Vanishing Point: Spatial Composition and the Virtual Camera
[this paper was presented at the University of NSW school of media theatre and film in 2007. The presentation slides and an audio podcast of the lecture are available for download on the Podcasts Page of this site. The paper was subsequently published in the Sage Journal 'Animation' and a high-quality PDF version of the journal article is availible HERE]
Introduction: Staging the Camera
“It is quite feasible to produce a film without actors, but a film without a camera is a sheer impossibility.” Vagn Börge, 1962.
Whilst the above statement holds an invariable truth linked to the very nature of what cinema is we are also confronted by its confidence to re-consider not just how we define but also how we perceive the camera as apparatus, the camera as vehicle and the camera as entity.
If we root the notion of a camera in the apparatus of image capture, and moreover think first and foremost of the camera as physical tool, then the statement is not only rendered incorrect in a modern context but arguably has always been a fallacy. If we consider the foundation of cinema in the simple illusion of movement, the persistence of vision – the Zoetrope, flip-book, shadow puppet and cartoon animation – then the cinema has a long history of being camera-less. However, considered more broadly - the camera as vehicle and entity, as positioning construct for perception and engagement - then we have a firm statement that links cinema inextricably with perspective, with point-of-view and, more inclusively, with perspective embodiment.
Hand in hand with the notion of the camera as perspective vehicle is a long established compositional framework based on staging action, movement and subject for the camera. Distinctly borrowing from a theatrical sensibility there is a central pillar of frame-based arrangement of visual elements in cinema whereby the position of subject is dictated by the position of the camera; or else the position of the camera dictates the arrangement of the subject(s). In either case the relationship between the camera and depicted visual content is one whereby the ‘scene’ is staged for the camera’s benefit and, by proxy, the viewer’s benefit.
But in the digital age of infinitely flexible production means, where live-action photography is increasingly seen as an option rather than a given, and the virtual camera moves fluidly in a 3D constructed and virtual space, we are presented with a very different compositional sensibility. The traditional mise-en-scene, defined simply and effectively as “whatever appears in the film frame” (Cook, 1985: p151), cannot holistically account for either the process of constructing and depicting a cinematic scene from a 3D virtual camera, or of the new spatial awareness on the part of the viewer that arises through the virtual camera’s incorporeal, non-physical nature. The lack of physicality and subsequently the ability to depict, engage and navigate cinematic space in a new and omnipotent manner, forces us to reconsider our perceptions of both what the camera is and what it does. Further it makes ever narrower and less functional the distinction between what is animation and what is not.
One suggestion that goes a long way toward providing a conceptual framework that draws upon older established theoretical thinking to conceptualise a current state of affairs, is the notion of digital cinema as a neo-baroque movement.
With borders constantly being rewritten, (neo)baroque vision provides models of perception that suggest worlds of infinity that lose the sense of a centre that is traditionally associated with classically ordered space. Rather, the centre is to be found in the position of the spectator, with representational centre changing depending on the spectators focus… It is the audience’s perception and active engagement with the image that orders the illusion… the result being the articulation of complex spatial conditions. (Ndalianis, 2000: p358)
Here we have not only a creation of space as primary compositional framework (above and beyond the cinematic frame) but the camera, as an agent of perspective, becomes a compositional ‘element’ rather than a compositional ‘tool’. Through the virtual camera, and constructed or composited 3D spaces, digital cinematic forms no longer stage for the camera but stage and compose the camera itself as a form of specific purpose scenic content. The constructed space becomes the macro-frame work, what I’ve termed the mise-en-space whereby camera ‘objects’ are composed into the space to serve as a viewer aware, spatial-frame, extending well beyond the momentary framed window.
The entity of the camera becomes not an apparatus of moving image capture but rather a spatially specific vanishing point of perception. The virtual camera is the spatial ‘moment’ of temporal consciousness composed into a scene. The physical camera has long been associated with the ‘eye’, as a simulation of the ‘eye’ in an alternate world, (either anthropomorphically or otherwise) and in doing so is rooted to a physical and tangible exploration and depiction of space that obeys rules of physics. This can be seen in almost any physical moving camera shot from the dolly-based “Cabiria movements” (Cousins, 2004: p30) to Steadi-cam and hand-held where the camera navigates and adheres to the space it occupies.
