How it works
- Drop an isolated vocal take. The gate works best on a dry vocal stem — MP3, WAV, FLAC, AIFF, OGG or M4A up to 500 MB, decoded locally.
- Check the gate. The threshold is set automatically between the take's background noise and the voice level; adjust it, the minimum silence length, padding around phrases, fades and how much to lower the gaps — a summary tells you how many gaps will be muted.
- A/B and download. Compare the original and gated versions, then export in the same format. By default the file length doesn't change, so the take stays in sync; tick “Shorten long gaps” only if you want a shorter file (podcast, voice-over).
Features
- Sync-safe by default. Gaps are muted, not cut out: the file keeps its exact length, so your take stays aligned in your DAW.
- Optional: shorten long gaps. For podcasts and voice-overs, every gap longer than a length you choose (0.25 to 3 s) is cut down to it, with a short crossfade at each join — words and the padding around them are never cut. The file gets shorter, so it no longer lines up with the original.
- Automatic threshold, precise controls. The threshold is measured on your take (and stays adjustable), with minimum silence, padding around phrases, smooth fades against clicks, and gap reduction: mute, or lower by 6 to 60 dB to keep some air.
- Breaths and mouth noises too. Short, unpitched sounds between phrases are treated as silence; anything with pitch — sung or spoken words — is kept.
- A/B preview. Switch between the original and the gated version before exporting.
- 100% in your browser. Your vocal never leaves your computer — analysis and rendering run locally.
FAQ
Is this silence gate free?
Yes. Without an account you get 2 free uses per day across AudioKit's free tools; with a free account, 5 per day. Tuning the gate and A/B-ing is always free: a use is only counted when you export. AudioKit Premium removes the daily limit.
Is my vocal take uploaded to a server?
No. The file is decoded, analyzed and rendered entirely in your browser (Web Audio API + ffmpeg.wasm) and never leaves your machine. The only network call is a tiny anonymous counter that tracks your daily free quota.
Which audio formats are supported?
MP3, WAV, FLAC, AIFF, OGG and M4A files up to 500 MB. The gated file is exported in the same format as the source: lossless stays lossless, lossy formats are re-encoded at high quality.
Will it cut into my words or change the timing?
Not by default. The padding setting keeps a margin of audio around each phrase so attacks and tails survive, and the minimum-silence setting stops the gate from firing in short gaps between syllables. And since silences are muted rather than removed, the file length — and your sync — doesn't change. The one exception is the “Shorten long gaps” option, off by default: it removes part of the long gaps on purpose, so the file gets shorter.
Can it remove or shorten long pauses?
Yes, with the “Shorten long gaps” option (off by default): every gap longer than the length you set — 0.5 s by default, from 0.25 to 3 s — is cut down to that length, with a 10 ms crossfade at each join so there's no click. Words and the padding around them are never cut, and the summary shows the new length before you export. Use it for podcasts and voice-overs, not for a vocal that has to stay in sync with a beat.
Does it remove background noise from the voice itself?
No — and that's by design. A gate mutes the gaps between phrases (breaths, mouth noise, room tone) but can't touch noise that sits on top of the voice while it's sounding. To denoise the voice itself, use our AI-powered Enhance Voice tool instead.