Mini Audition 5.7: clean it up, hand it in

Editing a recording is half the job. The other half is cleaning it up and handing it over in the shape someone asked for: −16 LUFS for a podcast, −23 for broadcast, 24-bit WAV with the right metadata, the same treatment on every episode. Mini Audition 5.7 does that half too.

Noise reduction

Effects › Repair › Noise Reduction takes steady noise out of a recording — hiss, hum, a fan, the hum of a room — under the voice too, not only between phrases as a gate does. Select a stretch of the noise alone and choose Effects › Capture Noise Print (⇧⌘P), and it knows exactly what to take. Without one, it learns the noise from the quietest moments as it plays.

Reduction, Sensitivity and Smoothing set how much and how carefully. Switch Output to Noise only to hear what it’s taking away: if you hear the voice in there, back off.

Spectral editing

Some noises aren’t steady: a cough, a phone, a chair, a squeak. On the spectrogram, drag up or down as well as across, and you select just the frequencies they’re in, for just as long as they last. Then, from Edit › Selected Frequencies:

• Fill from Around (⌥⌘F) brings those frequencies down to what they sound like just before and after, so the room carries on where the cough was;
• Remove (Delete) takes them out;
• or Lower them by 6, 12 or 24 dB.

The voice or music under it stays as it was. Every edit can be undone, and the band stays selected to go again.

Will it pass?

Transport › Loudness Report (⌥⌘L) measures the track, or a selection, the way a delivery spec does: integrated loudness, loudness range, true peak, sample peak, and the loudest moment. It checks the result against streaming, Apple Music and podcasts, EBU R128 and ATSC A/85, tells you how much gain each needs, and warns you when that gain would push the peaks over the ceiling, so it needs a limiter first. Normalize to any of them from there, or copy the report into a delivery note.

The true peak is the level between the samples, where a file that never touches 0 dBFS can still clip once it’s converted. It’s the number specs cap at −1 dBTP.

Export it the way they want it

Export now writes WAV (16-bit, 24-bit or 32-bit float), AIFF, FLAC and AAC up to 320 kbps, at the track’s rate or converted to another. Going down to 16 or 24 bits adds dither, so quiet passages and fades don’t turn grainy. A WAV can carry Broadcast WAV info: a description, the originator, the date, where it starts in its source, and its loudness, as broadcasters and archives expect.

A whole series at once

File › Batch Process (⌥⌘B) takes a folder of files through the same steps without opening them: your effect chain (Audio Units included), leveling to a loudness target or the true peak to a ceiling, and export in any of those formats. A file whose peaks would go over the ceiling at the target is kept a little quieter and marked, so nothing clips in a batch nobody listens through. Each file shows its loudness and true peak when it’s done.

Get it

Mini Audition 5.7 is free, for macOS 15 or later. Every change is in the changelog.