Long pauses and dead air make a recording feel slow. This tool scans the file, finds silent regions below a dB threshold, and trims them — keeping the natural breaths and short pauses while cutting the awkward ones. Standard step in podcast and voiceover post-production.
Podcast, voiceover, lecture, interview — anything voice-heavy.
Up to 200 MB
Threshold sets what counts as silence. -50 dB works for clean studio recordings. Drop to -40 dB if there's room tone or noise to ignore. Minimum duration controls which pauses get removed. Set 0.7-1.0 s to keep natural pacing; set 0.3 s to tighten aggressively.
ffmpeg filter: silenceremove
ffmpeg makes two passes — one to trim leading silence, one to remove silences anywhere in the file longer than your minimum. Short pauses inside the threshold survive untouched.
Output: 16-bit WAV, trimmed duration
It depends on the noise floor of your recording. Studio mic into an interface, treated room: -50 dB. Untreated room with HVAC noise: -40 dB. Phone mic or laptop mic: -35 dB. Lower numbers (more negative) = more aggressive — quieter sounds get trimmed. Higher numbers (less negative) = more conservative.
Only if the threshold is set higher (less negative) than the quietest parts of your speech. If you set -30 dB and a soft consonant happens to dip to -32 dB, that consonant gets cut. Safe practice: set the threshold 6-10 dB below your average speech level. Listen to the result; if soft words got chopped, lower the threshold.
No. The minimum duration parameter protects pauses shorter than the threshold. Set 0.5 s and any silence shorter than half a second is preserved; longer silences get removed. For tight, fast pacing use 0.3 s; for natural conversational pacing use 0.7-1.0 s.
Yes. Manual trimming (with the Audio Trimmer or a DAW) lets you decide each cut. Silence Remover is automatic — it processes the whole file by rules. For 1-2 cuts in a song, trim manually. For removing dozens of awkward pauses across a 60-minute podcast, automatic is the only practical option.