♪ Music — Vocal synthesis
ACE Studio
An AI singing voice workstation that turns MIDI and lyrics into vocal performances.
Tool
Music / Album
PROTO (Holly Herndon)
Hybrid Showcase consented data Unverified
Holly Herndon's 2019 album made with a human vocal ensemble and Spawn, an AI she built with Mat Dryhurst and Jules LaPlace and trained on their own community's voices.
Why it's hybrid
The AI is treated as an ensemble member trained on consenting voices; Herndon says AI makes up only a small share of the audio.
Unverified: this breakdown is compiled from public sources and has not yet been confirmed by the creator. Corrections and takedown requests are handled within 48 hours via Submit.
Hand vs. AI
| Stage | Human did | AI did | Tool |
|---|---|---|---|
| Composition | Herndon composed and directed the ensemble | None (Spawn framed as a performer, not composer) | — |
| Training data | Herndon, Dryhurst, the ensemble and a public session recorded voices and sounds | Spawn learned from those recordings | Spawn |
| Performance | Vocal ensemble sang most of the material | Spawn reinterpreted and resynthesized voices | Spawn |
| Medium | Music |
|---|---|
| AI role | transform |
| Training data | consented · source |
| Craft | vocals, ensemble, composition |
| Creator | Holly Herndon |
| Last verified |
Related
♪ Music — Stem separation
Demucs
An open-source music source-separation model from Meta research that splits songs into vocals, drums, bass and other stems.
Tool open-source
▲ Games · ♪ Music — Voice & SFX
ElevenLabs
An AI audio platform for text-to-speech voices, voice changing and sound-effect generation.
Tool freemium