Tensed Computer Improviser

The Tensed Computer Improviser (TSC) is an audio-corpus-based computer improviser developed in MAX. It relies on machine learning to sort and label grains of audio according to a variety of sound characteristics; in performance it listens to and analyses the incoming audio signal, reacting with a variety of behaviors informed by what it hears.

The Tensed Computer Improviser is an open-source piece of music-generation software. This software is developed in MAX with integrations of the Flucoma, ml.star, and GO abstractions. Requires MAX (paid version not necessary).