The virtual camera, by contrast, moves beyond this into new conceptualisations of space and the viewer’s connection to, or immersion in, it. Through intangibility beyond the depicted space the virtual camera becomes a simulation of ‘I’ rather than ‘eye’; a simulation of viewer derived presence in space rather than an anthropomorphically based viewing apparatus. The animated virtual camera simulation of ‘I’ is simultaneously both a party of the scenic composition and beyond it.
In understanding the impact of the virtual camera there are a range of frameworks that the concept can, and needs to, be tested against and conceptualised by. What follows is a positioning of the virtual camera within the contextual frameworks of a number of established cinematic concepts consistent with both animation and live-action sensibilities. The division between the two is increasingly narrow through, if nothing else, popular and mainstream coexistence and so it is an attempt to understand where the virtual camera re-mediates older notions of the camera, as both concept and apparatus, and where the virtual camera forges its own language and aesthetic pathways. However, in order to comprehend those more philosophical and ephemeral elements of the virtual camera it is important to position the technical and production methodologies of the virtual camera as construct drawn from the media form where the virtual camera is native and organic; gaming.
Gaming and Production Process
At a surface level the virtual camera presents, as its name implies, a non-physical device possessing all the traits and characteristics of a physical camera. It is used to achieve the same ends of capturing and presenting a ‘perspective’ in a virtual computer generated environment. Indeed the software-based emulation of camera optics and artificially imposed constraints of mechanics and physicality is still the prevailing aesthetic of the virtual camera in much motion picture production.
Software virtual cameras are invested with many of the mechanical properties of physical cameras to simulate a common visual language; depth of field, lens flare, focal length and aperture. Similarly, very few productions utilising a virtual camera (either in composited, hybrid live-action or pure 3D animation) venture very far from the accepted and well entrenched cinematic camera language of ‘dolly’, ‘track, ‘pan’, ‘wide-shot’, ‘mid-shot’, ‘close-up’ in accordance with physical space and objects. The number of cinema productions that have, or are, embracing the omnipresent, ethereal, non-physical nature and unique language constructs of cinematic motion embedded in the virtual camera, is likewise very small. However the key influences on cinematic form and, more importantly, cinematic expectations on the part of viewers are arguably being exerted externally from hybrid, new and technologically more sophisticated areas such as computer and video gaming rather than internally from cinema itself. In computer games the virtual camera is native and innate and so as the virtual camera finds its way into traditional cinema it is to gaming that we must look for the future of cinematic form.
Traditionally the language of narrative-based cinematic gaming relied heavily on traditional cinematic form with the ‘cut-scene’ - the shifting of perspective from pro-active first person to passive, montage-based, third person - as the staple for progressing storyline. Doom3 (2004) and Return to Castle Wolfenstein (2001) among numerous others, both before and since, employed this technique to chapterise game play and advance plot line.
However we now have a methodical evolution in gaming towards the establishment of a more independent cinematic language distinct from traditional cinema. We see an seminal example of this in Half Life 2 (2004) where, for the first time a First-Person Shooter genre game possessed no cut-scenes whatsoever and the player’s perspective was a single, unified first-person POV unbroken for the entire duration of the game; from prologue to epilogue. Here the game developers of Half Life 2 – Valve Software – have chosen to deliberately dispense with the cinematic tenants of montage, of individual and discreet shots in sequence, which have been traditionally dominant in gaming. At the other end of the spectrum is The Sims 2 (2004) which utilises a virtual camera able to be as intimate as a close-up or abstracted as a birds-eye view, defying physicality and spatiality in its movements and positioning between the two extremes.
These two examples illustrate the extent to which the virtual camera can function as a depicter of spatiality. Indeed many games, such as Max Payne (2001) hybridise between these two extremes with both restriction in spatial movement and, at the same time, a user controlled camera that defies physical elements and presents depiction based not on perspective or immersion but on best vantage point to the view the action. In this example, common to a great many varieties of gaming, the thrill of the game is as much in the ‘seeing’ and the ‘observing’ of feats of action and scenic vista as it is in controlling them. Following this line of thought it can also be asserted that the power and influence of the virtual camera (in regard to elements of depiction and omnipotent perspective) derives as much, if not more, from the construction of the virtual environments in which the camera will reside than of the traits, movements and composition of the virtual camera itself. This returns us to the concept of an evolution in cinematic-form from a staging for the camera to a staging of the camera; from the composition of shot, frame and mise-en-scene to a more holistic, more flexible, composition of space itself. A composition that relies not on the framed perspective to construct meaning but on a composed perspective within space from which meaning is explored.
