Soft Fascinations in Shallow Depth

Imagine stepping off the streets of Tokyo into a dark and quiet gallery space. In the far corner is a five-meter-wide panoramic image, glowing in high-definition, four times wider than it is tall. The upper and lower edges of the frame are blurred, softening the border between the projected light of the image and the darkness hugging it from above and below.

Fade from black to a series of lush floral compositions in soft tones. The white petals of a lotus flower fill the entire frame as everything in the image slides slowly to the right. This partially dissolves into two more views of the same bloom, now drifting slowly downward, revealing a large grouping of mint-green stamen rising from the center of the blossom. As the different images slide across each other, their diagonally radiating petals mesh, casting shifting shadows across the surface of the screen.

The framing hugs the petals tightly. The flower never appears in its entirety, pulling viewers in toward the gentle curves and soft textures. Some parts of the image are semitransparent, giving a partial glimpse of the layers underneath. The petals appear supple and porous as felt. But there is something unnatural about their pristine high-definition surfaces. They glow in an uncanny space between photo-realism and a vacuum of digital purity.

A minute in, dark reflections appear across the surface of the image. These sliding patches of shadow eventually coalesce into the outline of a face. More shadows appear, slowly drifting horizontally across the screen. A title appears superimposed across the bundle of stamen as the music swells for the first time: “Swimming in Qualia.” The flower image fades to black, and then, slowly, the title follows it into the darkness (Figure 13).

These are the opening moments of Swimming in Qualia, an ambient video installation presented as part of the STILL/ALIVE group show at the Tokyo Metropolitan Museum of Photography (December 22, 2007, to February 20, 2008). The visuals are by Ise Shōko. The music is by Steve Jansen (1959–), electronic music composer and former drummer for Japan (the English rock band).1 Through close attention to the role of image compositing in this and related works, in this chapter I focus on how ambient techniques of self-care are translated into a digital audiovisual register.

My emphasis is on the compositing of what I call shallow depth, a space of relationality both literal and metaphorical seeking to situate a narrow space of autonomy within a complex assembly of external controls. As this chapter examines, the shallow depths of ambient video present viewers with a degree of freedom, allowing attention to drift in and out while providing a sustained mood of indeterminate calm. The depth of the visual compositions provides a hint of the unknown, but not enough to overwhelm the relaxed and freely drifting awareness cultivated by the shifting surface layers. Yet these tightly controlled digital images, built through the airtight vacuum of software compositing, form a potential challenge to the presumed autonomy of the contemporary viewer. The ambient videos I examine avoid this threat by seeking out contingent freedoms in the footage imported into the compositing space, using these to determine the shape of the final video. This, too, allows a shallow space of autonomous movement to drift in through an otherwise overdetermined digital landscape. In both ways, the shallow depths of ambient video provide coordinates for the postindustrial self, freely floating through and among accumulated layers of data.

Compositing in Shallow Depth

In installation form, Swimming in Qualia has two video channels: the first, “Ascent,” consists of a slow-paced loop of around twenty-four minutes in length. A second projector displays a much shorter loop, “Glimpse,” six minutes long and organized around a faster but still indeterminate cycling. Both channels are mixed from the same set of imagery: rippling waters, shifting silhouettes, foggy forests, railroad tracks slipping by, and a lush landscape of lotus blossoms and leaves. The two parts are projected on adjacent perpendicular walls, so the two images meet flush in the corner of the room. The longer section, “Ascent,” was later projected as a single-channel work during a live concert by Steve Jansen at the Meguro Persimmon Hall in Tokyo (February 29, 2008). Live musicians arranged in front of the screen performed an improvisatory version of Jansen’s score.2

From the felt-like flower petals to the gently rippling water appearing later in the piece, most of the material in Swimming in Qualia registers to vision as soft to the touch. What is not soft in itself is wrapped in a substance that is: trees in the forest are bathed in fog; railroad tracks are blurred with movement. Colors are saturated and smoothed through shading and added grain, and a soft twilight wraps the environments in warm tones. The visuals glow gently in the darkness of the projection space. Soft focus blurs hard lines, particularly around the edges of the image itself (Figure 14). Ise avoids hard cuts in favor of gradual dissolves. Rhythm, particularly in the “Ascent” portion of the work, manifests as a slow ebb and flow as the dynamic pulse of sounds and images gradually accelerates and decelerates over time. The video appears to breathe softly with this undulating pulse as the images drift across the screen.

