T. Fischer, K. De Biswas, J.J. Ham, R. Naka, W.X. Huang, Beyond Codes and Pixels: Proceedings of the 17th International Conference on Computer-Aided Architectural Design Research in Asia, 629–636. ©2012, Association for Computer-Aided Architectural Design Research in Asia (CAADRIA), Hong Kong

Moving in FilMic SpaceS

Relating camera movements to spatial archetypes in architectural animations

MAHA ZeINI Al-SAATI, DAvID BoTTA AND RoBeRT WooDBuRySchool of Interactive Arts and Technology, Simon Fraser University, Surrey, Canadamalsaati@sfu.ca

abstract. Architects sometimes use architectural animations to explain their designs. To probe the practice of architectural film/animation, this paper assembles filmic spaces as a conceptual tool that connects spatial archetypes with camera movements.

Keywords. Architecture; animation; representation; film; camera.

1. introduction

Architectural animations are often shaped by CAD software, which depends on the creation of paths, offering various degrees of freedom, little guidance, and few editing options. This leads to conventions of depicting movement in space through devices such as fly-throughs, combined with quickly changing angles, rotations and swirls that tour the building, speeding up at times and slowing at others. Techniques from cinema can yield richer spatial experience.

This research examines spaces in architectural animations in terms of both form and camera movement, encompassed by the term “filmic spaces”. In order to achieve this goal, we present a study in two stages. In the first stage, we review relevant literature on architectural animation that ties dif-ferent types of “spatial forms” to various “camera movements”. We develop a “filmic space” concept based on Cullen’s serial vision (Cullen 1995), who views spatial experience as a transformation of form revealed through move-ment in space. We then discuss the outcome: filmic spaces.

630 M. Z. Al-SAATI, D. BoTTA AND R. WooDBuRy

Instead of regarding space as a static form, we view space as one that changes from one state to another through camera movements. To understand the relationship between spatial forms and camera movements, the second stage comprises a close-reading analysis of films, which looks for filmic spaces, and understands how spaces and camera movements are related.

2. literature review

The design of an architectural animation is strongly tied to the intended goal. These can include non-commercial goals, such as exploring design alterna-tives and decisions, or documenting a historic building. Commercial goals can address data-communication, educating and informing about the building form, design features, and construction methods in order to promote design to clients and developers. Animations are also used to make an impression and convey an experience to occupants who are being sold a dream or lifestyle (Rafi 1998, Serrato-Combe 2004, Koutamanis 2005, Sanguinetti 2005, Kwee et al. 2006, Swathika et al. 2006, Alvarado 2008).

Architectural animation design is also influenced by the audience. For instance, if the audience were developers of large-scale projects, the animation will tend to have bird-eye views and fly-throughs to demonstrate the overall design. Meanwhile, when the animation addresses potential occupants, the aim is often to sell them an experience or a life style, and we would find more low-angles and immersed points of view. But if we were to move beyond these general impressions on architectural animations, how does previous research address the issue of architectural animation design?

Many previous studies on architectural animations relate camera move-ment to space types (Goldman and Zdepski 1987, Goldman and Zdepski 1988, Kwon and Nagakura 2004, Alvarado et al. 2005, Chatzitsakyris and Nagakura 2005, Calderon et al. 2006, Alvarado 2008). Since spatial form can become overwhelmingly complex, they start by identifying simple forms like: linear, layered, bent, broad, and tall spaces (Kwon and Nagakura 2004); arena, and maze (Nitsche 2008); path, edge, district, node, and landmark (Lynch 1960); composition, rhythm, scale, proportion and verticality (Calderon et al. 2006). Following the tradition of previous literature, we turn to film to perform close-reading analysis. This is to understand the nature of filmic spaces as a combi-nation of camera movement and spatial form.