On a practical and technical level this is exactly the process of constructing software-based 3D graphics environments (for either games or animated films). A live-action film goes into production with a clear story-board of shot and sequence mapped out in specific detail relative only to the framed perspective; the predictability of viewer positioning. As a result, the construction of a set, of an environment for those framed perspectives, is a process of building in the backgrounds of the pre-defined and prescribed frames. This is distinctly a process of staging for the camera whereby the largely pre-ordained position of the camera is the axis about which all else revolves and is concerned.
The design of a 3D environment, by contrast, is built on the premise of an un-predictable, infinitely flexible and variable camera that often has no limitations of movement and direction. Here it is the environment, the world, the space that is ‘composed’ by the creator. The camera, more viably referred to as a simple position of perspective, is then immersed into the scene after the major act of ‘composition’ has already taken place. A staging of the camera rather than a staging for the camera.
The influence on more traditional, non-interactive, filmmaking from both game-based technology and game-based aesthetics is twofold. Firstly, under the weight of a massively popular culture of computer gaming, viewer expectations of what is possible, of how a camera can and even should move, are fundamentally altered. Traditionally a camera flying through a wall to the other side might be considered to have broken the suspension of disbelief undermining the truthful, credible believability of the scene. This moment of fantasy defying the physical could be said to remove the viewer from an intimate engagement with the screen events as ‘real’. However, now and into the future, it may be seen as native and unremarkable for a camera to move in such a way, a natural mode of cinematic representation driven by a popular, game-driven, aesthetic of expectation.
The second is that the production processes of cinema (which have arguably changed very little over the past hundred years) undergo a substantial alteration. Live-action cinema production process and workflow is built substantially on the inflexibility of the cinema form. Traditionally, once a live-action image is captured, the amount of alteration that can take place on an image is rudimentary at best – colour grading, merging with other elements (titles, other images or image parts), blending and fading. The camera cannot be relocated or re-positioned, the time of day cant be changed, the sun or clouds cant be moved, rain cant be taken away or added in.
However in the digital production environment of computer generated 3D spaces all elements remain largely flexible at any time in the production process; indeed the very notion of ‘Production’ and ‘Post-Production’ stages become mostly irrelevant for wither animated films or live-action ones. A scene can be ‘re-shot’ at any time, the identical environmental elements maintained or changed , the properties of a virtual camera lenses altered and modified. The possibilities of this technology driven production model go beyond increased flexibility and adaptability of the form, rather they introduce a new conceptualisation of what ‘composition’ is in the context of a legacy of photography. Where once the photographic image from the photographic camera was the inextricable base-element of cinema, and thus the lens and the frame where the unshakable mode of composition, the move to a cinema that exists without a camera object, without a lens, without a staging for the camera imperative, forces us to reconsider the stringency of the frame – of the mise-en-scene itself.
Whilst non-game cinematic form, produced with these spatially-based 3D computer generated technologies, doesn’t possess the unpredictable camera of gaming, it none the less uses the same processes and construction methodologies to design, build and compose holistic spatial entities. Subsequently there is a shared method of primary composition between gaming, animation and cinematic form through a shared technological process. Unavoidably this must lead to a shared aesthetic based as much on viewer expectations as on a shared production line process of making it.
The Diegesis of the Camera
Diegesis is a simple enough concept; a cohesive narrative space, a coherent construction with its own internal logic. An object or event’s diegetic state is, broadly speaking, a reference to its perception of belonging (or not) to a consistent and accepted framework – generally a narrative. A diegetic object is not at odds with the environment it occupies; a diegetic narrative element has a natural status within a ’story’.
