← Blog
Scheduled — goes live September 1, 2026. Only people with this link can see it until then.

Remove Silence From Video or Podcast Audio, Free

September 1, 2026 · 6 min read · by the ClipCraft team

The fastest way to remove silence from video or podcast audio is to stop scrubbing for it. ClipCraft's Silence Remover takes any audio or video file, finds every stretch quieter than a threshold you control, marks each one in red on a waveform, and exports the tightened file, all in your browser with nothing uploaded anywhere. I fed it a 35-second test take with six long pauses while writing this: it flagged all six, showed "-0:12.5 removed" and a new length of 0:22.4 before I touched anything, and the defaults were right.

How the detection works

The tool decodes your file's audio track and measures loudness in 21-millisecond windows, then applies three rules you can see and change. Anything quieter than the silence level (default -40 dB) counts as silence. Only silences at least as long as the minimum gap (default 0.8 seconds) get cut, so the natural half-beat between sentences survives. And each cut leaves a little padding (default 0.15 seconds) on both sides, which is the difference between a tightened take and one that sounds like the air got vacuumed out of the room.

The Silence Remover with a podcast waveform, six red regions marked for cutting, and stats showing 12.5 seconds removed
My 35-second test take at the default settings: six silences found, 12.5 seconds gone, 22.4 seconds left.

Tuning the three sliders

The sliders update the red regions and the stats live, which makes dialing them in fast. Drop the silence level toward -60 dB if your room has hum or breath sounds getting flagged; raise the minimum gap if you want short pauses kept for pacing. When I raised the gap from 0.8 to 2 seconds on my test file, one shorter pause got spared: five cuts instead of six, 11.2 seconds removed instead of 12.5. There are also two green handles on the waveform that limit the analysis to a selection, useful when the intro music is supposed to have quiet in it and only the talking needs tightening.

The Silence Remover after raising the minimum gap to 2 seconds, sparing one short pause and cutting five silences
Same file, minimum gap raised to 2 s: one pause survives, and the stats update as you drag.

Preview before you commit

The preview button plays the result, not the original: playback hops over every red region so you hear exactly how the tightened version flows. This is where you catch the one cut that clips a breath weirdly, nudge the padding up a notch, and listen again. Nothing renders until you hit export, so the loop is quick.

Exporting: audio cuts clean, video re-records

For audio files the export is sample-accurate: the kept segments are joined digitally and encoded to MP3 at 192 kbps or 16-bit WAV, so the cuts are exactly where the red regions were. Video works differently, and honestly: the tool replays your video while re-capturing it, skipping the gaps, which means a video export runs in real time (a 10-minute trimmed result takes about 10 minutes, tab open) and re-compresses the picture. For a podcast that's uploaded as audio anyway, export the MP3 and you lose nothing. An export costs the standard small tool charge: 2 tokens on the free plan, 1 on Saver, none on Creator or Studio, and the public tool page works signed out with 4 anonymous tokens. Plans are on the pricing page.

Why in-browser matters here

Podcast episodes get edited before they're published, which means the file you're cleaning up is usually unreleased. Every cloud silence-cutter asks you to upload that file to someone's server first. This one doesn't have a server side at all: the decode, the detection, the preview and the encode happen in the tab, and the network stays quiet. If you recorded the take with our screen recorder or miked up through the vocal recording flow in SoundCraft, the whole record-clean-publish chain never leaves your machine.

What it doesn't do

Podcast editors usually run this pass last, after the content edit. Cut your bad takes first, then let the threshold sweep the remaining dead air; doing it in the other order means re-detecting every time you rearrange something.

Dead air is the easiest quality win in audio: nobody ever complained a podcast moved too well. Drop a file in and check the "removed" number on your own episode; mine average around a minute per twenty. A free account covers plenty of exports a month, and the tool itself has no upsell inside it.

Cut the dead air, keep the episode

Threshold, minimum gap and padding you control, a live preview, and sample-accurate MP3/WAV export. 100% in your browser; unreleased episodes stay on your machine.

Open the Silence Remover