Transformers Inc.Kawamura Electrical LaboratoryKEF Audio (UK) LimitedKenwood U.S.A. CorporationKlark Teknik Group (UK) PlcKlipsch L.L.C.Laboratories for InformationL-Acoustics USLeitch Technology CorporationLindos ElectronicsMagnetic Reference Laboratory (MRL) Inc.Martin Audio Ltd.Meridian Audio LimitedMetropolis GroupMiddle Atlantic Products Inc.Mosses & MitchellM2 Gauss Corp.Music Plaza Pte. Ltd.Georg Neumann GmbH Neutrik AGNVisionNXT (New Transducers Ltd.)1 LimitedOntario Institute of Audio Recording

TechnologyOutline sncPacific Audio-VisualPRIMEDIA Business Magazines & Media Inc.Prism Sound

Pro-Bel LimitedPro-Sound NewsPsychotechnology, Inc.Radio Free AsiaRane CorporationRecording ConnectionRocket NetworkRoyal National Institute for the BlindRTI Tech Pte. Ltd.Rycote Microphone Windshields Ltd.SADiESanctuary Studios Ltd.Sekaku Electron Ind. Co., Ltd.Sennheiser Electronic CorporationShure Inc.Snell & Wilcox Ltd.Solid State Logic, Ltd.Sony Broadcast & Professional EuropeSound Devices LLCSound On Sound Ltd.Soundcraft Electronics Ltd.Sowter Audio TransformersSRS Labs, Inc.Stage AccompanySterling Sound, Inc.Studer North America Inc.Studer Professional Audio AGTannoy LimitedTASCAMTHAT CorporationTOA Electronics, Inc.TommexTouchtunes Music Corp.TurbosoundUnited Entertainment Media, Inc.Uniton AGUniversity of DerbyUniversity of SalfordUniversity of Surrey, Dept. of Sound

RecordingVidiPaxWenger CorporationJ. M. Woodgate and AssociatesYamaha Research and Development

In this issue…

Inconsistent Loudspeaker Cone Displacement

Low-Frequency SpatialEqualization

Vector Sound Intensity Probe

Digital Audio BroadcastingEvaluation

Standards:Radio Traffic Data in BroadcastWave Files

Technical Council:Technical Committee Reports

Features…

114th Convention Report,Amsterdam

Documents

Journal AES 2003 May Vol 51 Num 5