Remove Silence from Audio Online — Free
Automatically detect and remove silent segments from voice recordings, interviews, and podcasts. Tune the threshold and minimum gap to match your recording conditions.
Drop audio file here or click to browse
MP3, WAV, OGG, FLAC, and more
How Automatic Silence Removal Works
The tool scans your audio from start to finish, measuring the energy level (RMS amplitude) in small overlapping windows. When the measured level drops below the threshold for a duration longer than the minimum silence duration, that segment is identified as silence and removed from the output. A small padding value is preserved around speech segments to avoid clipping the beginnings or ends of words.
After detection, you can see how many segments will be removed and how much total time will be saved. Click Detect & Preview to listen to a preview, adjust the settings if needed, and then export the cleaned audio file.
Tips for Tuning Silence Detection
- Quiet recordings: Lower the threshold (e.g. −50 to −40 dBFS) so soft background noise counts as silence.
- Noisy rooms: Raise the threshold (e.g. −30 to −20 dBFS) so background noise is not mistakenly treated as speech.
- Natural pauses in speech: Increase the minimum silence duration (e.g. 0.5–1.0 s) to preserve intentional pauses while removing long dead air.
- Avoid cut-off words: Increase the padding (e.g. 0.1–0.2 s) to keep a small buffer around each speech segment.
Podcast-Specific Use Cases
- Remove dead air from the beginning and end of a recorded episode
- Clean up filler pauses and awkward gaps in a solo recording
- Tighten the pacing of a lecture or educational recording
- Batch-process interview recordings to remove hold music or connection gaps
Frequently Asked Questions
How does the silence detection threshold work?
The threshold is measured in dBFS (decibels relative to full scale). Audio segments with an RMS level below this value are treated as silence. Lower values (e.g. −50 dBFS) are more sensitive and will catch very quiet sections. Higher values (e.g. −25 dBFS) will only remove clearly silent passages.
What threshold works best for voice recordings?
−40 to −30 dBFS usually works well for voice recorded in a quiet room. If the tool is cutting parts of speech, lower the threshold. If background noise is being kept, raise the threshold.
Will words be cut off at the edges?
The padding parameter adds a small buffer around each detected speech segment. Set it to 0.05–0.15 s to preserve natural-sounding word edges.
Can I remove silence from music files?
Yes, but the results may vary. Musical silence (rests, dramatic pauses) may or may not be intended. For music, a higher threshold and longer minimum duration is usually safer. For spoken word, the defaults work well out of the box.