Touchups is where you measure each song's silence boundaries and integrated loudness so that AutoCast can play your library cleaner — trimming dead air at the head and tail of files and matching every song to a consistent on-air loudness target. It's two features that work together but that you can run independently.
Librarian never modifies your audio files. Every Touchups operation writes its results into the library database and leaves the original audio untouched. AutoCast reads those values at playback time and acts on them — by skipping past leading silence, ending playback before trailing silence, and applying a per-song volume adjustment to bring the song to your station's target loudness.
Touchups offers two ways to apply these measurements to your songs:
Two operations, each independently togglable:
Librarian scans the file and finds where audible content actually begins and ends. It writes two values into the library:
You can choose to trim only the start, only the end, or both. The audio file itself is unchanged — only AutoCast's playback behavior changes when it reads these values.
Librarian measures the file's integrated loudness using the broadcast-standard ITU-R BS.1770 algorithm and writes two values into the library:
−18.3.Voice or Music, telling AutoCast which of its two loudness preferences this song should be matched to.At playback, AutoCast computes the gain needed to bring the song to the matching target (set in AutoCast Preferences) and applies that gain in real time. A song measured at −20 LUFS, tagged Voice, with AutoCast's Voice target set to −14 LUFS will play +6 dB louder than the file natively is — without re-encoding the file.
Use Voice for talk content (announcer tracks, liners, sweepers, news, weather) and Music for songs and instrumental beds. The two targets exist because spoken word and music typically benefit from slightly different loudness levels — voice wants to feel "up front", music wants room to breathe.
Aggressiveness controls how eager Silence Trimming is to call something "silence." A single slider drives three internal values together so you don't have to think about them individually:
| Slider | Threshold | Min silence length | Pre-roll left in |
|---|---|---|---|
| 0 (gentle) | −50 dBFS | 500 ms | 80 ms |
| 5 (default) | −60 dBFS | 300 ms | 50 ms |
| 10 (aggressive) | −70 dBFS | 100 ms | 20 ms |
Lower values are gentler — only obvious, lengthy near-silent passages will be trimmed. Higher values are more eager, cutting closer to the audible content. The default of 5 works for most material; bump it up if you have files with long quiet intros that should be tighter, or back it off if Librarian is trimming things that should stay.
"Pre-roll" is a tiny bit of audio left in before the first detected sound, so the trimmed playback doesn't start with a click or an abrupt jolt. At default, you get 50 ms of lead-in; at aggressive settings, only 20 ms.
By default, Touchups skips songs that already have values in the relevant columns — running again is a no-op for any file that's been scanned before. Tick Re-analyze already-scanned files to force a fresh measurement on every selected song. Useful when you've changed the Aggressiveness or LUFS target and want existing entries updated.
Two action buttons at the bottom:
TrimStart, TrimEnd, LUFS, LUFSTarget) for the selected songs. Confirms with a dialog before doing anything. Audio files are untouched — this just removes the annotations Librarian wrote.Restore Original applies to the songs currently selected in the middle pane. If nothing is selected, the button does nothing. If you want to clear Touchups data across your whole library, select all songs first (⌘A).
Once the values are written to the library, AutoCast honors them every time it plays a song:
target − measured and applies that many dB of gain to the song's playback. Songs measured quieter than the target play louder; songs measured louder than the target play softer. Every song ends up at roughly the same on-air loudness without any change to the file on disk.AutoCast also enforces a Clip Safety Ceiling on the gain value — a quiet file that would otherwise need +20 dB of gain to hit target gets capped, with optional soft limiting on the output, to prevent clipping or distortion. Configure this in AutoCast's Audio Levels preferences along with the Voice and Music loudness targets themselves.
If a song doesn't have any Touchups values written, AutoCast plays it exactly as before — no trimming, no gain adjustment. Touchups is purely additive: existing libraries keep working unchanged until you choose to scan something.