The Depth of Adventure, the Volume of Discovery: Feeling 3D Space


[This blog is based on my 2016 SCMS presentation]

Stereographer Brian Gardner (Coraline, Life of Pi) describes creating “an emotional undercurrent” through depth and volume. He writes:

“when the 3D goes deep behind the screen, you get this large empty space, this feeling of the grandeur of God, the vastness of the possibilities. I generally like to put that right at the act one climax, as the character goes off to explore the new world, to convey that sense of adventure. That’s an example of what I mean by using 3D depth to underscore emotional dynamics.” [“Perception and the Art of 3D Storytelling” Creative Cow 6, 2, 09:

This way of describing things absorbs stereoscopic space into a formal system of underscoring, akin to music. Professional manuals echo this by stressing the dangers of “fatigue,” “eyestrain” and the importance of correlating depth brackets with narrative development. Comfortable viewing and story-centeredness are often presented as interdependent aesthetic ideals that will ensure stereoscopy a place in Hollywood aesthetics – a place in support of dominant formal registers. But must 3D stop at the point of underscoring, supporting, and repeating things already available in 2D?

The question of exactly what 3D contributes to the film experience has been rich and productive for film scholars. This line of inquiry is best exemplified by Miriam Ross’s book 3D Cinema: Optical Illusions and Tactile Experiences. She argues that stereoscopy engages viewers in a distinctive affective register, a “hyper-haptic” relationship to the image that is peculiarly corporeal as well as ocular.  Here, I contribute to an ongoing discussion by mapping some formal qualities and techniques artists have used to harness this relationship. By looking closely at three-dimensional volume, a perceptual quality that can be felt but not always recognized, we can better understand the historical process of narrativizing stereoscopic effects. Dimensional volume can create emotional engagement with a 3D story world distinct from the 2D version.



First we should say something about the nature of volume in the stereoscopic space. Depth refers to distance between fore and background in a frame or the distance between objects within dimensional space. Volume is related, and a little more complex. In part, volume depends on the aspect of an object revealed by the camera. Rudolf Arnheim proves useful here, because of all the aesthetic theorists he brought a keen sense of the film as a 2D plane emulating 3D space. Volume in 2D is related to what he called the choice of aspect: a camera can present either a flat square or a cube depending on how many facets of the object its position reveals. Shading also provides cues about surface shape, hence modeling light can present the camera with more or less apparent roundness. These are compositional qualities of volume, which persist in the stereoscopic space. Related to compositional volume is ocular volume, the perceived roundness determined by choice of lens. A long or narrow lens tends to magnify all parts of the image equally and compress peripheral information. The “telephoto” effect. Wide angle lenses increase access not just to the front but to the sides of the object, and pull the planes apart through 2D cues, creating both roundness and depth. And then there is movement, both within and of the frame, which contributes kinetic volume. The rate of movement of objects in relation to one another gives us information about their position in space. Meanwhile, movement itself reveals more aspects and alters overlap giving us cues for roundness. This is kinetic volume.

Compositional, ocular, and kinetic volume cues exist in 2D and 3D, but stereo provides additional variables. One is interaxial distance, or the spacing of the right and left lenses.  Interaxial manipulation, moving the cameras closer or farther, can shift the way a visual space feels, its habitable expanse and tactility, without dramatically changing composition or entailing visible movement.


This is a kind of ocular volume specific to 3D. Increasing the interaxial, for instance, can open up space between objects, reveal the roundness of surfaces. 3D also offers a new contribution to kinetic volume by allowing filmmakers to move not just the frame but the window that surrounds the image independently of the space represented. This is the Dynamic Floating Window, which accommodates objects in negative parallax by tilting one side of the stereoscopic frame toward or away from the viewer.


The creation of volume and depth in 3D depends on the interaction of all of these variables, many of which are also available in 2D. Perceivers cannot readily segregate them. This is important to filmmakers, because it allows 3D to hitch a ride on familiar techniques and qualities. It creates a sort of safety zone of compatibility between 2D and 3D versions of the same film. As Nick Jones explains in his recent Cinema Journal essay, expressive 3D effects tend to operate in accordance or even augment 2D cues. Jones observes that films like Dredd and Resident Evil: Afterlife

“orchestrate stereoscopy to make it both elemental to the visual experience and also, paradoxically, entirely redundant, thus allowing the film to be a cogent piece of entertainment in either 3D or 2D” [“Variation within Stability Digital 3D and Film Style” Cinema Journal Fall 2015, pp. 52-73].

The shared territory of volume and depth cues encourages thinking about 3D as an adjunct to existing formal grammars, a means of amplifying extant storytelling techniques. But I am particularly interested in the ways that 3D volume and depth expands the nature of cinematic engagement, the way it can create a new way of feeling the image.



As I’ve argued elsewhere, stereoscopic depth is a good candidate for filmmakers seeking to bind 3D to storytelling tasks.  For instance, from its opening shot, James Cameron’s Avatar inscribes character subjectivity into depth and volume. Avatar mobilizes a two-world scenario, with more pronounced depth and volume in Pandora’s natural world than the human’s military base. The visual shift is cleverly motivated, in part, by the Navi’s relatively wide inter-ocular compared to humans.






