Free Online Silence Remover
Cut dead air from podcasts and voice recordings.
Drag & drop your audio file here
or
Max file size: 100 MB · MP3, WAV, FLAC, OGG
Original
beats per minute
musical key
Anything quieter than this counts as silence. -40 dB suits speech; raise it (toward -20 dB) for noisy rooms.
Only pauses at least this long are cut. Shorter gaps are kept, so natural speech rhythm survives.
Keeps a little audio on each side of every detected speech region, so word endings and breaths are never clipped.
Processing…
Result
How it works
The Silence Remover tightens up spoken audio by cutting the dead air between words and sentences. It scans your file in 20 ms frames, measures the energy of each frame, and drops every silent stretch longer than your minimum-silence setting — pauses, room tone gaps, and the silence at the start and end of a recording.
Three controls keep it honest: the threshold decides what counts as silent, the minimum silence protects natural speech rhythm (short gaps stay), and hangover pads every speech region so word endings are never clipped. The result line tells you exactly how much was removed, e.g. "Removed 12.4 s of silence".
This is a buffer (offline) tool: the whole file is analyzed and rebuilt with your settings, then returned as a downloadable file. Everything runs client-side in your browser, so your files never leave your device and nothing is uploaded anywhere.
Frequently asked questions
How does it decide what is silence?
The audio is scanned in 20 ms frames and the RMS energy of each frame is compared against your threshold. Frames quieter than the threshold are "silent"; silent runs longer than the minimum-silence setting are cut, shorter ones are kept so the result still sounds natural.
Will it cut quiet speech or fade-outs?
It can, if they fall below the threshold — that is what the hangover control is for. Hangover keeps a margin of audio around every detected speech region, protecting soft word endings and breaths. If quiet passages disappear, lower the threshold (toward -60 dB) or raise the hangover.
What is the hangover setting for?
Hangover extends each detected speech region by a few tenths of a second on both sides. Speech rarely starts and stops cleanly at a threshold — without hangover, the first and last milliseconds of words can get clipped. 0.1 s is a safe default.
Is there a file size limit? Is it private?
Files up to 100MB are supported. Detection and cutting run entirely in your browser - your audio never leaves your device.