Algorithmic Composition

As an algorithmic composer, my work has used machine learning techniques, developed new strategies for creating prolation canons, and derived material from my original mp3 processing algorithms. An example of this work is my composition Heard which appears on Ravello Records. It uses an original automatic transcription program implemented in Pure Data to create a machine listening transcription of a field recording. This became the foundation for an entire series of works exploring different ways of listening, including transcription by ear, live auditory scores, interactive electronics, and machine listening. The original score can be viewed on Ravello Records website and the recording is available on Spotify.


My article on using Probabilistic Latent Component Analysis to derive polyphonic musical structures from field recordings is available in the .works section of my website.

My original algorithmic prolation canon Carrauntoohil is below. A journal article on this work and related prolation canons in Larry Polansky’s music is in preparation.





My chamber music also involves algorithmic processes. I am very interested in generative algorithms for performers–simple, score-based instructions that can generate complex musical results. An example, my guitar quartet Orbits is below, from a performance by the Ukranian Kharkov Guitar Quartet.