To share a dark space with these audiovisuals is to be pulled out of the task-oriented, identity-based self of everyday life and toward an impersonal immersion in pure sensation, as promised by the title of the work. Qualia is a philosophical term referring to the subjective contents of sense perception, the various qualities perceptible in objects only through the medium of the human senses. The soft porousness of Swimming in Qualia’s stimuli produces a blurring effect on the subjective level. The feelings produced are highly sensual, but the objects of these feelings shift and slide as the piece progresses, never stabilizing around a single focus. These ambiguous sensory impressions provide an experience of “swimming” with the qualia themselves, rather than a more direct encounter with the objects giving rise to them. In so doing, the video avoids reference to any more stable form of reflective human identity, allowing audience attentions to disperse and a more impersonal, phenomenological mode of awareness to emerge.

Unlike the more closely integrated rhythms of the works discussed in the previous chapter, Swimming in Qualia sits at a further remove from the pressures of everyday life. Audiences leave the noise of the Tokyo streets to slip into the cool darkness of the exhibition space and drift among these indeterminate sensations. The lush yet muted textures and the deliberately slow pacing of the video guide these dissolving bodies and pull them toward a more relaxed, meditative, and yet mobile state. People in the audience are together in their shared orientation to the screens but dissolved as people, blending with the larger atmosphere. And yet, at the same time, this escape from self-concern produces a form of subjectivation eminently suited to the neoliberal focus on self-care.

The spatial logic of Swimming in Qualia has similarities with older traditions of image composition that render depth not through orientation to a distant horizon but through a layered accumulation of surface textures. Complicating this push toward flatness, however, is the presence of photo-realistic three-dimensional space. This brings us to the central aesthetic problem in ambient video compositing: how to blend flat, two-dimensional graphic surfaces with three-dimensional photographic images.

The seeming opposition between “surface” and “depth” has long structured discourse on modern Japanese aesthetics. The contrast is often narrated as a choice between a native and/or “postmodern” mode of composition, “in which depth comprises layers of planes without regard to graduated perspective,” and the geometric mapping of three-dimensional space arranged around a vanishing point, the singular essence of the Cartesian self.3 In this reading, perspectival depth is associated with centralized power, control, and hierarchy (famously materialized in Jeremy Bentham’s panopticon), whereas (as early postmodern critics liked to argue) the play of surface layers is inherently more free, unsystematic, and outside hierarchy.4 As Thomas Lamarre notes in his work on the animation table, however, there is nothing inherently less rational about organizing depth not through gradated perspective but through distributed planes. In fact, most science and engineering has long since moved on from the fixed perspective of Cartesianism to explore more layered and flexible ways of mapping the world. Envisioning a three-dimensional space as a series of two-dimensional layers provides a much more precise (and machine-friendly) mapping of relations between points in space. Lamarre gives the example of assembly diagrams produced in orthogonal perspective, where a three-dimensional object is mapped in two dimensions by tilting it forty-five degrees from the page and envisioning it as a series of parallel two-dimensional layers—as in the “exploded view” showing how the components of a machine fit together.5 But there is more to this than simply a greater compatibility with machine vision. Visual perception researchers have argued human vision may be fundamentally oriented toward distinguishing between layers, with the sensation of depth only a later, higher-order addition inferred through the shifting of surface planes. As James Jerome Gibson argues, “What we see is not depth as such but one thing behind another.”6

If this is the case, why were postmodern theorists able to see a world of distributed planes as something beyond hierarchy and less susceptible to control? Lamarre argues this seemingly “flat” space could be understood as more unsystematic and free only through contrast with the Cartesian model and the totalizing gaze it implied. In Japan this cofiguration was especially overdetermined, with the latter identified with “Western” colonial modernity and the former with a pre- and/or postmodern Japan simultaneously more authentic and more futuristic. Postmodern Japanese thinkers sought to position the distributed-plane model outside modern systems of control and surveillance when, in fact, this could not be further from the truth.7

While insightful, Lamarre’s explanation does not fully account for the broader postindustrial turn to imagining space as a series of distributed planes, with or without the need to reject the Cartesian or invest in facile assertions of Japanese particularity. I suggest one significant reason layered space came to appear more “free” starting in the late 1970s was it served as a useful spatial model for the emerging neoliberal self and its own desire to be distributed, flexible, and indeterminate. As Rose describes, the rise of neoliberalism marked a shift from a psychological model of “hidden depths” at the core of the person to an understanding of the self as a “flattened field of open circuits.”8 In other words, neoliberalism no longer seeks to ground identity deep within but rather conceives of the self as a series of distributed nodes—a composite of data layers. This transformation comes with a biopolitical twist: distributed and dispersed, people are liberated from the all-seeing gaze of centralized power—providing an actual, verifiable sense of freedom—while their bodies are opened up to new forms of data manipulation and surveillance far below the threshold of a unified perceiving subject. The “flat” rhetoric of postmodern self-determination and autonomy obscures this rise in environmental controls, emphasizing only the gains in personal freedom and mobility. But the two cannot be separated. To return to Foucault’s comments from The Birth of Biopolitics lectures discussed in the introduction, this is a society “in which the field is left open to fluctuating processes, in which minority individuals and practices are tolerated, in which action is brought to bear on the rules of the game rather than on the players, and finally in which there is an environmental type of intervention instead of the internal subjugation of individuals.”9 While Lamarre finds possibilities for divergence within this “distributive field,” the intervals between layers also allow for a more precise fixing of bodies in space.

