Two sub-tabs. The Detect tab now holds two tools: Music Sleuth, described on this page, which identifies songs from their audio, and the Baddie Detector, which finds damaged audio files and lets you replace or delete them.
Desktop only. The Import tab and all of its sub-pages are part of Librarian Desktop and are not available in Librarian Mobile. To bring new audio files into your library, use Librarian on your station Mac.
Sometimes a file's name and its embedded tags both come up empty — a generic rip labeled Track 04.mp3 or 00-NoTitle.mp3 with nothing usable inside it either. That's what the Detect sub-tab — nicknamed Music Sleuth — is for. It listens to the actual audio, fingerprints it, and looks it up in an online music database to identify Artist, Title, Album, and Genre.
Detect needs an internet connection to look anything up. It doesn't touch your files themselves — it only writes to the library database (the same place Embedded Info and Filenames write to), so filenames on disk are never changed.
Select the songs you want identified in the middle pane, then click Go!. For each selected song, Librarian:
Each match is saved to the library the moment it's found, not batched up until the end — so if you're running Detect against a large batch, you can check a song's Attributes right away without waiting for the whole run to finish.
Unlike Embedded Info, Detect does not ask whether to fill blanks only or override existing values — a confident match always overwrites whatever's currently in the Artist, Title, Album, and Genre fields for that song. If you've already hand-corrected metadata for a song, don't run Detect on it unless you're fine with it being replaced.
The box below the Go! button fills in live as Librarian works through your selection — one line per song, showing what was found (and its match confidence) or that nothing confident turned up. It's worth a scan after a big batch: songs it couldn't identify are exactly the ones still worth a manual look, or a try through Embedded Info or the AudD fallback below.
AcoustID's database is free and crowd-sourced, which means excellent coverage for mainstream, well-circulated recordings and spottier coverage for older, obscure, or vinyl-sourced material — the kind of thing a lot of station libraries are full of. A song not being found doesn't necessarily mean anything is wrong with the file; it may simply not be in AcoustID's database yet. That's exactly the gap the AudD fallback below is meant to help with.
If a song comes up empty through the free AcoustID lookup, Librarian can optionally try a second, paid service — AudD — which draws on a larger, commercially licensed fingerprint database and tends to catch songs AcoustID misses.
AudD is pay-as-you-go — roughly half a cent (US$0.005) per lookup at standard rates, billed only for songs actually sent to it. Because Librarian only calls AudD when the free AcoustID lookup has already failed, you're never paying for songs the free service already found.
Every time you click Go! with AudD Fallback turned on, Librarian shows a confirmation dialog naming how many songs are selected before it starts — a reminder that a paid service may be used, so a big batch doesn't rack up lookups without you noticing.
If your token is missing, mistyped, or your AudD account has run out of funds, the results log will show:
⚠️ AudD is turned on, but connection was refused. Confirm your API token is correct and your account is funded.
This message appears once per run (not once per song) to keep the log readable. If you see it, check your token and account balance at dashboard.audd.io before running Detect again.