In the context of cinema there are a number of largely independent and established articulations of diegesis but the most commonly expressed, beyond diegetic adherence to narrative itself, is that related to the specifics of sound within the mise-en-scene. A diegetic sound therefore is simply one where the source of the sound, in both space and time, is tangible and (in most cases) visible within the mise-en-scene; an actor’s voice as they speak, foley sounds directly connected with visible actions. Conversely a non-diegetic sound is one where the sound’s source and presence in time and space is outside of the mise-en-scene, beyond the spatiality of the scene. Key examples in this case being a film’s musical score or a voice-over narration.
In the shadow of the virtual camera, possessing a fundamentally different engagement with and depiction of space, there is opportunity and impetus to reconsider the parameters of the camera itself in the context of diegesis. The notion of a diegesis of the camera is not unknown but its common reference is narrow and problematically specific to the depiction of a camera as prop within a scene and the POV of such a camera. In Boogie Nights (Anderson, 1997) for example, during Dirk Diggler’s first porn movie scene, a 16mm film camera is distinctly part of the scene’s diegesis – both as a physical prop commensurate with the locale and environment and also as a tangible POV for the viewer who is taken visually into the camera lens from which to view the scene as the perceived end-user will see it. This is effectively a repositioning of the viewer from seeing the scene play out from an abstracted (non-diegetic) camera entity to that of being repositioned in the perspective of a diegetic object with camera properties; two fundamentally different camera entities. However this notion of a camera-apparatus diegesis, I would argue, is fundamentally different to a more sophisticated notion of camera-positioning diegesis which examines the notion of the camera’s role in depicting a scene. In other words allowing us to pose questions whether the camera spatially and physically engages with a scene or transcends the spatiality of the scene.
If we define the positioning-diegesis of the camera using the same compositional methodology as for sound diegesis (that a diegetic element is identifiable as a part of the mise en scene) then a diegetic camera is one that is identifiable by the viewer as physically belonging to and occupying the scene. This is not to say that a diegetic camera is necessarily one that characters and objects in the mise-en-scene are aware of; the diegetic camera can remain ‘invisible’, but it is none the less adherent to the accepted spatiality and physicality of the scene.
There are many examples of such a diegetic camera that constructs for the viewer a perception mechanism obeying the spatiality as an occupant. One such example is the much quoted steadicam shot from Martin Scorsese’s Goodfellas (1990) where the camera follows two characters through a long winding walk via the back door and kitchen of a nightclub to their table in the main room. Along the way not only does the camera navigate corridors and rooms but also is forced to open doors and stop its movement to allow other objects and characters to pass by before it can continue its journey.
In this example the camera, whilst invisible to the occupants of the mise-en-scene, is none the less diegetically occupying, negotiating and responding to the composed cinematic space. That said, the diegetic camera is not by necessity reliant on technology developments such as the steadi-cam and a diegetic modality of the camera can be just as tangible with tripod mounted or fixed cameras that occupy specific and tangible diegetic space. Bazin, in discussing the work of Jean Renoir, has commented on the notion of the wandering camera invoking this concept of invisible diegetic presence.
the camera acts like an invisible guest wandering about the salon and the corridors with a certain curiosity, but without any more advantage than its invisibility… the camera even gets trapped in a corner, where it is forced to watch the action from a fixed position, unable to move without revealing its presence. (Bazin, 1972: p87)
Looking then to the antithesis, the concept of a non-diegetic camera, we arrive at one that transcends and is beyond the spatiality and physicality of the scene; positioned omnipotent to it. This is essentially the camera equivalent of a voice-over in the sense that the camera is not divorced from the mise-en-scene, it depicts and might be considered to visually comment on the scene or add to the scene as Voice-over does, but, by the same token, doesn’t share the scene, isn’t derived from the scene and doesn’t occupy a position of spatial diegesis respondent to the scene.
A distinction must then be offered here between the long-take cinema of Bazin and the often continuous perspective of the virtual camera. For Bazin the strength of photo-graphic-based cinema was its reality which is most readily evoked by deep focus and long takes resonating as the antithesis of theatricality. Indeed for Bazin, it is the camera’s obvious diegetic positioning in the scene that gives the image power: “The aesthetic qualities of photography are to be sought in its power to lay bare the realities.” (Bazin, 2004: p15)
On the surface, the continuous and un-broken nature of the virtual camera – despite being routed in non-photographic processes - is much like the long-take in defying the artificial nature of montage and attempting to present a truthful visual capture of a ‘reality’ through unbroken depiction. But what separates the virtual camera from the long take, and indeed from much of the body of conceptual ideas of Bazin, is not the division between the photographic and the animated (a division increasingly difficult to articulate or discern) but a core value of fantasy embedded in a changeable state of diegesis. The virtual camera, by its very nature of physical intangibility, is rooted in a depiction of fantasy and the impossible.