Nowhere is this relationship between environmental control and personal flexibility more vivid than in the realm of motion graphics compositing software. Lev Manovich’s Software Takes Command describes how a layered form of spatial organization gradually became dominant as software engineers sought to incorporate practices of graphic design, animation, filmmaking, motion graphics, and photography into the unified space of moving-image compositing programs. By the late 1990s, previously incompatible media genres became integrated into a hybrid and mutually intelligible computer workflow. This was thanks to the emergence of a range of different image-compositing programs, including Photoshop, Illustrator, Flash, Final Cut, After Effects, and Maya. The techniques and traditions of previously distinct fields—for example, graphic design and digital video—were now available in a single unified interface. The consolidation of these varied media into a small number of compatible software platforms effectively blurred the professional boundaries between different fields as well, allowing even small firms and freelance designers to operate in several fields simultaneously.10 As noted in the previous chapter, this was the dawn of the Japanese eizō sakka, multimodal image creators working in a number of different media at once—and layering them together within a single software space.

For much of the twentieth century, bringing different media types within a single work (for example, hand-drawn animation, live-action imagery, and typography) meant editing them sequentially (moving from one form to another) or combining them within a single, static image. Compositing moving images of different kinds simultaneously was both labor intensive and technically complex. With the advent of motion graphics software, however, artists gained the ability to mix numerous moving-image layers within the same software space, precisely adjusting their position, opacity, and movement across a work’s running time and between its image layers. As Manovich notes, with the advent of moving-image compositing, visual media were catching up with the studio revolution in recorded music, where multichannel recording had generated a new spatial musical aesthetics from the 1970s onward.11

The creative process shifted from a resolutely time-synced mode of moving-image construction, with the frame as the indissoluble unit of a linear montage sequence, to a nonlinear software canvas composed of groups of objects arranged across a layered spatial field, each of which may be discreetly modified in size, shape, and movement. The hard cut became less important in this new aesthetic, in exchange for visual forms organized around the “continuous transformation of image layers” and rhythms of “rewriting, erasing, and gradual superimposition.”12 As with the rise of multitrack nonlinear music production, the shift in focus to a nonlinear model of distributed layers in motion graphics brought the practice closer to the world of “flat” two-dimensional media. Painting, animation, photography, and 2D design became increasingly prominent reference points for the emerging aesthetics of digitally composited motion.13 With digital compositing software, all forms of audiovisual media were brought under the guiding framework of the distributed multiplanar image.

Even with this new hybridity, however, a larger aesthetic problem remained: how to integrate the animation and graphic-compositing techniques of the past with the flexibility and portability of the video camera and its lens-based perspective on existing three-dimensional environments. By the early 2000s, compositing software had begun to incorporate 3D computer graphics and architectural modeling as well, enabling software users to directly manipulate 3D shapes. As Manovich argues, this was a qualitatively different approach from the essentially 2D space of 1990s compositing. Whereas 2D compositing software treats everything as a flat 2D layer, 3D compositing treats even 2D layers as objects to be positioned in 3D space, not unlike the arrangement of screens in a gallery. These two modes of compositing each foreground different aesthetic concerns: as noted, the former draws heavily on the history of 2D media, although with increased degrees of control. The latter draws more from architecture.

In both cases, however, the software remains object oriented, or what Manovich calls “modular.” Each element in a composition, whether placed on a 2D layer or in a 3D space, is defined as an object with particular values such as size, orientation, transparency, and an array of other textures and effects. This includes image data imported from outside the software, such as video clips and photographs. Every object imported into the software space is immediately situated in virtual depth in relation to other objects. Instead of the black box of the frame (or the cel, in cel animation), the malleable software object becomes the irreducible compositional nucleus within compositing software. From the perspective of the software, “a media composition is understood as a set of independent objects that can change over time.”14

Manovich does not emphasize this point, but I want to draw attention to the persistence of the object here as the one element resisting dissolution within this new regime of spatial organization and control. The properties of the object may be radically transformed, and the object can be duplicated, split, or deleted at will. But within the logic of the compositing software, the frame of the object remains absolute. For compositing software, the object is an indissoluble building block, there to be manipulated but never compromised in itself.

