I have completed five ambient video works: Rockface, Streaming Video, Long Falls, Winterscape, and Cycle. Each of these is a visual poem, exploring the relationship between form, content and visual flow in the context of the landscape as environment. The visual material includes studies of water snow, ice, sky and peaks that play out across a range of scales - from the wide expanses of the full mountain landscapes to the equally rich visual pleasures of closer shots and finer detail. Compelling as these images are on their own terms, this work is not a paean to mere visual beauty. At its heart, my work is a repudiation of the fundamental logic of film and video construction. The moving image has historically relied on the hard cut as the dominant instrument for the ordering of shots. My work rejects this limitation. I completely abandon the hard cut, along with its reductionist illusion of instant and invisible transition. My style never hides transition - rather it celebrates it. In doing so, the work transcends the borders - both spatial and temporal - that have traditionally bound the moving image. The spatial borders within the frame are broken, as image is first deconstructed into component parts, and then rebuilt within a recombinant visual aesthetic. The temporal borders between the shots are obscured, and the traditional hard cut is abandoned for an ongoing cycle of flow, metamorphosis and transformation. This is an evolution in the fundamental treatment of the moving image. A combination of spatial montage and visual transformation has replaced the standard cinematic conventions of temporal montage. Combined with the natural imagery, the visual flow creates its own momentum. Subject and technique mutually support a dialectic of change and connection. Landscape, detail, and form are visually deconstructed and then fused together within an organic unity of space and time. We have seen that time itself is treated as a plastic material in this work. Just as significantly, the pace of change, though constant, is slow. Images are held on the screen long enough that detail can be examined, texture is revealed, composition can be appreciated, even as it flows and slowly transforms. Because of the treatment of time, space, and image, the piece is as much a video painting, or a living photograph as it is a traditional moving image work. In the spirit of intermedia [5], it traverses media boundaries and merges aesthetic forms and viewer reception practices. This form of ambient video draws on a range of artistic practices: a photographer's eye for landscape, detail, composition, and light; a painter’s sense of color and shape; a filmmaker's concern about time and interval; and a video artist's ability to combine moving imagery into dynamic collage and flow. These ambient works reject the standard cinematic imperative to seize viewer attention, and then to hold it unrelentingly. Instead, like a painting or a photograph, the choice of when and how long to interact with the piece is owned by the viewer. This moving image art doesn’t command, it first seduces, and then rewards our attention. Ambient Video and Sustainable Replay I did see one potential problem for ambient video art that is conceived along the lines outlined above. An ambient piece that is intended to hang in a space of recurrent viewing (the home, an office, a public space) will necessarily be seen many times. This final creative challenge - that visual pleasure sustains over repeated viewings - is indeed a non-trivial problem, as Brian Eno recognized. For my own work, no matter how strong the imagery, and how intricate and aesthetically pleasing the transitions - after a number of viewings the images, the sequencing, and the transitions will all be remembered and anticipated. A certain amount of memory and anticipation will initially add to the viewing pleasure. However at some point the extended repetition of identical sequences may lead to a diminished aesthetic return and a lowered sense of visual pleasure. I decided to explore the potential of computation to address this problem - and to increase the re-playability of my ambient video art. A generative ambient video piece can rely on simple computational operations to continuously vary the sequencing and combinations of both the ambient shots and the visual transitions. Since I incorporated many of the same visuals as my last linear video (Cycle), I titled the new generative work Re:Cycle. The Re:Cycle Engine - a Recombinant Generative Video Presentation System Re:Cycle incorporates a variation on the aesthetic strategies of the earlier works (strong imagery, manipulation of time base, and careful use of video layers and layered transitions). Re:Cycle maintains the first two aesthetic strategies, but explores the development of a recombinant aesthetic that applies random processes to video layers, transitions, and sequences. The commitment to a recombinant video system based on random process is an exercise in generative art. Re:Cycle is driven by a closed generative system, relying completely on two databases for its operation. The first is a database of ambient video shots, the second is a database of video transitions. The twenty video clips in the shots database are all visually strong, at least sixty seconds long, and shot in the same general region (the Canadian Rockies). Many present the cyclical and visually interesting motion of clouds or water. The time frame in several of these shots has been manipulated in order to give the motion even more visual interest. There are four transitions in the transitions database. One of the four is a luminance transition. This transition will use the brightness values within the shot to drive the change from one shot to its successor. The incoming shot will appear first in the brightest sections of the current shot, then in the mid-range brightness areas, and finally in the darkest areas. When the transition is complete, the second shot has replaced the first completely. (Figure 2 below) The other three transitions work in a similar fashion, except they are based on chrominance values, not brightness. There are three chrominance transitions: red, blue and green - corresponding to the video color palette. Each of these starts the transition in the areas of the shot with the highest chroma value in the selected color, and continues the transition down through the range of chroma saturation until the transition from one shot to the next is complete.
Figure 2: Stages of Luminance Transition Figure 3 below shows how the generative system uses these two databases to structure the video presentation. When a shot is on the screen, the system selects a new shot at random and selects a new transition at random. It uses the transition to drive the change from one shot to the next, and then repeats the process indefinitely.
Figure 3: Generative Engine This generative engine has the potential - especially if the database of shots is large enough - to drive an ambient video art work that can run indefinitely, and still provide interesting visuals and transitions. The resultant doubly-randomized video stream will generally not repeat particular shot sequencing with any frequency, and will generally provide a different transition for each change. The recombinant aesthetic will play out both temporally and spatially. The random sequencing will provide temporal recombination, while the interplay of random shot and random transition selection will drive each shot change with a fresh spatial recombination. |
|||