There are numerous examples of the use of a non-diegetic camera and the rise of 3D graphic environments in cinema production has made them a significant factor in contemporary film making stemming from their intrinsic use in game design. The opening shot from Angles in America (Kushner, 2004) might be seen as a key example where the camera moves in one continuous take from a cloud bank high over the city of San Francisco, across continental United States, over Chicago and St Louis to New York, folding distance and time as it goes as traditional montage editing would do only it is a continuous and unbroken cinematic take. This shot seems to celebrate all the power of unbroken time in depicting space diegetically that is at the heart of Bazin’s long-take realism but does so completely engaged in an act of obvious and apparent fantasy.
Here the camera can be said to occupy a non-diegetic positioning as it transcends the physicality and spatiality that defines all other elements of the scene. It is not subject to the physical space, kinetic velocity or temporal depiction in the scene - people walk at a normal pace down below, cars drive at normal, real-time, speed but the camera folds time/space between each city it passes over. The camera is positioned in an omnipotent, non-diegetic position that is outside if the scene’s spatiality. The camera moves in a way that defies time and space; ethereally beyond it. The unbroken shot is visually the same as any long-take cinematic composition, indeed embraces and exploits many of the strengths that long-take presents in contrast to montage, but is none the less conceptually opposite to the Bazin philosophy of the long-take cinema.
Other defining examples can be seen in a number of David Fincher films, Fight Club (1999) and Panic Room (2002). In Fight Club, during the opening scenes, the unnamed character (played by Ed Norton) narrates as the camera, starting from outside the window of a sky scraper, flies directly down to road level, goes through the road itself to the underground car park beneath, then travels perpendicular through numerous solid objects, through the insides of a van, over a series of explosives and ultimately to final position in close-up on a second set of explosives in a more distant location under another building elsewhere in the city. Similarly, but on a different scale, in Panic Room the camera moves in a long continuous take through a stairwell space vertically, through a room, over the surface of table at a distance that would be physically impossible with a ‘real’ camera, through physical objects and into the inner workings of a door lock. In pure animated films we very often see the same sense of camera-work beyond spatial physicality and the chase sequence through the door factory in Monsters Inc (2001) is just one of many examples.
In all these examples the camera takes on a distinctly non-diegetic nature whereby it exists outside of the space-time diegesis of the scene, outside of physicality itself. The camera does not adhere to a diegetic and perceivable construction of space. The camera doesn’t share the spatial conventions or restrictions that all other depicted cinematic elements such as the characters, environs, physical objects and events in the scene do. As with Angles in America the camera is omnipotent and beyond the scene it depicts.
One method of conceiving this phenomenon of intangible and abstracted presence in a case-study such as Panic Room is the concept of the virtual camera as the anthropomorphism of space itself via the camera entity. Whilst the free flowing movement of the virtual camera presents as intuitively anthropomorphic with an engagement of space that is intimate, personal and directed - possessing properties associated with a ‘presence’ - the virtual camera’s dislocation from physicality moves this notion of an anthropomorphic embodiment into a new sphere. A mode of simulated anthropomorphism of the viewer via the space itself, through the eyes and awareness of the space.