While the underlying code might remain hidden, the software object itself is fixed in place, transformed into a set of data points. When humans encounter an object in lived 3D space (outside software), it comes conditioned by what Gibson calls the “ambient optic array,” the visual cues helping us situate an object within a wider environment like light, shadows, and the edge between one object and another behind it. With nonlinear software compositing, these atmospheric components of visual perception threaten to dissolve into a world of seemingly discreet and isolated data objects with no inherent relation to one another. As Gibson emphasizes, with human perception the world only ever reveals itself partially: as a person moves through space, some perspectives are revealed, others are concealed.15 In contrast, in the nonlinear, distributed field of motion graphics software, the ambient array has been eliminated in favor of a distributed field of complete visibility. Objects become data sets that are fully present, all of the time.

If graduated perspective can be understood as the core visual strategy for the Cartesian quest to be all-seeing, the software object layer—composed, mapped, moved, and manipulated over space and time through an ever-expanding range of variable data points—makes a good candidate for envisioning how visibility functions within contemporary societies of control. Rather than fix it in relation to a single point, an object is covered over with metrics until little space is left to imagine anything outside the frame.

Ambient videos like Swimming in Qualia respond to this situation by developing specific multiplanar strategies to open out a shallow space for the unknown within these overcoded contexts. I focus on two strategies here. First, soft fascinations give viewers the freedom to determine what degree of attention they will give to a work while implying much more lies beyond. Second, contingencies are introduced into the vacuum of software space via footage of found movement in lived space.

Soft Fascinations

Building on the work of William James as well as contemporary neurobiology, environmental psychologists Stephen and Rachel Kaplan describe two forms of human attention used in everyday life: voluntary and involuntary. A person calls on voluntary, or directed, attention when attempting to focus on something by inhibiting awareness of other parts of the environment of equal or greater interest. This inhibitory mechanism is susceptible to fatigue from overuse. Attentional fatigue is common in urban environments, which often present exhaustive demands on directed attention. Crossing a busy street safely, for example, demands attending to many different objects moving in many different directions while inhibiting attention to many others.

In contrast, involuntary attention does not fatigue. It occurs when attending to objects of inherent fascination in the absence of competing stimuli. The Kaplans describe two varieties of involuntary attention: hard fascination, which arrests a person’s entire concentration, and soft fascination, which supports attention but does not demand it. The Kaplans’ “attention restoration theory” proposes soft fascinations serve to relieve attentional fatigue by resting the inhibitory function as well as providing a reflective space for attention to wander freely.16

The Kaplans identify several specific aesthetic qualities inherent to environments of soft fascination, including being away (the feeling of being outside everyday space and time), extent (the feeling the environment extends beyond the horizon of perception), and mystery (the promise of further intrigue). The Kaplans’ clinical research demonstrates how exposure to this combination of environmental qualities effectively contributes to the restoration of directed attention, with benefits in both convalescent settings and the everyday attentional demands of urban life.17

The Kaplans’ theory provides important clues for understanding how ambient media are able to support various levels of attention while generating restorative moods. First, they argue familiar aesthetic criteria like otherworldliness and mystery are integral to the experience of soft fascination. This implies when standard forms of BGM and BGV “strip away all sense of doubt and uncertainty” (to paraphrase Eno), they not only eliminate their interest as works of art but also undermine their ability to provide restorative moods as “ignorable as they are interesting.” In other words, the presence of uncertainty in an ambient work is not incidental to its ability to provide attention restoration but integral to it. Feelings of being away, intimations of a world beyond the perceptible, and a sense of mystery are all inherent properties of ambient “healing.” When a work of BGM or BGV fails to maintain this mystery, it also fails as an effective source of attention restoration.

Second, the Kaplans’ emphasize the contingent and unpredictable aesthetic details of the nonhuman world are often the best providers of soft fascination, leading to their emphasis on “nearby nature” (parks, trees, and other green spaces) as a solution to urban attentional fatigue.18 This helps explain why phenomena still not fully subject to human control so often feature in ambient works: the movement of clouds, the rippling of water, the patterning of plants, the voices and gestures of nonhuman animals. What makes these materials effective tools of ambient subjectivation is not the fantasy of “nature” as a pristine place outside human civilization (a clichéd image found in many more-generic works of BGV) but how these objects draw viewers away from their human identities and hint at a less familiar world of indeterminate borders. By cultivating perceptual uncertainty, soft fascinations point to a world of effects and influences beyond what is immediately apparent to human viewers. This confluence between attention restoration and aesthetic interest is key to understanding how ambient media reconfigures earlier aesthetic styles as practical resources for self-care.