3. Close reading films

Close-reading required the careful study, analysis and interpretation of the particulars in a film. Through disciplined immersion, the perceptive senses are

631MovINg IN FIlMIC SpACeS

attuned to observe and extract repeated patterns (Kain 1998, Mantex 2011). We selected film clips with strong architectural components and spatial fea-tures, and those featuring spatial exploration. We searched for spatial arche-types and noted their connections to certain camera movements. We stopped encoding when we have reached saturation, a state when no more can be learnt from additional viewing. These films include: a scene from Marie Antoinette (Coppola 2006) in which Antoinette explores Versailles for the first time, a chase scene in Brazil (Gilliam 1985), in which Sam runs through a series of narrow arcade spaces and ends in a funeral home; a scene from The Glass house (Sackheim 2001), in which two teenagers explore the extraordinary home of their caregivers; a scene from Running Scared (Kramer 2006), in which an abducted child searches the apartment of his kidnappers for an exit; A scene from Duplex (Devito 2003), where a young couple explore a duplex house before deciding to buy it; the Spectra Town scene from Big fish (Burton 2003); several scenes from The Fall (Singh 2006), which show monumental buildings; scenes from Gladiator (Scott 2000), including when the city of Rome is shown through a bird-eye view, and when the gladiators are awed by the monumental scale of the coliseum.

The outcome of our analysis combines camera movements, spatial types and characters into what we call filmic spaces. We argue that such triples provide elements with which to conceive and develop architectural animations.

4. Filmic spaces

Filmic spaces associate spatial forms with specific camera movements. We start with simplicity before moving on to complexity, and thus identify basic camera movements corresponding to three primary shapes: (1) point, com-prising a stationary camera position in space, observing occurring actions, but not affected by them; (2) line, where a camera travels along a path from one point to another. lines comprise dolly forward or backward, pedestal up or down, and tracking-sideways left or right; (3) Circle, where a camera rotates around a central point. Primitive motions for circles include a pan (left/ right outward-rotation), an arc (inward-rotation), or a tilt (upwards or downward-rotation, Figure 1). We then proceed to the application of these basic camera movements to convey spatial archetypes. These basic camera movements can either reinforce explicit spaces, or create implicit spaces: ones that are not physically defined. In the following sections, we demonstrate several filmic spaces, with references to key films in which they occur. We give each filmic space a title in the form of CAMeRA:SpACe:CHARACTeR, describing respectively the camera movement, space type depicted and role of character. When a character is not involved, the last element of the triple is null.

632 M. Z. Al-SAATI, D. BoTTA AND R. WooDBuRy

Figure 1. Basic camera movements include point, line and circle.

4.1. poINT:AXIS:CouNTeRpoINT

A stationary camera observes characters moving against a spatial back-ground. The characters amplify spatial qualities. In Marie Antoinette, poINT:AXIS:CouNTeRpoINT is used to convey a major axis at versailles and to convey the vastness of space in with respect to human scale (Figure 2).

Figure 2. A stationary camera observes characters moving across the screen, revealing space in this scene from Marie Antoinette (Coppola, 2006).

4.2. LINE:PATH:FOCUS

One of the horizontal camera movements is the dolly, which penetrates the depth of space. In this pattern, a dolly movement aligns with the passage of

633MovINg IN FIlMIC SpACeS

a character through the space. The movement may observe the character as a third person or as a first person. The character serves to reveal the path and to give meaning to its traversal. Examples of such paths include a road defined by floorings and ceiling (Figure 3a), or an arcade defined by vertical elements on the edges of a path, creating punctuation between solid and void (Figure 3b). A path can also transform from one state to another. For instance, in Figure 3c, a clearance turns from narrow to wide, dark to bright, and inside to outside, giving a feeling of relief. The tunnel gives the reverse effect. Another horizon-tal camera movement is Figure 3d, tracking-sideways, which emphasises the width of space. It tracks subjects through silhouetted foreground elements to establish the context in which they are situated, or reveal the perspective depth of space. In Figure 3e, a combination of tracking-sideways and dolly move-ments help create a reveal effect, by having a visual surprise concealed behind the gate, then revealing it through camera movement.