Put in more simple and pragmatic terms this example scene from Panic Room can be viewed as a point-of-view shot from the perspective of the house and its holistic internal space. If the same shot had been captured using a physical camera it would have been forced to comply with physics of the space as well as the physicality of the camera apparatus. This compliance tangibly and unavoidably positions the camera as an entity occupying the space, rather than an entity of the space, and thus the viewer as embodied navigator of that space. Being however a virtual camera outside of physical constraints the POV of the virtual camera’s exploration itself becomes an act of ’seeing what the space sees’ rather than what the characters can or might see. This shift has implications for the perception of the diegesis of the camera and the perspective of the viewer as occupant/viewer/spectator/player/watcher/observer…
The Mediated and Unmediated camera
If we accept the notion that the camera (as entity rather than apparatus) can occupy a non-diegetic position in regard to cinematic space (delivered via the ethereal nature of the virtual, non-physical camera) then we might also re-consider the mediated nature of viewership in cinematic form. As has been explored in many contexts of cinema theory, the camera (as apparatus), along with the medium of delivery (the screen in various sizes and forms) and the subsequent depicted space (visible and off-screen), together construct a spatio-temporal experience that is by nature mediated in relation to viewer perception. In other words the viewer’s experience of the drama, narrative, events and actions of the cinematic construct are mediated by the intermediary nature of the camera and its associated mechanisms in the depiction of space and events.
The physical camera, as an apparatus that conforms to space and implements its own tangible elements of mechanics (lens aberrations, depth of field, shake and movements that are dictated by physical constraints), invariably presents a mediated experience for the viewer. The camera apparatus is the medium through which the experience passes to the viewer and subsequently the viewer is regulated to a position forever one step removed from the scene and space by a recognition of the camera’s identifiable mechanical properties.
This is a long established concept of cinematic reception but where the consideration of a non-diegetic camera potentially prompts new thinking is the idea of an un-mediated position of cinematic viewership. A position whereby the absence of spatio-physical constraints and a transcendent, omnipotent positioning of perspective constructs a cinematic presence of the viewer in the scene that presents as unrestricted, unconstrained and un-mediated in the context of diegetic elements. Where traditionally the camera apparatus serves as a mechanism of mediation between the depicted world and the viewer by introducing tangible and visible camera specific elements, the virtual camera, by contrast, introduces no such elements. This can be seen as a substantial and even dramatic shift in cinematic awareness of experience whereby we move from a position of cinema as a scenic/spatial simulation of the abstracted ‘eye’ to a fundamentally new conceptualization that is, in essence, an emulation of a spatially abstracted ‘I’; a consciousness of presence, a spatially removed perspective but one none the less intimate and evoked as non-diegetic viewer observation.
The presence of the viewer, the abstraction of the ‘I’ as scenic observer, divorced (possibly for the first time in cinema history) from apparent mechanical mediation, constructs an apparent purity of cinematic experience where engagement with cinematic space and cinematic depiction is total, un-regulated and intimate. This is not to say that by removing those camera elements that remind us as viewers of the artificial nature of cinema we move to a position of ‘total belief’ in the moving image as reality. The virtual camera by its non-physicality is just as obvious a cinematic device as rack focus and lens flare but we do evoke a very different conceptualization of the viewer’s perspective and presence within a given scene. The viewer’s position, non-physical, omnipotent and un-mediated by mechanical aberrations, delivers them an immediacy of presence both temporal and spatial where, as Branigan comments there is an intention to create “a continuous presence of the present” (Brannigan, 2006: p9); an intimate immediacy of experience.
Diegetic and Mimetic condition
Another way of considering diegesis through spatial composition and the virtual camera (one that reinforces the idea of a simulation of ‘I’) but one still inextricably connected to notions of Mediation and the ‘Medium’, might be to use opposing conceptualisations of Diegesis and Mimesis as the marks of distinction, the separation of viewership evoked by the virtual camera. To do this we can simply use the common definition of Diegesis and Mimesis as the difference between ‘Showing’ and ‘Telling’. Understood in narrative thinking, whereby important narrative elements are revealed and unveiled in very specific terms, in specific order, and in specific context of the scene and story; a context of restriction is enforced by the diegesis where the narrative relies as much (if not more) on what the viewer doesn’t know and what order in which they find out, as much as the details themselves. Subsequently with the diegetic camera in the space, a distinct positioning is created whereby the mediating elements - camera, space, physicality - are ‘telling’ the viewer the story.
Mimesis on the other hand is conceptualized from an act of ‘showing’ not ‘telling’ and so potentially possesses a stronger notion of a narrative immediacy of experience. This in turn implies an absence of mediation; an absence of a conveyor or interpreter between the viewer and the event. A direct and un-mediated presence where the event/scene/environment is not communicated to the viewer but is experienced by the viewer via an internal logic of omnipotence and mimesis that is divorced from diegetic constraints.