Swimming in Qualia cultivates soft fascinations through the permeability of each of its constituent parts. Transparency, softness, drifting fog, rippling reflections, and gentle lighting make permeable the boundaries between the different objects sliding around the screen. Ise draws on software-based image compositing only to push against its object/layer ontology, rendering a world of porous surfaces and blurred borders. The bodies of the viewers can then mirror this amorphousness as they attune to the surfaces of the work.

In the “Ascent” portion’s second sequence, overlapping black silhouettes of humans’ upper torsos appear as if gathered to watch something deeper in the image, gently lit by a turquoise background. This assembly of still and slowly moving bodies is arranged in superimposed layers of differing sizes with varying degrees of transparency. Some bodies are semipermeated by the turquoise light. The figures shrink in size with each receding layer, providing a perspectival illusion of shallow depth against a background color field while retaining a sense of flatness (Figure 15).

The figures’ sliding is situated in an uncanny valley between human and computer-generated movement, convincing in general contour but in its particulars always a little too smooth.19 This also holds for the dolphin swimming through the blue-green background late in the sequence, with its slightly unrealistic tail thrust. Just as Ise suffuses these semitransparent humans with an impossibly pure light, she mixes photo-realistic and computer-generated movement beyond the point of visible discernment, compositing a shallow space of indiscernible volume and substance.

In both museum and concert versions of the piece, the screen architecture of the work further undermines any clean distinction between its constituent parts. In the museum identical source material is recast in continuously varying combinations on the two adjacent screens. The two loops’ disparate lengths result in a continuous shifting of the relationship between the two screens, scrambling attempts to discern an overarching narrative or a firm beginning and end. The perpendicular angle between the two projections forces museumgoers to either approach both walls from an oblique perspective or favor one in a frontal approach and relegate the other to peripheral vision. In either case, as curator Ishida Tetsurō points out in the STILL/ALIVE exhibition catalog, it is impossible to focus fully on both screens at once.20 By eliminating the possibility for complete frontal attention, the screen architecture of Swimming in Qualia blurs the border between foreground and background forms of awareness.

The concert version cultivates a different form of perceptual ambiguity. A mise en abyme emerges in the visual echo between the live space of the concert hall (with its audience seated in rows facing musicians arranged in front of a large screen) and the layered silhouettes visible within the screen during the second section of “Ascent,” assembled in rows of their own. With the sudden appearance of the dolphin in the deepest layer of the image, the glowing turquoise backdrop is reframed as a tank filled with water, and the silhouettes become legible as an audience in the darkened space of an aquarium. The tank’s light wraps the dark frames of the silhouettes in front of it, much as the live theatrical audience watching this scene see those seated in the rows before them bathed in the light of the video’s own projected glow.

Steve Jansen’s ambient score provides further resources for ambient subjectivation via soft fascination. Reverberant sliding synthesizer motifs slow in speed as they trickle like birdcalls from the speakers across the length of the piece. Through these slowing gestures, the music provides a sensation of continual deceleration even as it flows steadily onward. As with Hosono’s “Fossil of Flame” (described in chapter 2), these slowing cadences provide a sonic framework for the audiences’ own gradual physical relaxation.

In the theatrical version of Swimming in Qualia, this drift toward stillness culminates in an extended section near the middle where the musicians put down their instruments and the screen fades to an indiscernible darkness. Slowly eased into deeper immobilities, the audience sinks into a collective rest feeling more like a fall back into sleep than a break in the flow of the piece.

When the musicians resume playing and the screen begins to glow again, the audience returns from this interstitial slumber with attention refreshed and even less able to discern where the piece ends and their own reflective drift begins. Swimming in Qualia aims for this blurred space between dissolution and drift, implying an unknown beyond while remaining unthreatening enough for attention to relax and recover. Characteristic of ambient video, the work emphasizes the uncertainties of the image but ensures this uncertainty never threatens to overwhelm. By providing immersive spaces of contained and “safe” complexity—another form of shallow depth—ambient video serves as a soft refuge for sensory reintegration.