But the film opens before we are introduced to Pandora, with the hero’s dream of soaring and plunging over a forest as though he were flying.

Slide015 Slide016 Slide017 Slide018

Voiceover efficiently flags the formal pattern.

“When I was laying there in the VA hospital with a big hole blown through the middle of my life I started having these dreams of flying, I was Free.”

We are told how this display of kinetic and ocular volume is meant to feel as we are watching it, and it is immediately contrasted with Scully’s “real” surroundings – a coffin like chamber of compressed space.

Slide022 Slide023 Slide024








Avatar’s visual movement, then, is not just between a grounded and a fantastical world, but also connected to character journey from a constrained and compromised human form to free and righteous Navi. The corporeal experience of depth and volume here gets bound to narrative meaning.

In 3D films like Avatar, Coroline, Tron, Oz the Great and Powerful, and Life of Pi, the division into two worlds motivates opportunities for expressive and spectacular dimensionality. This solution serves James Cameron’s extra-textual, assimilationist rhetoric, as it both draws attention to his manipulation of form within the film (characters AND viewers notice this difference) and announces that 3D can be yoked to character development. But if you were to see only a 2D version, you have still been told of the contrast between worlds through other means.

Subsequent films have begun to develop stereoscopic volume into an expressive grammar that can operate with or without a diegetic tether. This involves keying changes of volume and depth to an emotional experience, which can be aided by a depth script. Pixar’s stereoscopic supervisor Bob Whitehall explained that for UP:

‘We created a graph of how we would use stereo. In the beginning when the character is happy there is a deep space, then it flattens out when he loses his wife and then it slowly increases throughout the film. Just like the lack of color in dark scenes make the vibrant images stand out more, so do the flat scenes enhance the scenes where you are more aggressive with the 3D.’ [“Finding the Language of Stereo 3D” Studio Daily, 8, 20, 09:]

UP’s famous marriage montage demonstrates this emotional mapping of volume. As Carl and Ellie make a home together, Whitehall delivers deep compositions with well defined planes and fullness.

Slide027 Slide028Slide029 Slide030

Ellie’s consultation after her miscarriage contrasts in nearly every way – the camera moves laterally across flattened planes. Stereo volume, like color, has been emptied out, the characters are confined to a narrow aperture that isn’t so much in the depth of the shot as sliced out of a single plane.


Depth and roundness return as Carl cheers Ellie up with promise of future travel.

Slide035 Slide037 Slide038

When Ellie dies, volume once again compresses. Her hospital room is handled like the Dr.’s office, with lateral tracking and movement from side to side rather than into dept

Slide039 Slide040 Slide041 Slide042


The change is most apparent if we compare rhyming shots from the start and end of the montage. Carl and Ellie’s wedding plunges from fore to rear, but the church after her funeral has shallow stereo (though the 2D cues remain about the same).

Slide044 Slide045

When Carl and Ellie first arrive at their new home we are given an expanse for them to travel between the street and the porch. But when Carl returns alone after the funeral, fence, porch railing, and doorway stack up against one another behind the screen plane.

Slide046 Slide047

Stereoscopic volume ebbs, along with color, light, movement, and Michael Giacchino’s score. On one hand the 3D manipulation is redundant with other expressive devices, perhaps emphasizing but not replacing 2D cues. On the other, this squashing of space is less observable than the drop in color saturation, change in lighting, deceleration of character movement. Less observable but nonetheless felt, as though the air goes out of the scene.



A more common way of drawing dimensional volume into a film’s expressive fabric involves punctual shifts from shot to shot. How to Train Your Dragon 2 offers a spectacular example. When Hiccup meets his long-lost mother and visits her dragon sanctuary for the first time, the thrill of discovery comes with a burst of color, light, and space. We begin by tracking Hiccup’s progress down a narrow tunnel, action confined to the center of the frame. Dark rock and shadow surround him, as our point of attention converges on his face within a constantly moving background.

Slide048 Slide049

Slide050 Slide051

His emergence from the cave wall delivers a series of arresting views each more dynamic than the last.

Slide056 Slide057Slide058

The dragon flock rotates around a stone column, circling toward and away from the camera as individuals enter and leave the frame.


Meanwhile the camera arcs first with the flow of dragon traffic, then in the reverse shot, against it. Both rotations accentuate kinetic volume. The world expands in all directions, and the viewer’s eye meets the harmonious flow of winged creatures.

Slide061 Slide062

Slide063 Slide065


Slide067 Slide068


Where Hiccup’s tunnel trip forces us to track a central point of negative parallax, views of the sanctuary allow attention to range wide; the tight, somewhat uncomfortable narrowing of focus gives way to open space ripe for exploration. In 3D the change encourages us to literally relax our gaze, un-flex our eyes’ muscles, as we drink in the light and color. This punctual change to mark a plot point recalls Brian Gardner’s description of the “feeling of grandeur.” Specifically, the moment reverses the opening of Avatar: our experience of constraint and containment in the tunnel gives way to freedom and spatial exploration. But where Cameron built this into his film’s formal and thematic arc, in Dragon 2 it has become a passing punctuation, a momentary expressive burst available to the animators.

