Audizer
Clean-up and recording

Strip silence settings that don't chop your words

For a vocal, start with a threshold of −45 dB, remove only gaps longer than 350 ms, and keep 60 ms of padding around each phrase. Lower the threshold if soft word endings are being cut. Raise the minimum gap if the take is being chopped into too many pieces.

What the three settings do

Threshold. Audio below this level counts as silence. It has to sit above the room noise and below the quietest thing you want to keep. Measure the noise floor in a gap: if it's at −58 dBFS, a threshold of −45 to −50 dB is safe.

Minimum silence. The shortest gap that will be cut. Without it, every tiny pause between syllables becomes an edit. 300 to 500 ms keeps phrases whole.

Padding. Extra audio kept before and after each detected phrase. Words start with quiet consonants and breaths, and end with decaying vowels and room sound that fall under the threshold. Padding saves them.

Starting points by source

SourceThresholdMinimum gapPad before / after
Lead vocal−45 dB350 ms60 / 100 ms
Spoken word, podcast−40 dB500 ms80 / 150 ms
Backing vocals−40 dB300 ms40 / 80 ms
DI bass or guitar−50 dB500 ms20 / 200 ms
Toms−30 dB200 ms5 / 300 ms or more
Snare top, to cut hat bleed−25 dB100 ms3 / 150 ms

Getting it right

Run it and look at the result before you listen. Good output has one clip per phrase, each starting just before the waveform rises. If you see dozens of slivers, raise the minimum gap. If clips start right on the first loud peak, add padding in front.

Then check three things by ear:

  • Phrase starts. "H", "f" and "s" at the front of a word are quiet and get clipped first.
  • Phrase ends. Held notes that fade and words ending in "m" or "n".
  • Breaths. Decide whether you want them. More front padding keeps them; less removes them.

Every cut needs a short fade, 5 to 20 ms, or it can click.

When to skip it

Don't strip silence on anything with continuous sound: a strummed acoustic part, a pad, room mics, overheads. Don't use it on a vocal recorded in a noisy room if the noise is audible under the singing, because the noise will switch on and off with each phrase and draw attention to itself. Reduce the noise first, or leave the gaps and turn them down 10 dB.

Quick answers

Is strip silence the same as a noise gate?
They do a similar job. Strip silence makes permanent, visible edits you can adjust clip by clip. A gate works in real time on every playback and can chatter on borderline material.
Why does strip silence cut off the ends of words?
The threshold is too high or the padding too short. Lower the threshold 5 dB and raise the padding after each phrase to 100 ms or more.

Audizer does this when you ask.A Mac DAW with an engineer built in. Say what you want and it makes the edit. Free download.

Get Audizer

Updated 2026-11-15. Corrections: support@intuzu.com