|
Video 5 Sanctuarium Program Notes
Sanctuarium was composed using algorithmic congruence procedures; that is, the video and computer music were generated using similar granulation techniques on both the local and macro-dimensional levels. The piece bears a superficial resemblance to the Sky Gods of Tavarong in that each video applies granulation techniques to the same set of 2x2-inch (abstract) glass plates. However, the build-up of the initial, pre-granulation montage differs in each case; the former uses only canonic procedures, while the latter employs multiple zoom-trailing and image-folding devices. Composition of the video imagery: Step I. Controlled flow processes were applied to thirty-one, 2-inch square, glass plates using inks, dyes, and viscous fluids; the resulting images formed abstract patterns bearing a resemblance to stained glass windows. Step II. Pairs of the abstract glass plates were sandwiched together, and then macroscopically photographed in digital form to extract segments of the individual images. Step III. Computer video software was used to sequence the digital stills and render them as a two-layer (cancrizans) montage of moving images lasting three minutes. Step IV. The resulting video was subsequently re-rendered and stored as a stream of single-frame bitmap images. Then, the 5400 images were grouped into separate collections of 4,5,6,7,and 8 contiguous frames, and were distributed according to a numerological schema. Two separate videos were again rendered and concurrently (20%-50) ramp-cross-dissolved over an 8’52” time period, a structure of increasing visual density that parallels the sound granulation directionality and formal shape. All of the sound sources for the music were drawn from audio samples I recorded of Chinese musical instruments - sona, sheng, kuchin, erhu, gongs, cymbals, and other small percussion. Most of the samples were processed (patterning congruent with the video frames) using sound granulation software to create drone-like data streams that were further transformed using digital audio plug-ins such as convolution, phase vocoding, time/pitch transposition, etc. A few of them were processed to preserve melodic or texture-character identities. (These appear as iconic references at several points in the audio score.) Subsequently, the results were sequenced and superimposed using standard audio multi-tracking software. Phil Winsor 2/24/2002
|