The same spectral-subtraction Noise Reducer engine, pre-set to a stronger reduction level that suits spoken-word podcast recordings.
The Podcast Noise Reducer is tuned for speech: it cleans fans, hiss, hum and room noise from interviews and solo shows while keeping voices natural, with an option to preserve the breaths and room feel that keep pauses from sounding dead.
It is built on the same spectral engine as the general Noise Reducer, but with speech-friendly defaults, an optional gate for extra-quiet pauses and automatic levelling to −16 LUFS.
Spectral Wiener-style noise reduction (2048-point STFT, 75% overlap, frequency/time-smoothed gain) with a noise profile from your room-tone selection or the quietest 12% of frames. With ‘Preserve natural breaths’ on, the strength is scaled to 85% so some noise floor remains; with it off, an expander (−52 dB threshold, −10 dB) hushes the gaps. Finally the audio is levelled to −16 LUFS (−1 dBTP).
Remove fan and computer noise.
Clean up Zoom/Skype recordings.
Clean voice-overs.
Reduce room tone.
| Engine | STFT spectral gating, 2048 FFT, 75% overlap |
|---|---|
| Profile | Room-tone selection or auto |
| Extras | Optional gate, −16 LUFS levelling |
The sound of the empty room, used to teach noise reduction what to remove.
No — reverb is different from noise.
It's the same engine with the strength pre-set higher, on the assumption that podcast recordings often have more consistent, removable background noise than music, where over-processing is more noticeable.
This tool runs entirely in your browser using the Web Audio API. Nothing is uploaded, stored or shared — there is no server to send your file to, so it works offline once the page has loaded.
Be the first to share your experience with this tool — your feedback helps others.