Here we might return to the case study of Panic Room with the concept that the virtual camera allows us to ’see what the space sees’ and, pushed a small step further, to experience the events of the impending break-in to the house as the space itself experiences them. This can well be interpreted as a Mimetic construct of engagement, one where the viewer is ‘shown’ rather than ‘told’ of the events as they unfold in an unbroken spatio-temporal manner. The difference between “what’s happening now” as opposed to “what happened then” (Cubitt, 2004: p32) enforced by unbroken time and un-interrupted space.
The question then is to consider whether the result is any different, particularly in the context of human perception and memory working much more like snap-shot montage than continuous perception. Certainly in recalling a scene a viewer is more likely to remember and recall the individual moments of the sequence rather than the sequence itself, having imposed an in-frame montage despite the continuous take. However there is, none the less, a distinct difference of experience in the moment of the sequence (if not in its post viewing recollection) that engages a very different user perception of cinematic engagement. An engagement that is derived from a mimetic discourse of the virtual non-diegetic camera.
The unbroken, continuous temporal and spatial nature of the virtual camera long-take removes the innately segmented nature of ’story-telling’ where by an element is introduced, established, undermined and concluded (there was this, it was there, then there was that, then this happened). A traditional approach to the same scene from Fight Club would invariably have been built of physical camera movements adherent and compliant to the space along with montage editing from establishing shots to close-ups of action, continuous through the series of events leading to the final entry point for the burglars. This methodology of establishing place, showing detail, depicting event, showing ramifications or results of event, withdrawing to larger scene, cut to next and repeat; follows a narrative internal logic of specific sequential unveiling that conforms to the idea of auteur control. This type of cinematic construction is essentially dictation – compositional elements dictated to the viewer with an air of hindsight and past tense narration; ‘this is how it happened’ instead of ‘this is happening now’.
A mimetic construction of the same scene, by contrast omnipotent via virtual camera, allows the viewer to see rather than be shown; that by virtue of an holistic and unbroken perspective abstracted from physicality the viewer is engaged as a presence, the simulation of ‘I’, into the scene. This is arguably exactly what the virtual camera in Panic Room allows in the way of viewer immersion. With unbroken, uninterrupted space/time, with neither cutting or physical restriction, the viewer experiences a Mimetic showing of events that is largely without elements of hindsight reconstruction or mediation. Davis comments on cinematic Mimesis as that which “involves a framing of reality that announces that what is contained within the frame is not simply real. Thus the more “real” the imitation the more fraudulent it becomes.” (Davis, 1999: p3) In other words the concept of mimetic form relies on the viewer perceiving of the depiction, the constricted imitation, as more than real, hyper-real, beyond real. This is necessary in order for the process of being ‘shown’ events to be experienced mimetically. This is acutely what the virtual camera does with its innate mode of fantasy as opposed photographic-based long-take rooting in corporeal reality. By abstracting itself from physical space, transcending spatial restriction it delivers to the viewer a moving image experience that is directly perceived as not ‘real’ but beyond ‘real’, drawn from a lack of spatial diegesis that would otherwise be present with the physical camera apparatus. Put simply, when the camera flies through a wall or into a door lock we are immediately presented with an obvious affectation of non-reality and as such, a greater opportunity to connect with the idea of being mimetically ‘shown’ a story rather than diegetically narrated one.
The Vanishing Point
So what is the virtual camera? Is it part of a natural evolution of optical emulation that simply shakes off the shackles of a physical constraint? Or does the Virtual Camera represent a shift more fundamental to our perception of both cinematic form and our place as viewers, observers and occupiers of cinematic space? A useful connection may be drawn between the virtual camera and what James Ferwerda has referred to as “Functional Realism” (Ferwerda, 2003). Ferwerda uses the example of a technical cross-section illustration to understand the notion of functional realism; a diagram that shows not just an indexical based representation of identifiable forms but one that positions the viewer’s perspective on the diagram in such a way that the vantage point is a physical impossibility. This perspective is perceived to be one which none the less ‘realistically’ represents the scene and provides exactly the visual information required by the viewer to conceptualise the depiction in real-world space. The specific example he gives is one of a hand removing a nail form a fence with a tool where not only is the hand, tool and nail shown but also, in x-ray-like cross-section, the depth and angle of the nail into the timber of the fence itself.
