Podcast audio cleanup

If you have separate tracks, yes: clean each one, then mix them. Every track gets its own noise removal and you keep control of the balance. A single mixed file works too.

Around −16 LUFS is the common target, and it is what the Podcast setting gives you, with true peaks kept under −1 dBTP so encoding doesn’t clip.

No. Where it hears music it turns it down under the voice, and only so far, so a music bed dips instead of cutting in and out. For a clean music level, add it after you clean the voices.

Yes. A long episode works the same as a short clip and the progress bar shows the time left. How long a file can be depends on your plan; see the limits on this page.

On a paid plan it gives you the transcript to write them from. Save the episode to your library and choose Transcribe: you get plain text for show notes and an accessible transcript, plus a version with timestamps, in 30 languages, or automatic detection. Only then is the audio sent to our speech AI, which keeps nothing; copy or download the result, because we don’t store it. Check names and terms before you publish.

No. The cleaning runs inside your browser, on your own computer. Your audio and video never leave your device; our server only receives the file’s length and format to check your allowance.

A free account cleans 30 minutes a month, no card needed. The paid plans, Creator (9 hours), Pro (30 hours) and Studio (100 hours) a month, take files up to 3 hours and 4 GB each. See the pricing page for details.