Victor Zheng

Fork is an AI composition tool that interprets PGN files of chess games and generates a live soundtrack in response to the game’s progress.

Fork seeks to generate an immersive soundtrack that reacts with clear correspondence to each player’s immediate interpretation of the game, while avoiding being entirely predictable with regards to gestures produced and obvious input parameters such as pieces being used, moves executed, etc.

To that end, Fork reads the number of attacks generated between pieces as the primary indicator of a game’s progress. If few attacks are generated, the music tends toward the tonic, while more attacks push the music away towards more unstable tonal zones.

Download: https://github.com/vzhyu/fork

Papers:

Zheng, Victor. “Fork: Composing a Live Soundtrack for Chess.” Paper presented at SuperCollider Symposium, Johns Hopkins University, 13–15 March 2025. https://supercollider-2025.github.io/papers/Zheng.pdf.

Zheng, Victor. “Chess and Music: Quantitative and Artistic Interpretation of Information.” In International Conferences Proceedings 2025: Information Systems and e-Society, 470–474. IADIS Press, 2025. https://www.academia.edu/128741447.

Photo credit: Purdue University Department of Music