-The objects distributed throughout the space according to Perlin noise.
for want of forest is large-scale, telematic audio-visual installation (currently in production) comprised of 40 two-meter acrylic tubes suspended vertically at varying heights across a darkened gallery space. Each object is illuminated internally by a a strip of addressable RGB LEDs, such that each tube becomes, effectively, a very low-resolution 1 x 60 pixel display. Collectively, the installation is conceived as a distributed, 3-dimensional assemblage of 2,400 pixels which the viewer may enter into and observe from various angles and vantage points.
Driving this display is a set of 4 to 6 networked cameras placed throughout the city of Buffalo, focusing upon scenes chosen for their color palette and varying amounts of activity, e.g. the horizon at sunset, a busy intersection, a large stand of trees. As the installation switches between camera feeds, the images are relayed back to the gallery site in real time, the resolution downgraded to 60 x 40, and the pixels mapped accordingly across the 2,400 LEDs of the installation.
Though a one-to-one correlation between the captured image and the colors displayed is retained, the installation is not intended to present a coherent depiction of the video source. Rather, the installation works against this, rendering the content as a slowly evolving, discontinuous color field that changes throughout the day. The audio component of the installation is simarly created, composed of drones derived from FFT analysis of microphones located at each camera node.
Aesthetically, the work operates by subjecting aleatoric information streams — audio/visual content over which there is no control — to a quantization procedure which is entirely predetermined. To loosely combine two historical visual art movements, the work aims to invoke an abstract pointillism, presenting the viewer with an experience that is at once both foreign and familiar.
The demonstration video below shows the capture of a live stream rendered across 40 60-pixel tubes. This test was to prove functionality across the network. Video is streaming over IP from a Raspberry Pi using Motion software. In order for the feeds to be seen on the network, it is necessary to forward the ports of routers at each site. This stream is ported via a VPS in Iceland to simulate network latency. The camera is running at 30 frames per second, capturing a cloudy day in Buffalo. The distribution of the objects, however, is randomize throughout the space as in the render above.