This provision of a sensory refuge is enhanced when an ambient video is ensconced within an art gallery, museum, or theatrical space. Often located in densely populated urban environments, these contained spaces are nonetheless protected from outside noise and immediate demands for focused attention. The mostly bare walls serve as a way not only to better absorb the media on display but also to block out the more cluttered and demanding environment waiting outside. In the case of projected video, lowered lighting and covered windows further deepen the gallery’s role as an intermediary space between the private and the public, the inward and the outward. But unlike the darkness of a movie theater, the gallery most often calls upon visitors to remain engaged in organizing their moment-to-moment movement, deciding where to position their bodies in relation to the screen and to other visitors. As noted, audiences can try out different angles of perception, deciding how close and how long to linger. In this, the gallery serves as an ideal space for a more autonomous encounter with soft fascinations, allowing visitors to approach them as they see fit (literalizing the sculptural quality of Satie’s Gymnopedies, discussed in chapter 1).

As presented at the Tokyo Metropolitan Museum of Photography, Swimming in Qualia is structured through nested pools of shallow depth. The perpendicular screens draw audiences closer to the image than is typical for such a large format projection, luring the mobile viewers into an unresolved search for a stable perspective. At the same time, the visible spaces within the projected images are, with few exceptions, focused on layered surface textures continually slipping between two- and three-dimensional space. Petals fold on top of one another; silhouettes slide in front of other silhouettes; trees emerge here and there in the middle distance. One seeming exception is the sequence focusing on the railroad tracks, in which an overhead shot shows two parallel rails zipping by horizontally at high speeds, with no overlays. Here, however, the image is presented in triplicate, with slightly different tracks side by side, pulling the spatial familiarity of the original three-dimensional shot back into the two-dimensional arrangement of a triptych.21 Ambiguous shadows occasionally pass across one of the inset windows, and rain falls on the “surface” of a transparent foreground layer, further flattening the overall texture. A similar two-dimensional splitting of the three-dimensional image occurs in a later lotus flower sequence. Here, the flowers and leaves break up into a series of vertical bands that slip and congeal as they slide laterally across the surface of the screen (Figure 16).

Swimming in Qualia constantly plays with this tension between two-dimensional graphic surfaces and a shallow three-dimensional depth of field. This latter space is itself established not through linear perspective (lines moving from near to far, from viewer into a gradually shrinking distance) but through accumulated surfaces occluding the deeper layers of the image.

Movement in Swimming in Qualia is mainly movement across rather than between these two-dimensional layers. The few dives into the image consist of software zooms rather than three-dimensional camera movement. In a zoom into a forest scene, for example, the size of the trees grow larger, giving the impression of moving forward in space, yet the perspectival relations between the different parts of the image do not change. What resembles movement in depth at the same time registers to perception as simply the enlargement of a flat, two-dimensional surface.

Feelings of drift in Swimming in Qualia are largely the result of this software-based interplay between perceptual depth and the surface of the projected image. The movements of objects within the frame have no clear beginning nor end and lack a visible instigator within the image. The different layers appear to be moved by an unknown hand, an absent creator, an invisible software instruction lurking unseen in the background of the image. Individual scenes fade out before the moving parts can settle into any static formation. The sheer number of moving surfaces and their variable speed and direction make it difficult to see everything as a gestalt. Just as the perpendicular angle of the two projections resists a singular focus, so does the indeterminate relation of the nested frames within each screen. Repeatedly displaced from any stable perspective, attention is left to drift among the qualia.

Swimming in Contingency

But even with all of these manipulations of transparency and depth, compositing software still presents a challenge to soft fascination: how can qualities like mystery and extent (the feeling of something beyond the horizon of perception) be achieved within the airtight vacuum of overcoded object layers? Visual effects notwithstanding, contingency has been evacuated in exchange for the precise mapping of each object’s properties and movements at every moment of the video’s running time.

Ambient video’s answer is to undermine the modular logic of the compositing software by seeking out ways to sneak in the material uncertainties of soft fascination from the outside world. Rather than take the software-coded object as a discreet and isolated set of parameters, ambient video artists leverage environmentally inflected objects discovered in existing environments as a means to introduce soft fascinations back into the vacuum of composited space. In this way, these objects can maintain a sense of mystery and depth, pointing viewers beyond the space of the software itself and back into the perceptual uncertainties of their original ambient optic array.

