MPEcho Optimizes Cover Generation with Phoneme Timing
The ISMIR 2026 paper MPEcho combines melody and phoneme information for controllable cover generation. Commenters noted that directly extracting real phoneme timing from reference audio, rather than predicting it, provides more stable lyrics control.
2026-08-04 ~ 2026-08-04 · 3 related posts
- ISMIR 2026 paper MPEcho adds phoneme conditioning to cut lyric errors in cover song generation — affige_yang · 2026-08-04
- Reviewer argues cover songs should extract real phoneme timing instead of predicting it — affige_yang · 2026-08-04
- Cover-song generation gets more reliable when phoneme timing is extracted — affige_yang · 2026-08-04