Transcribe a song and clone a voice from singing
When the sound is singing over music, the app first separates the voice from the music, then transcribes it with song settings, so whole lines are no longer skipped.
On this page
Music under a voice can make speech recognition skip whole lines of a song. VLPremiere Pro now handles songs on their own path.
Transcribing a song
- When the sound is singing over music, the app first separates the voice from the music with macOS voice isolation. Every word keeps its time.
- Then it transcribes with song settings: no part of the song is skipped, and an unclear pass is tried again.
- You do not need to choose anything: transcribe as usual. Speech without music is transcribed exactly as before.
Cloning a voice from a song
When you add a song recording as a voice, the clone is made from the separated singing voice, not from the band.
One step for you
The song settings live in the voice engine. When the Voiceover panel offers Update for the engine, press it once. Until then, the voice separation already helps.
Still need a hand?
Open a ticket and we'll reply in your account.