Figure 3. The linear camera movements, such as dolly and tracking sideways, emphasise spaces such as path, arcade, clearance, tunnel, grid and maze.

Figure 3 reveals these linear camera movements. The dolly movement is used for spaces like a road (Figure 3a), or an arcade (Figure 3b) created by people standing on the edges of the path in a scene from Marie Antoinette (Coppola

634 M. Z. Al-SAATI, D. BoTTA AND R. WooDBuRy

2006). In Figure 3c, a tunnel and clearance space is created in a scene from Brazil (Gilliam 1985). In Figure 3d, the tracking-sideways movement reveals the market place where the characters walk in a scene from A Series of Unfor-tunate Events (Silberling 2004), and also reveals the depth of the grid space in a scene from Manufactured Landscapes (Baichwal 2006). In figure 3e a combination of dolly and tracking-sideways camera movement follows the protagonist to reveal a visual surprise behind the gate in a scene from Tideland (Gilliam 2005), creating a maze-like space.

4.4. lINe:veRTICAl SpACe:Null

Tall spaces profoundly affect architectural experience. The standard horizon-tal orientation of moving images is ill-suited to convey such experience. Film-makers use camera movement to convey verticality. Such movements can also create implicit space that is experientially more than physically defined. For example, the boundary between a village and its surround may define a gate. Such implicit spaces may mark the threshold between two territories, or serve simply to introduce or exit a scene. Characters in this pattern play a lesser role or may even be absent – the focus is on conveying spatial effect.

Figure 4 reveals the vertical camera movement can emphasise the length of a shaft space (c), through (a) a tilt, as in Marie Antoinette (Coppola 2006), or (b) pedestal movement, as in Pan’s Labyrinth (Del Toro 2006). In (d), a scene from A series of Unfortunate Events (Silberling 2004) implies a gate space, as the camera pedestals up when a character exits the space.

Figure 4. The vertical camera movements emphasise the height of space.

4.3. CIRCle:RoTuNDA:Null

A rotunda space can be evoked by a rounded camera movement, such as an arc in or a pan movement. The court space evokes various rounded camera movement like arc or pan.

635MovINg IN FIlMIC SpACeS

Figure 5 reveals (a) The rounded movement enforces the feel of the rotunda, and includes movements such as (b) arc, observed in The Fall (Singh 2006) or (c) pan, in Marie Antoinette (Coppola 2006).

Figure 5. Rounded camera movements create a rotunda space.

6. Discussion and future work

This is a descriptive work that acts an introduction to the larger scope of our research. In this paper, we proposed the concept of filmic spaces in architec-tural animations. The proposed filmic space model combines spatial form with camera movement. Due to the limitation of this short paper, we have only explored two aspects, though they are not exhaustive. our larger framework explores further cinematic techniques like narrative, editing, composition, lighting, sound, and camera angles.

note

We used hand drawings for the figures in this paper for two reasons. First, copyright clearance for film segments can be difficult, while the copyright to hand-drawings lies with their creator. using hand drawings means that we were unconstrained in sampling films and thus avoided any bias due to arbi-trary difficulties in sourcing material. Second, hand drawing allowed us to emphasise drawing features to relate to points being made.

AcknowledgementsWe thank Dammam University of Saudi Arabia for financial support of the first author’s research. We also thank the architects who participated in this study. This work was partially supported through the Canadian Natural Science and engineering Research Council Discovery grants and Collaborative Research and Development programs; and the Networks of Centers of excellence program through the graphics, Animation and New Media Network.

ReferencesAlvarado, R. g. et al.: 2005, Filmic Development of Architectural Animations.Alvarado, R. G.: 2008, Analysis of filmmaking techniques. METU Journal of the Faculty of

Architecture, 25(2), 133–149.Baichwal, J.: 2006, Manufactured Landscapes, video recording, Zeitgeist video.

