About the Acapella Extractor
The Acapella Extractor uses the Demucs AI model to remove drums, bass and every other instrument, leaving only the voice. The result keeps the original timing, so it lines up with the song for remixes.
Everything runs on your device. The model downloads once, and after that you can isolate vocals from as many songs as you like.
How to use the Acapella Extractor
- Add a song.
- Click Isolate vocals. The first run downloads the AI model (about 40 MB) once.
- Listen to the vocals and compare them with the original.
- Export the acapella as MP3, WAV or FLAC.
Frequently asked questions
What is an acapella?
A track with only the vocals and no instruments. Producers use them for remixes and mashups, and singers use them to study a performance.
Will some music remain?
Usually very little. Reverb on the voice stays, and in dense mixes you may hear faint traces of instruments.
Which format is best for remixing?
WAV or FLAC keep full quality for editing. MP3 at 320 kbps is fine for listening.