The batch jobs worth running before you trust a library

Three background jobs turn a folder of files into a library you can schedule against. Run them once, benefit indefinitely.

A freshly imported library is a list of files. Turning it into something a rotation clock can select from takes three batch jobs, all of which run in the background.

1. Loudness analysis

Preferences → Broadcast Engine → Analyze all tracks walks the current playlist on a background thread with live progress, filling the loudness cache.

Until this runs, tracks play at whatever level they were mastered at, and the difference between a 1980s master and a modern one is audible and irritating.

Results cache to disk keyed by path, size and modification time, so this is a one-time cost per file. A warm-up service also queues missing entries after imports, limited to two concurrent analyses so it cannot compete with playback.

2. Cue point detection

Automation → Analyze Library Cue Points runs intro and outro detection across the whole music library.

Detection works in layers: RMS frames for leading and trailing silence, an energy envelope pass to refine positions, then a peak-transient pass for tightly trimmed files with no silence at all. Each result carries confidence scores and a review flag.

Manual cues can be preserved rather than overwritten. An operator who set a marker by ear outranks the algorithm, and the software should assume so.

3. Library analysis

The AI Library Analyzer runs deterministic library and cue checks, including a fix-missing-cues batch. Despite the name it does not require any model to be configured for the "Analyze library…" path — it is grouped with the AI tools by menu placement, not by dependency.

Checking your work

After these, the schedule confidence scan tells you what is still outstanding on a loaded playlist: missing files, missing cue points, tracks with no loudness cache, weak rotation category matches.

Ready, Review or Risk on the status bar, with a detail window behind it.

The order that makes sense

  1. Import folders; let metadata load in the background.
  2. Assign categories in bulk, using ADS, ID and VT prefixes for imaging.
  3. Run loudness analysis.
  4. Run cue point detection.
  5. Build a rotation clock and generate an hour.
  6. Read the preview timeline before committing.

Skipping steps three and four is the most common reason a first generated hour sounds worse than expected.

Related reading

Ready to hear it on your own station?

Tell us about your setup and we'll send you a personalised quote.

Request a Quote