Clean up a voice recording
Denoise a voice-only recording and bring it to a steady level, ready to publish or send. No upload, no account.
Runs on your device. Your file never leaves it.
Preset for this page
- Strength
- Voice
- Normalize
- Yes
The tool below starts with these settings. Change any of them in the tool.
How to
- Add the recording. Drop a voice file such as a podcast take, an interview or a voice memo onto the page.
- Keep the voice strength. Leave the strength on voice so the clean-up targets speech and leaves the tone of the voice intact.
- Leave levelling on. Keep normalize switched on so that the cleaned recording ends up at a steady, listenable volume.
- Listen to the pauses. Play the gaps between words; they should be quiet, with no watery or warbling artefacts, then save the file.
A speech-first clean-up for podcasts and voice work
Voice recordings need different treatment from video soundtracks. The goal is not to preserve ambience but to make one person's speech clear: remove hiss from a cheap microphone, reduce the hum of a computer fan and tame the room tone between sentences. This page uses the voice strength and switches on normalization, so one pass produces a recording that is both quieter in the background and even in volume.
That suits podcast takes, remote interviews, audiobook drafts, lecture recordings and voice-over for videos. It does not remove echo from a hard-walled room, nor a second person talking in the background, since those are not steady noise. For music with voices, the vocal remover is the right tool instead. If the result sounds over-processed, go back and use the light setting on the remove noise page.