What I’ve tried
- Using different tags and lyric phrasing
- Testing multiple genres and languages
- Reviewing the ACE‑Step documentation and UI tabs (Text2Music, Edit, Lyric2Vocal) for accent‑related controls
What I’m looking for
- Is accent control currently supported?
- If not, is it planned as part of voice‑cloning, LoRA fine‑tuning, or future controllability features?
- Are there recommended workarounds (e.g., training a LoRA on region‑specific vocal data)?
Why this matters
Accent control is important for creating regionally appropriate vocals, especially for songwriting, demos, and commercial use cases.