Live Coding With Crowdsourced Sounds And A Virtual Agent Companion
summary
Live coding uses machine learning and music‑information‑retrieval to pull crowdsourced samples from Freesound.org into a constrained improvisation: four dice generate random IDs, a trained filter predicts good vs bad sounds and sometimes blocks retrieval creating a deliberate paradox, and a similarity step fetches alternatives when available. A virtual companion agent negotiates choices alongside the human coder, and every downloaded clip must be actively transformed to become part of the set, foregrounding legibility, agency and negotiability.
AI-written summary · about
chapters (4)
- 0:00Introduction
- 17:37Tool development process
- 19:01Randomness and sound filtering
- 20:05Improvisation and manipulation
related talks
17mLive Coding Performance
47mLive Coding Music Performance
51mThe Power of Live Coding: Humanizing Technology
47mProgramming In Time - Live Coding for Creative Performances
30mHarmonizing Code and Melody: making electronical music in a matter of speaking!
46mLive Coding Music: Can You Code? Become a Composer with WebAudio and WebMIDI!
32mLive Coding Music Performance with Sonic Pi
08mNo merge conflicts