Audio5 min

How to Remove Silence from a Podcast Automatically

Detect and remove long silences from interviews and podcasts automatically — configurable threshold and padding, all in your browser.

Silence detection and removal run locally via Zro7 Remove Audio Silence using ffmpeg.wasm. Interviews never leave your browser.

To automatically remove long silences from a podcast or interview, use Zro7 Remove Audio Silence. It runs FFmpeg's silenceremove filter to detect stretches quieter than a threshold you set, and trims them — with optional padding so cuts don't feel abrupt. Everything happens in your browser.

Why silence removal saves hours of editing

A one-hour interview with typical pauses is often 6–10 minutes of silence. Manually cutting each pause takes 2–3× the audio length in a DAW. Automated removal takes 30 seconds and lands within a few frames of what a human editor would do — good enough for podcasts, voiceover drafts, and rough cuts.

Tuning the two knobs that matter

  • Threshold (dB) — anything quieter counts as silence. -40 dB is a good start for clean recordings; -30 dB is more aggressive; -50 dB preserves subtle breath sounds.
  • Min duration (seconds) — silences shorter than this are kept (natural breathing rhythm). 0.5s is default; 1s is safer for conversational feel.

Steps

  1. Open Remove Audio Silence.
  2. Drop the audio (or video — the tool operates on the audio track).
  3. Pick threshold and minimum silence duration.
  4. Optionally add padding (e.g. 200 ms) around each cut for smoother transitions.
  5. Preview the trimmed waveform; click Export.

Chain with other cleanup

The Podcast Cleaner tool combines this with denoise + normalize in one click — the fastest path from a raw interview recording to publish-ready audio.

Frequently asked questions

Will it cut in the middle of a word?

Not with sensible settings. Threshold detects breath-level silence; minimum-duration prevents chopping micro-pauses inside speech.

Does it also work on music?

Yes but be careful — songs with intentional soft passages may be trimmed. Use a lower threshold (-55 dB) and longer min duration (2s+).

Can I keep some silence for pacing?

Yes — the tool can leave the first N ms of every detected silence untouched, mimicking a natural pause instead of a hard cut.

How large a file can I process?

Multi-hour recordings work. Silence detection is a fast filter; the bottleneck is the final re-encode.

Any upload?

No. ffmpeg.wasm handles everything inside your browser.

Related posts

← Back to blog