Mute or dim parts of a track

The three jobs people bring here

Nearly everyone arrives with one of three problems. The first is censoring: a surname, an address or a swear word said out loud that needs to vanish before the recording goes anywhere. The second is ducking: background music or hum that should sit lower under someone speaking. The third is one loud moment — a burst of laughter, a cough into the microphone, a lorry passing the window — that jumps out of an otherwise fine recording. All three are the same move here: select the stretch, choose how loud it should be, apply. You never re-record anything and the rest of the file is not touched.

Working in selections

The workflow rewards a little care. Zoom into the waveform until the moment you are after is easy to see — a spoken word is usually a distinct little cluster of peaks — then drag across it. If your ear says the selection starts a fraction late, nudge the time fields; they accept hundredths of a second. Apply, listen back, and if it is not right, undo and try again: every edit is one undo step, so experimenting is cheap. An amber marker stays on the waveform wherever you have worked, which makes it easy to review a file with five or six fixes in it before exporting.

The honest bit about 150% and 200%

Boosting is more constrained than cutting, and it is worth knowing why. The whole-file setting offers 150% and 200%, and a hard limiter stops the result from distorting — but a limiter works by flattening the loudest peaks, so a big boost trades some punch and dynamics for loudness. Boosting also raises everything: a quiet voice comes up, and so does the hiss, the fridge and the room around it. And if the original was already clipped at recording time — flat-topped, crackly peaks — no amount of volume work can restore what the microphone never captured. Boost in moderation, and treat 200% as a rescue setting rather than a default.

In, out, and what pairs well with it

You can feed it audio in any format the browser plays, or a video — its soundtrack is lifted out automatically and the result comes back as an MP3 at 128, 192 or 320 kbps. The free tier covers files up to 1.5 minutes and 20 MB; a day pass or unlock lifts both. Two neighbouring tools finish common jobs: if the file needs a pitch change as well, do the level work here first and then shift it in /voice-pitch-changer — a balanced recording shifts more cleanly. And if you are fixing sound inside a video that should stay a video, /video-mute-tool applies the same dip without separating the audio from the picture.

How it works

Drag on the waveform to select a part (the time fields let you fine-tune to the exact hundredth of a second), pick a volume for it and add it as a zone. Repeat for as many zones as you like — one dimmed to 33%, another fully muted. Every zone edge gets a soft 15-millisecond ramp so the result never clicks. The final file downloads as MP3 with the quality you choose.