Silicon Remembers Carbon (Version 1).
The reflection of the bicycle was drawn into the video mix by the presence of someone walking along the upper edge of the image.

Version 1 (1993-5)

The central element in Silicon Remembers Carbon is a large video image projected down onto a bed of sand on the floor of the installation space. Visitors' movements subtly affect the mixing and dissolving of video images and sounds. Each visitor leaves traces which affect the experience of the work for later visitors. The installation presents a fragile illusion, a consensual hallucination, requiring the visitors' participation for its continuation, through their body movements, a willingness to blur their eyes slightly to hide the scan-lines, and their ability to project depth into the flat image. They are offered a range of possibilities from sustaining the illusion by creating and maintaining distance, to dispelling it by stepping into the illusionary space itself. For the artist, the visitors' movement through this range of possibilities represents a more important interaction than the direct interaction with the technical system itself.

A 4 meter x 3 meters video image is projected from above onto a bed of sand. There are 1/2 meter walkways forming a frame around the image. The interior of the enclosing space is completely black. A speaker hangs in each of the four corners above head height. Four modified infra-red sensors register the amount of movement occuring along the four sides of the image. Two cameras look across the image from opposite corners from floor level, registering movement only if someone in the installation steps onto the sand and begins reflecting the light of the projected image.

When one enters the empty installation, a single video clip is playing out across the sand. Most likely the image is of water (beach waves, water under a bridge in Toronto, passers-by reflected in the canals of Hamburg, shadows and reflections in street puddles...) The image creates the illusion of depth, the sand sinks underneath it. It is not clear that the 'screen' is sand. People look around then reach surreptitiously down through the depth of the image to the dry sand.

Movement along a walkway is sensed by one of the infra-red sensors which sends an analog voltage roughly representative of the amount of movement. This signal causes a second image from the other disk to be dissolved in along the side where the movement took place. The second video clip is selected by a program from the possible clips on the disk based on the side of the frame where the current movement began.

The new image usually contains shadows or reflections of people along the edge of the clip that is visible. One tends to interpret those reflections and shadows as the being generated by people actually in the room, either oneselves or others, rather than as being present in the image itself. So this installation is some sort of fake reflecting pool, an inversion of Narcissus's experience. Whereas Narcissus's tragedy is that he cannot recognize himself in his reflection, the visitors to the space would find themselves identifying with shadows and distorted reflections that had only circumstantial relation to them. The identification is be momentary, and elusive. My intention is to play along the boundary of identification.

Version 2 (2000)
 
 

'Live Virtual Shadow' in Version 2, Lowry Centre.

In the second version, instead of laser-discs, the video source is made up from 2 streams of MPEG-2 compressed video, resident in RAM on a computer. The live mix between the two video sources is defined by a third video stream composed in real-time by a second computer. This mix-defining stream is made up in various ways of cross-frame dissolves (as in the first version) and silhouettes. The silhouettes are derived from the images of 4 video cameras positioned one on each side of the projection as though looking over the shoulders of people standing around the image. The sand/screen is lit with infra-red light by 4 infrared illuminators and the cameras have infrared-pass filters. Thus the sand is seen by the cameras a brightly lit back-drop against which people's shadows are dark. Instead of casting shadows of darkness, people cast shadows of video, cutting the hidden layer of video into the visible one.

Version 2: The shadow of the person walking by is smearing in a second image, in which a recorded shadow appears. (the second image is an image of water... making it seem as if the pavement has been moistened by the shadow and therefore become reflective...)

At any given moment, there is one video stream that is termed the 'ground' image. In the absence of people in the space, this image will fill the screen. The presence or movement of people in the space, near the image will cause the generation of silhouette mixes or cross-frame dissolves. The nature of the video clip to be mixed in is determined by the side of the image experiencing the most activity. From each 'ground' clip, there are separate mixing possibilities for each of the 4 sides of the 'ground' clip. For each mixing possibility, there is also a mixing 'mode' (i.e. cross-frame dissolve, solid silhouette, smeared motion silhouette).

This new version plays three layers of shadows against each other.

'Real' shadows cast by the bodies of visitors to the space when they interrupt the video projection.

'Recorded' shadows and reflections in the video footage.

'Live Virtual' shadows of video cast by the bodies of  visitors.

The result is an ambiguous play of shadow relationships, where identification appears and disappears...
(Next)

1995
"Press Enter", the Power Plant Art Gallery, Toronto, Canada.
"ArsLab 2", Torino, Italy.

2000
permanent installation in "ArtWorks", the Lowry Centre, Salford Quays (Manchester), UK.

Copyright 2000 David Rokeby / very nervous systems / All rights reserved. 12/11/00