|Landscape Temporal Variations
Manipulating Landscape Time
|View Flash videos|
Landscapists have had to select a given moment in time in which to represent a landscape. They may have employed elements of lighting and atmosphere from different times or days but a painting or photograph of a landscape irrevocably implies the selection of a particular, unique moment in time.
The advent of film and video cameras freed the artist from the constraint of representing the landscape at a single moment in time. Moving image recording allows the presentation of a landscape over a range of time, whether direct cinematic "real time" as in Warhol's "Empire" or in Wolfgang Staehle's work where discreet samples of time are presented sequentially. Time lapse cinematogrphy has become commonplace as a means to temporally compress time, to allow the perception of otherwise glacial phenomena.
I first become interested in compressing landscape time as a means to visualize proposed land art pieces which would evolve very slowly over time. These pieces would generate clouds in geometric patterns but those patterns might only be discernable over durations of hours. I began to do time-lapse studies of clouds and the landscape, and became fascinated with this process of temporal modification.
A View of the East River from Lower Manhattan
still from HD video (detail)
All evidence of human activity has been removed from a midday scene of New York City
I then began to wonder about performing other manipulations on landscape time. I drew on my experience as a computer software engineer to select various digital image processing techniques which were appropriate to my intent as an artist.
One aspect of temporal compression which I found undesirable is the fact that the very rapid motions in a scene become frenetic and stacatto-like. A well-known example is the film "Koyaanisqatsi - Life Out of Balance" where this attribute of time-lapse cinematography is used to convey and reinforce the message of an out of control society. In temporal compression, rapid movement such as from human activity, or even the movement of vegetation in the wind, becomes discontinuous. In engineering terms is it "undersampled", leading to "aliasing". Objects appear to jump from place to place instead of moving smoothly.
I sought a way to remove these rapid motions within the landscape, leaving only the slower, more gradual environmental changes that I was interested in. I decided to employ the technique of averaging frames, literally adding or superimposing adjacent frames, as many as several hundred. Elements which change rapidly, on a frame by frame basis, are "averaged out", blurred into indistinction, while more slowly changing processes remain. Curiously, this is a digital image processing technique which is used to improve degraded images, such as those from distant space exploration vehicles, by removing noise or "snow". In this case, it is being used as a temporal filter, to remove elements in the scene which change frequently. The result can be erie and disquieting. A busy citiscape can be made suddenly devoid of any evidence of human occupation. Cloud forms and vegetation become strangely modeled. But most importantly, the slower invironmental processes which had been below our limits of perception, are now dominant.
A View of the Concourse, Grand Central Station
still from HD video (detail)
Here the depiction of human activity is proportional to the lack of motion.
The Relationship to Photography
These works, while often distributed or exhibited using the medium of HD video, have a closer relationship to photography than to video. They are more "moving photographs" then video. They are recorded using a still camera not a video camera. The image sequences are recorded at very high digital camera resolution. This is a much higher resolution than even High Definition Video and they must be down-rezzed to be presented in that format. They are not a video clip but rather a sequence of still photographs. They are presented in a rapid succession like video or film, but the absence of a sound track reinforces their photographic nature. They are trans-media and lie in an indefinite region between still photography and video.
The fact that time is manipulated within the sequence and various complex forms of time averaging and temporal filtering uccur further removes them from the medium of video and places them closer to photography. A selection process has occurred on the time-dependent elements of the original scene and unlike video or film, not all of that temporal content is presented. One can consider photography as a rarified subset of film or video, one in which only a single frame is presented. (Think of a photographer using an SLR with a high speed motor drive or high speed sequential shooting with a digital camera and then selecting only one frame). This work forms another such subset, where only certain temporal frequencies are presented and others are filtered out. Many frames are presented, but not all of the temporal events which occurr in the original sequence are present. A form of temporal editing has occurred, similar to the selection of a single moment in time in a photograph. However in this case an ensemble of temporal frequencies has been selected comprising only a subset of the temporal frequencies in the original scene.
A series of digital camera images are taken at intervals of a few seconds throughout the course of a day. The camera and computer which controls it is usually left unattended for a period of days or even weeks, continuously recording images. A single day is selected from the sequence and encoded as a video and evaluated. When these images are presented at video frame rate they compress the course of the day into a period of a few minutes. Decisions are made as to what content in the sequence is to be manipulated, enhanced or suppressed, and multiple computer image processing passes are performed. Often, different elements in the scene are composited from different processing passes, so the clouds in the sky might be simply time compressed, while the ground is time averaged at one rate and selective elements such as a river are averaged at another. In this way several different temporal filters are applied to the same scene simultaneously. The processing is computationally intensive, and processing of a few minutes of material can consume days of computer time on the fastest desktop computers. The final images are then encoded as an HD video. The technique occupies a convergent domain between photography, digital imaging, and video. The final HD video is intended for presentation on a large format flat panel display, referencing landscape painting. The ultimate presentation format is intended to be the new generation of ultra-high resolution digital cinema projectors, intended to replace 35mm film projectors in theatres. As of this writing these projectors are very expensive and access is extremely limited. As they begin to be produced in volume and enter distribution it is hoped to be able exhibit this material using this preferred technology.
New York, New York