Before giving a few examples of how this functions in Swimming in Qualia and other recent works of ambient video, it is worth noting this incorporation of external contingencies into controlled environments has an extensive history. The Sakuteiki (Records of garden making) is the world’s oldest known text on the aesthetics of gardening. Written by Tachibana no Toshitsuna (1028–94) sometime in mid- to late eleventh-century Japan, the work focuses on the selecting and placing of stones as the primary act of garden formation. The author advises the would-be gardener to “follow the request of the stone” (ishi no kowan ni shitagahite) in choosing where to place it. In Toshitsuna’s time a stone’s shape was thought to express its own intention and dictate its proper use and position in the garden. For example, taking an upright stone from a mountain and placing it upside down in a garden was considered taboo and thought to lead to the death of the garden’s owner. If a stone had landed upside down after tumbling off a mountain and enough time had passed for it to weather and settle in its new position, however, the stone could be used in this new orientation without the threat of evil spirits: “This weathering is not the work of man. Because the stones have weathered naturally, they can be set or laid in the garden as they are found in nature without impediment.” Over nine hundred years later, worries over evil spirits may have faded, but the visual fascination of this contingent “weathering” has not. Ambient video works toward a similar incorporation of the contingent texture of found objects into the highly controlled space of the digital “garden.”22

The interplay between the contingency of found objects and their subsequent framing is also a well-established theme in more recent camera-based arts. Ise is also a photographer, and despite her reliance on compositing, her video work in many ways shares more with the techniques of street photography than it does with feature filmmaking and other forms of prescripted moving-image media.23

Whereas a still photographer might be searching for found compositions frozen in time, video calls for a search not only for compelling images but interesting movement patterns as well. There is an abiding interest in Ise’s work in found movement, in unscripted physical events encountered in a public landscape and surreptitiously recorded as they occur in front of the camera. As noted in chapter 3, Ise’s pieces often begin with her wandering around local environments looking for interesting visual rhythms. Her videos are then built around this found material, whether from the city streets, the zoo, the park, the forest, or the sky. By letting these found movements guide her later compositing choices, Ise uses the software to reach back toward the contingencies of the space before her camera.

The use of found 3D movement to guide composited 2D effects was also popularized by video artist and composer Takagi Masakatsu (1979–), who began developing these techniques after disbanding his initial live video project with Aoki Takamasa, Silicom. Takagi’s Journal for People series of video works (2001–2) uses an array of heavily effected imagery of figures sliding smoothly through space: ice skating, swimming, jumping, and running, among other forms of movement.24 The most provocative combination of 2D and 3D in the collection does not feature human movement, however, but the circular rotation of a carnival ride. Light Park #2 begins with medium and close-up shots of a rotating swing ride, the kind where individual swings hang by chains from a large carousel shaped like a spinning top. As the carousel picks up speed and lifts off the ground, the swings (and the riders they carry) are pulled centrifugally outward. The video presents images of the rotating swings in high contrast, with the seats, chains, and riders presented in glowing and slightly ghosted white silhouettes against a background of solid black (Figure 17). The removal of nearly all visual texture from the image, save for the uniform gray blurring at the border between the two colors, simultaneously flattens the composition and reveals the fundamentals of the 3D rotational movement with all the more clarity.

Video artist Kawamura Yuki (1979–) explores similar overlaid rotations in a series of video works from 2005.25 Slide and Port both draw upon footage of objects rotating on an axis tilted away from the vertical surface of the image. Slide eliminates most visible details from the original video, a medium shot of a revolving door. The only remaining elements are the gray outlines of light reflected off the spinning glass, rotating at various speeds through empty white space. Near the middle of the piece, the pace shifts to introduce a strobe effect, with individual figures walking toward and away from the camera. The figures themselves are turned into washed-out fields of color, overlaid one on the next with various layers of transparency, while over each figure various fragments of refracted light (akin to lens flares) float around like particles in the air. Port begins with a similar shot of a revolving door, but here the framing focuses on one section of the threshold, with the spinning axis off to the right. The figures and the outlines of the door itself are again drained of detail, the passersby visible only as shadows reflected within the spinning glass. Depth is obscured to the degree it is difficult to discern which direction the door is spinning. The edge of the door itself appears to be sliding left and right in 2D space, despite the visual cues signaling rotation. As with Takagi’s work, orbital 3D movements captured by a camera guide the postproduction 2D effects, but in ways that leave much of the original scene ambiguous. Jour de reve brings in a similar carnival theme, this time focusing on a spinning carousel heavily overlaid with layers of color and translucent water droplets (Figure 18). Again, the original footage is heavily obscured, allowing the rotation of the carousel to determine the vectors of the swirling colors and shapes overlaid upon it.

