Agent given a vocal-synth repo builds branching acoustic-tube singing synthesis
TinfoilTricorn · x · 2026-11-10
Developer jamesonhodge showed off a vocal synth that models the vocal tract as a network of branching acoustic tubes, instead of stitching and pitch-shifting phoneme recordings like Vocaloid. He's tuning it against recordings of real singers to improve naturalness — and is already impressed by how cool it sounds.
In a related repost, gavinpurcell handed the repo to an advanced model (Fable 5.2 via his Claude agent) and let it run free, marveling at the creative choices frontier models now make on their own.
More from coding & agent
- He Vibe-Coded a Calorie Tracker to Avoid Paying: 'Most Apps Have No Moat' — weswinder · 2026-11-10
- From Frameworks to Full Agent Platforms: Who Wins the Agent Platform Market — prasannaalahoti · 2026-11-10
- Hermes Agent Matches Claude Code on Benchmarks but Costs More per Task — EXM7777 · 2026-11-10
- Invest in engineering, not agent factories: Geoffrey Huntley says Nix and fast builds get uplifted by model drops — sloppenheimer · 2026-11-10
- Coding tip: tell the new model "This codebase is yours" to make it bolder, plus always read the model card — sloppenheimer · 2026-11-10
- Local Qwen judges stream clips for free: a zero-token auto-clipping pipeline — TangeloOk9486 · 2026-11-10