This piece consists of a system whereby a recorded piece of music can be manipulated by graphically re-interpreting its own score.
This is accomplished by loading a page or pages into a video buffer, which converts the light and dark areas of the printed page to digital data (a video matrix). The scale and position, as well as the resolution of the images are in full control of the user/performer. The data produced by this matrix is a representation of the displayed image and no data specifically related to the file specifications (header, etc) is included. The output of this floating-point video matrix is then transformed into a stream of audio signal, which in turn are constantly read and re-recorded into an audio buffer. The information stored in this buffer is then used to control individual frequencies in a 512 frequency spectral filter; the value of each sample contained in the constantly changing buffer effects the volume level of one of 512 frequencies. The file being filtered is a recording of a performance of the score being displayed on the screen. In this manner, one form of documentation of a piece (the audio recording of a musician performing the piece) is manipulated by another form of documentation (a digitized version of the printed score) by means of the spectral domain (512 channel spectral filter).

The video example here uses the score of Brian Ferneyhoughʼs Unity Capsule as source material.

Loading more stuff…

Hmm…it looks like things are taking a while to load. Try again?

Loading videos…