While opening up greater degrees of porousness between 2D and 3D space, these four works simply replace the mechanics of the software with the geometrical mechanics of an object in front of the camera. A different approach to this same problem is to bring in movement less predictable than the rotation of an object on a stable axis. Takagi’s work often focuses on the undisciplined and lively physicality of young children. Works such as Rama (2002) and Aura (2003) organize a complex series of 2D color and pattern layers around footage of running, playing, skipping, and tumbling kids as they move toward, away from, and around his camera.26

Animal movement, particularly bird movement, emerges as another prominent source of spontaneous motion. Takagi’s Birdland (2001–2) plays with the partial revealing and erasing of images of crows on electrical wires and in trees. As in Light Park #2, image detail is reduced to solid color fields—this time, black silhouettes against a white background. The compositing intervention in this piece centers on the active drawing and erasing of the original image (lines emerging and branching off from each other, growing rhizomatically into new shapes and connections, only to disappear again). Amid this 2D mutability, attention falls to the occasional glimpse of the 3D movement of the crows themselves—a turning head or a leap into flight.

Ise also has a range of pieces organized around the contingencies of animal behavior. In Number of Blinks (Mabataki no kazu, 2003), the eyelid movements of a primate shot in close-up and presented in heavily saturated pinks determine the editing of the piece itself.27 As noted in the previous chapter, a central section of +Intersection (2002) features overlaid images of semitransparent flamingos wading around a shallow pond. In these videos, as with the aforementioned examples, the blurring of color and the shifting of semitransparent layers mesh the animals’ movements through 3D space with the sliding of their silhouettes across the surface of the image.

Over time ambient video artists began developing more subtle approaches to 2D/3D porousness not dependent on such dramatic manipulations across the entire canvas of the image. This shift to exploring more localized transformations within the visual frame was aided by technological changes in image delivery. In The Language of New Media (2002) Manovich predicted larger screens and higher resolutions would generate an aesthetic of “spatial montage,” where only parts of an image move at any given time, allowing attention to slowly shift to different parts of the frame. As Jim Bizzocchi points out, this is exactly what has happened with the proliferation of large high-definition televisions in the past decade.28 The flatter screens blend more easily into their surroundings, while the higher resolution has allowed video to begin to rival painting or still photography in visual detail, reducing the need for constant movement and variety to hold viewers’ attention. Rather than relying on editing and narrative to maintain interest, high-definition video could focus on a more gradual shifting of image layers.

Along with Swimming in Qualia, Ise explores this new approach in a series of four looped pieces entitled Passage (2008).29 The final loop, Noema, consists of a single image of the upper part of a nearly leafless tree shot against a dull, gray background. Large and small branches fork out to fill up the entirety of the frame, with dozens of black birds perched upon them. In stark contrast with the other works I have described, Ise’s software-based manipulations of this image are simple and restrained. Every time a bird flies off from the tree or returns from out of the frame to alight on a branch, a thin, red line appears in the bird’s wake, tracing a path through space on the surface of the image (Figure 19). These lines persist even after the birds have landed or departed, so the screen gradually begins to fill up with a tangle of red-threaded flight patterns.

Noema makes for a stark contrast to Takagi’s Birdland, in which bird movement emerges only in brief glimpses while the image layer undergoes continual and forceful manipulation by the virtual brush of the software. In Noema, compositing choices are led entirely by the unscripted movement of the birds. After determining the basic structure, Ise allows the contingencies of the found movement to determine the ultimate form of the graphic 2D surface.

While the distributed field of compositing software can be understood to embody the larger biopolitical drive for environmental surveillance and control, ambient video seeks to mobilize the attentive freedom of soft fascination and the contingencies of found movement to carve out shallow spaces of autonomy within this coded world. This ambiguous fusion of analog contingency and digital control can be understood as the visual equivalent of the return to acoustic instrumentation in Hatakeyama’s Minima Moralia and other recent ambient music discussed in chapter 2. Instead of bringing objects fully into the malleable world of digital data, ambient media seek ways to open the enclosed virtual space of the computer to the contingencies of the wider (as yet uncoded) environments outside.

Time spent swimming in these shallow depths has real effects on viewers, slowing the breath, allowing for attention restoration, and opening up more flexible ways of relating to moving-image media. At the same time, this emphasis on soft fascinations and spatial contingency serves to compensate for just how shallow this constricted space of subjective autonomy really is. Ambient video can imagine a self free to float with the water and drift with the wind, yet the atmospheres on offer are as tightly controlled as any have ever been.

What does it mean, ultimately, to be subject to such carefully designed atmospheres? In order to pursue this question further, the final two chapters return to the social and historical context of ambient subjectivation, focusing on how ambient media force us to rethink the ethics of emotion regulation (chapter 5) and confront the cultural politics of “therapy culture” more broadly (chapter 6). In the process we will look at how ambience functions within the more narrative contexts of the feature film and the novel.