In this case our acceptance of the ‘real’ is defined not by the diagram’s connection to human sight and real-world positioning and depiction but rather by the diagram’s usefulness to inform a real-world understanding. Functional Realism is “defined in terms of the Fidelity of the information the image provides. If an image lets you do the task you need to do, and allows you to perform the task as well as you could in the real-world then, for that task, the image is realistic.” (Ferwerda, 2003)
This notion of Functional Realism becomes particularly pertinent to contemporary cinematic form in the context of the virtual camera transcending physical spatiality. Moreover functional realism is equally applicable to both live action-based and animated films along with the increasingly common hybrid ground between. The functionally realistic image makes “it possible to show viewpoints that would be difficult or impossible to photograph… (with) artificial transparency to depict important features that would be hidden in photographs.” (Ferwerda, 2003)
This notion of Functional Realism makes for a powerful framework for conceptualising the virtual camera within a greater context of realism tied to the practical and unique properties of the virtual camera over the physical camera. More importantly, or at least more philosophically, this understanding of functional realism forms the final element in reconsidering what the virtual camera is outside of any relation in nomenclature to a mechanical camera. With the apparatus removed, a non-diegetic framework opened up for spatial occupation, a viewer perception of un-mediated access experience and a composed space, what remains is the conceptualization of the virtual camera as spatially placed vanishing point of perception; the spatio-temporal moment of a scene where vision perspective converges to a refined, accessible point.
This vanishing point is highly mobile, deliberately functional, perspectively omnipresent, un-mediated and tangibly focused on an viewer experience that is pure and unique cinema. The cinematic vanishing point provides a vision that is consistent across the ever widening diversification of cinematic delivery platforms (from the hand-held and mobile to the domestic and theatrical) and delivers an experience that is cinematically autonomous, unable to be obtained from any other art form. All these elements demand a consideration of the virtual camera as not just a hybrid, evolutionary or progressive movement from the physical camera but rather one of profound conceptual shift in what we see and how we see it.
—
Cited Texts and Works
Anderson, Paul Thomas. (1997), Boogie Nights. Dir P. T. Anderson, New Line Cinema.
Bazin, A. (1973) Jean Renoir. Trans W W Halsey II and William H Simon. New York: Simon and Schuster.
Bazin, A (2004) What is Cinema? Volume 1. Trans Hugh Gray. California. University of California Press.
Bolter, J & Grusin, R. (1999) Remediation: Understanding New Media. Cambridge: MIT press.
Börge, V. - Translation Snell (1962) ‘Cinema: A Camera Within Us’, The London Magazine, vol. 1 (no. 12)
Branigan, E. (2006) Projecting a camera: language-games in film theory. New York: Routledge Taylor & Francis Group.
Cook, P. (ed.) (1985) The Cinema Book. London: BFI Publishing.
Cousins, M. (2004) The story of film. London: Pavilion
Cubbit, S. (2004) The Cinema Effect. Cambridge: MIT Press
Docter, P. (2001) Monsters Inc. Dir. Peter Docter, David Silverman. Pixar.
Ferwerda, J. (2003) ‘Three Varieties of Realism in Computer Graphics’ , URL (consulted February, 2006): www.graphics.cornell.edu/pubs/2003/Fer03.pdf
Gray Matter Interactive/iD Software/Nerve Software (2001) Return to Castle Wolfenstein. Activsion.
ID Software (2004) Doom3. Activision
Koepp, D. (2002) Panic Room, Dir Fincher, D. Sony Pictures.
Maxis. (2004) The Sims 2, EA Games.
Ndalianis, A. (2000) ‘Architectures of the senses: Neo-Baroque entertainment spectacles’, in Rethinking media change; the aesthetics of transition. Cambridge: MIT Press.
Pileggi, N. (1990) Goodfellas, Dir M. Scorsese, Warner Bros.
Remedy Entertainment. (2001) Max Payne, GodGames.
Uhls, J. (1999), Fight Club, Dir Fincher, D.
Valve. (2004) Half Life 2, Sierra.