Animated family films offer elastic worlds of loose verisimilitude that stretch and bend to the technology, as well as precise moment-by-moment control of parallax and convergence. Verisimilar, photorealist drama confers 3D more grounding, which generally narrows opportunities for stereo manipulation. Since Avatar, the most successful photorealist dramas have been otherworldly, and in the case of Gravity, every bit as animated as a Pixar release.  Still, character-centered drama is something like the last frontier of assimilation for stereo. If you can make it there, you’ll make it anywhere. Hugo, Life of Pi, and Gravity belong to the distinguished circle of Oscar winning 3D films, and, though none of them has kicked off a significant production trend, they each make a case for stereo in mature, critically respectable fare. I want to turn to Hugo and Pi to explore subtler, less redundant alignments between stereo volume and emotion. I’ll leave Gravity aside for the moment, partly because I’m short on time, and partly because I ask my students to work that film out on their own.


3D for Scorsese is a way of rethinking cinema – of renewing it, bringing back the wonder of seeing things move. It also gives new life to some of the director’s favored stylistic flourishes. In Hugo the characters develop by discovering the beauty and art of cinema, a cinephilic journey that we are invited to join. For example, to underline the moment of union between the orphan Hugo and his new adoptive father George Méliès, Scorsese employs a “vertigo” shot: the simultaneous zoom-out and dolly-in made famous by Hitchcock and elaborated by Scorsese in Good Fellas. The trick amounts to a 2D manipulation of ocular volume in which the background appears to encroach or recede across the shot. Scorsese’s 3D version differs in intent and visual effect. Rather than express the pictured character’s psychological state, here the dolly/zoom is connected to Hugo’s point of view. It is what he sees as Méliès speaks to the audience and begins the show.


Slide072 Slide074

Slide074 Slide075

Slide077 Slide078


Slide080 Slide081

Slide082 Slide083

Slide084 Slide085


In this context the technique takes on the feeling of enchantment and intensified attention, rather than terror or disturbance. Stereographer Dimitri Portelli enhances the dolly/zoom by widening the interaxial across the shot. As the background falls away, the air between planes becomes apparent and the actor’s features gain palpable roundness. Where the 2D version of the effect connotes intensity and psychic stress through contradictory spatial cues, the 3D version makes an odd kind of embodied sense. It is the experience of spatial expansion and tactile proximity. Less a filmmaker’s stylish intervention (though it remains that) than the world opening to the character and viewer. Appropriately, this is prelude to Scorsese’s display of dimensionalized Méliès films that directly address the viewer as attraction. A return to exhibitionist early cinema.


But the dolly/zoom, is embedded in character subjectivity, and it provides a spatial experience distinct in feeling and function from 2D versions of the technique.



Finally, in Life of Pi, Ang Lee and Brian Gardner offer a variation on volumetric change within a shot. During his interview with Japanese insurance adjusters, Pi spins a rational if ugly explanation of his time at sea. Events that we have seen brilliantly rendered in plunging parallax and surreal abstraction are reframed by a seemingly unembellished monologue.  Lee strips the image and gives us the character’s face in an over-exposed field of white. Shading cues are virtually absent.

Slide091 Slide092

Across Pi’s speech, we experience a very slow graphic change. Distance between viewer and character collapses.

Slide093 Slide094

Slide095 Slide096

Pi’s face progressively gains volume until it is bulging from the center of the screen. The effect appears to be some combination of dollying forward, increasing the interaxial and widening the lens.

Slide098 Slide099

Slide100 Slide101

Stereo manipulation intensifies the move into close-up, but does it so slowly and within so spare an image that the change seems isolated from 2 dimensional cues. We find ourselves physically near the narrator, sensing the creep in but only noticing the new shape of space when a cut back to the frame story snaps characters behind the screen plane and out of our personal space.

Slide102 Slide103


The dimensional volume gives Pi’s account an almost embarrassing intimacy absent from 2D. Though this is the less appealing version of events, and despite the weight of cinematic spectacle given to the scenes at sea, Pi’s gentle emergence, the plastic proximity of his face, renders the speech as an emotionally raw confessional. In 2D the tone is somber frankness, which continues when we move back to the present day. In 3D, the image flattens upon our return to Montreal, the spell of the story is broken, the world deflated. Pi’s truth is still a story told in a different space.



These examples illustrate some narrative potentials of dimensional volume. On the spectrum of 3D effects, volume change can be unusually subtle, shaping our experience of the story world with quiet force. This might allow it to reach beyond the realm of redundant emphasis, in which stereoscopy repeats functions shouldered by other registers. Not only is our visual experience different from 2D, our physical sense of space and our emotional understanding can be distinct. We can glimpse 3D’s potential to alter storytelling not only by degree but by kind.







This entry was posted in Uncategorized. Bookmark the permalink.