Diagram
Every audio clip consists of automatically selected layers of previous sound clips and the most recent clips looped based on a simple algorithm. The speed of the clips are varied by the algorithm based on a sequence of generated prime numbers, which makes the loops non-repeatataive when played together.
Ecasound, Sox, Ffmpeg, Mencoder, Mplayer, Maccam & Jack wrapped in Python and Bash script is used for creating the installation. Thanks to Michael Murtaugh and Calum Selkirk for technical support.