636 M. Z. Al-SAATI, D. BoTTA AND R. WooDBuRy

Burton, T.: 2003, Big Fish, video recording, DreamWorks and Sony pictures Releasing.Calderon, C., Nyman, K. and Worley, N.: 2006, Architectural cinematographer (AC): creating

architectural experiences in 3d real-time environments, International Journal of Architec-tural Computing, 4(4), 71–90.

Chatzitsakyris, P. and Nagakura, T.: 2005, The Man with the Movie Camera: An Event-Driven Approach to Architectural Design, Master of Architecture Thesis, Massachusetts Institute of Technology, Cambridge.

Coppola, S.: 2006, Marie Antoinette, video recording, Sony pictures Releasing.Cullen, g.: 1995, The Concise Townscape, Reinhold, New york.Del Toro, g.: 2006, Pan’s Labyrinth, video recording, Warner Bros., picturehouse and New

line Cinema.Devito, D.: 2003, Duplex, video recording, Miramax Films.Gilliam, T.: 1985, Brazil, video recording, universal pictures and voyager Company.gilliam, T.: 2005, Tideland, video recording, Recorded picture Company.Goldman, G. and Zdepski, S.: 1987, Form, color and movement, in Bancroft, p. J. (ed.), Proc.

ACADIA 1987 Conference, Raleigh, 39–50.Goldman, G. and Zdepski, S.: 1988, Abstraction and representation: computer graphics and

architectural design, in Bancroft, p. J. (ed.), Proc. ACADIA 1988 Conference, Ann Arbor, 205–215.

Kain, P.: 1998, How to Do a Close-reading, Writing Center, Harvard university, Cambridge.Koutamanis, A.: 2005, The continuity style in architectural dynamic visualization, in Angulo, A.

and Vasquez de Velasco, G. (eds.), Proc. 9th SIGraDI Conference, Lima, Vol. 1, 468–472.Kramer, W.: 2006, Running Scared, video recording, New line Cinema.Kwee, v., Bruton, D. and Radford, A.: 2006, visual expressiveness in educative architectural

animations, in lee, y. T. et al. (eds.), Proc. 4th GRAPHITE Conference, Kuala lumpur, 191–194.

Kwon, K.E. and Nagakura, T.: 2004, Filmic Architecture: On Motion Perspective in Archi-tectural Synthesis, Master of Architecture Thesis, Massachusetts Institute of Technology, Cambridge.

lynch, K.: 1960, The Image of the City, MIT press, Cambridge.Mantex: 2011, What is close reading? – a brief guide to advanced reading skills. Mantex. Avail-

able at: http://www.mantex.co.uk/2009/09/14/what-is-close-reading-guidance-notes.Nitsche, M.: 2008, Video Game Spaces: Image, Play, and Structure in 3D Game Worlds, MIT

press, Cambridge.Rafi, A.: 1998, Computer Animation for Architectural Visualisation. PhD Thesis. University of

Strathclyde, glasgow. Sackheim, D.: 2001, The Glass House, Sony pictures Releasing.Sanguinetti, p.: 2005, The integration of cinema in the digital design studio, in Angulo, A. and

Vasquez de Velasco, G. (eds.), Proc. 9th SIGraDI Conference, Lima, Vol. 1, 286–290.Scott, R.: 2000, Gladiator, video recording, DreamWorks.Serrato-Combe, A.: 2004, Something’s gotta give: architectural animations, SIGraDi 2004,

porte Alegre, Brasil.Silberling, B.: 2004, Lemony Snicket’s A Series of Unfortunate Events, video recording, para-

mount pictures and DreamWorks.Singh, T.: 2006, The Fall, video recording, Roadside Attractions.Swathika, A., yoon, S.-y. and Saleh uddin, M.: 2006, Architectural animation and cine-

matic interpretation, in Kaga, A. and Naka, R. (eds.), Proc. 11th CAADRIA Conference, Kumamoto, 197–202.