Silence Remover
Find and cut the silent gaps from a recording, with a threshold you control.
Cut The Dead Air, Keep The Pauses
Any recording made in one take is full of gaps: the pause before you started talking, the moment you lost your place, the silence at the end while you reached for the stop button. Removing them by hand is slow and mind-numbing, which is why most people leave them in.
The important distinction is between dead air and a pause. The gaps between words and sentences are what makes speech sound human, and stripping them produces the breathless, over-edited effect that makes a podcast exhausting to listen to. This tool keeps short gaps by design.
Key features
- Threshold in decibels — stated the way every audio tool states it, so −45 dB means the same thing here as elsewhere.
- Short gaps are kept — anything under your minimum stays, because the pauses inside a sentence are not dead air.
- Padding around speech — each kept region is extended outwards so words are not clipped at their edges.
- A visual map — the waveform with the removed sections marked, so you can see the effect before committing.
- Crossfaded joins — 5 ms at each splice, which is what makes the cuts inaudible.
This tool runs entirely inside your browser using native Web APIs. Your files and text are never uploaded to a server, never logged and never shared with third parties.
How to use: Silence Remover
- Choose a recording
MP3, WAV, M4A or OGG, decoded in the page. - Set the threshold
Watch the map: the red areas are what will go. −45 dB suits a quiet room. - Protect the natural pauses
Raise the minimum gap until the pauses inside sentences stop being marked. - Download the trimmed file
A WAV with the gaps removed and every join crossfaded.
Technical specifications
| Processing location | Entirely in your browser — no server round trip |
|---|---|
| Data uploaded | None. Files and text never leave your device |
| Price | Free — no account, no trial, no usage cap |
| Category | Audio |
| Works offline | Yes, once the page has loaded |
| Browser support | Chrome 90+, Edge 90+, Firefox 90+, Safari 15+ |
| Interface languages | English, 中文, हिन्दी, Español, العربية |
Frequently asked questions
What threshold should I use?
Start at −45 dB, look at the map, and move it until the marked areas match the silences you can see in the waveform. A quiet room with good mic technique may need −55 dB; a noisy room or a laptop microphone may need −35 dB, because its noise floor is higher than real silence.
Why does the result sound rushed?
The minimum gap is too short, so natural pauses are being treated as dead air. Raise it to half a second or more. Speech without pauses is technically shorter and much harder to listen to.
Why are the first words of sentences clipped?
Consonants like p, t and k start very quietly and fall below the threshold before the vowel arrives. That is what the padding control is for — it extends each kept region outwards. Try 0.15 seconds or more.
Can I hear where the cuts are?
Not normally. Each join is crossfaded over 5 milliseconds, which removes the click a hard cut produces. What can give a cut away is a change in room tone either side of it, which no crossfade can hide — that is a recording problem rather than an editing one.
Why is the output a WAV rather than an MP3?
Because re-encoding to MP3 would mean a second lossy compression on top of whatever the source already had. A WAV is larger but preserves exactly what came out of the edit, and can be encoded once at the end of your process instead.