Remove Silence from Video and Audio

Shorten quiet pauses in a recording, with control over every suggested cut. Review the waveform, build the edited file and download locally. Free, with no upload, account or added watermark.

Processed on your device · No account · No added watermark

Choose audio or video

Detect low-volume sections, review the suggested cuts, then build your shortened preview.

Audio analysis and export run on this device. Processing libraries download when needed; your recording stays local. No AI model or transcription is required.

How to remove quiet pauses

Select a local recording and analyze it. A video needs an audio track. The waveform and proposed cuts appear after analysis.
Set the pacing before manual edits. Start with Balanced: −35 dB, a 0.5-second minimum pause and 120 ms padding. Natural and Tight offer different starting points.
Review each proposal. Select a pause, listen around its boundaries and uncheck anything worth keeping. Adjust start and end times where needed; check the time saved.
Choose the output. Select WAV, MP3 or video with audio, then build the shortened preview. Leave the tab open during export.
Play the complete result and download. Listen for clipped words and watch any picture jumps before saving.

Example: keep a quiet demonstration in a tutorial

A screen recording may contain a two-second pause while you wait for a menu, then another pause while you show an important setting. Both can be quiet enough to detect. Remove the waiting time but disable the cut over the demonstration, so viewers still have time to see it.

Keep the delivery natural

If words lose their endings, lower the threshold or increase padding. If phrases feel rushed, increase minimum pause or try Natural. Short audio fades soften joins; they cannot restore a word removed by a poor cut.

Trim a section first for a shorter test. Use background noise removal for steady hiss or Audio Cutter for freely chosen sections. Add subtitles after the cuts, and browse all free tools.

Silence removal questions

Do I need to upload my recording or sign up?

No. Analysis and export run in your browser, without an account or added DojoClip watermark. Processing libraries download when needed; your recording stays on your device.

Does it remove “um,” “uh” or background noise?

It detects quiet sections from volume, without AI or transcription. It does not recognize filler words, repeated sentences or mistakes. Removing silence shortens time; it does not clean noise underneath speech.

Which recordings and limits are supported?

Use audio or video with a browser-decodable mono or stereo audio track at 8–96 kHz. There is no fixed input duration or file-size cap, but long recordings can exceed memory or output-format limits. At most 300 cut proposals are supported.

What do the three detection controls change?

Threshold (−60 to −15 dB) sets quietness; higher values are more aggressive. Minimum pause (0.2–5 seconds) excludes shorter gaps. Padding (0–500 ms) keeps breathing room at interior cut edges. These settings propose cuts, not speech-aware decisions.

Can I keep a pause or revise a cut?

Yes. Uncheck a cut or edit its start and end, then use Undo or Redo to compare. Changing a preset or detection setting regenerates proposals and replaces manual adjustments; Undo can restore the previous edit.

Why can’t I build a result?

A shortened result needs at least one active cut and some retained content. Music or noise can prevent detection; an aggressive threshold can mark everything. Review the settings and selections. If there are more than 300 proposals, increase minimum pause or shorten the source.

What can I download?

WAV or MP3 for audio. Video input also offers video with sound, with the same intervals removed from both. Video is re-encoded up to 1080p and 30 fps, preferably MP4 with WebM fallback. Only primary video and audio tracks are included.

Is the shortened preview the actual edited file?

Yes. Building it renders the export rather than just skipping pauses in playback. The result player and download use the same file. Changing cuts or output clears it; rebuild and review the new version.