Indie dev builds vocal synth modeling the vocal tract as branching acoustic tubes
TinfoilTricorn · x · 2026-11-11
Developer jamesonhodge demoed a self-built vocal synth that models the vocal tract as a network of branching acoustic tubes — unlike Vocaloid's stitched, pitch-shifted phoneme recordings. He's still tuning naturalness against real singer recordings but is impressed by the results. The clip circulated with an AI-generated take on what Starship's "pez dispenser" would sound like if space weren't a vacuum.
More from Fun
- Economist jokes new QJE Insights journal rejects papers before you even submit — alejandroll10 · 2026-11-11
- Remember when we made images with these? OpenPose skeletons become a nostalgia meme — cocktailpeanut · 2026-11-11
- As Models Get Smarter, Prompts Devolve Into "ok do it now i believe in you" — petergyang · 2026-11-11
- Dev's "hand-written lines of code" counter is a kitten weighing zero — tlakomy · 2026-11-11
- Peter Yang: building a kart racer with Fable all morning — actually a good game, not one-shot slop — petergyang · 2026-11-11
- Fable 5.2 one-shots a photorealistic Central Park night photo recreation with p5.js brushes — yuwen_lu_ · 2026-11-11