Interviews and lessons
Keep an interview soundtrack for listening, notes and transcription.
Start with your own video
Choose your own mediaSeparate vocals and accompaniment from video or audio with local Demucs. Listen and save both tracks without uploads or credits.
For music practice, soundtrack organization and editing. Quality depends on the recording and mix.
Keep an interview soundtrack for listening, notes and transcription.
Start with your own video
Choose your own mediaListen to vocals and accompaniment separately.
Results depend on the source mix
Choose your own mediaContinue editing with independent WAV tracks.
Originals are preserved
Choose your own mediaSeparate vocals and accompaniment from video or audio with local Demucs. Listen and save both tracks without uploads or credits.
Try the toolSave vocals and accompaniment separately.
Extract the video soundtrack before local model separation.
First setup downloads components and weights; processing uses your device.
Prepare your input, choose settings and review the result.
Download the desktop app and install separation components.
Choose a file containing vocals and music and set the output format.
Listen to vocals and accompaniment and save them locally.
Common questions about usage, pricing, files and results.
More questions? Contact usFFmpeg handles extraction and conversion; Demucs performs vocal separation.
Installation needs internet to download components and weights. Processing works offline afterward.
No. Reverb, chorus and complex mixes can leave residual sound. Preview the results.
The model uses local computing power. Duration and hardware affect speed; tasks can be cancelled.
A few related tools for the work that comes next.
Free to use, no sign-in. Your files stay on your device.