Music Library Doctor
What is

What is acoustic fingerprinting for music libraries?

A content-based hash of the audio signal itself — stable across formats, bitrates, and tag spellings. It's what makes cross-format duplicate detection possible.

The problem

Filename-based duplicate detection misses duplicates with different filenames. Metadata-based detection misses duplicates with different tags. The same recording stored as `Artist - Title (320).mp3` at 320 kbps and `artist_title_192.mp3` at 192 kbps looks like two completely different files to any tool that compares names or tags. The only way to know if two files contain the same recording is to look at the audio itself — and that's what acoustic fingerprinting does.

How Music Library Doctor does it

  1. 1 Pick a fingerprint algorithm. MLD uses Chromaprint, the open-source algorithm developed for MusicBrainz Picard.
  2. 2 Decode the audio. The fingerprinter doesn't care about format — MP3, FLAC, ALAC, AAC, OGG all decode to PCM samples.
  3. 3 Compute the fingerprint. Chromaprint runs a spectral analysis over short windows of the audio and produces a compact hash that summarizes the spectral envelope over time.
  4. 4 Compare fingerprints. Two files with similar fingerprints contain similar audio. MLD's matcher tolerates small differences (different encoders produce slightly different PCM output) and groups files that are acoustically identical.
  5. 5 Score and decide. MLD's Group Scorer ranks copies inside each duplicate group by bitrate, file size, format, and folder location, recommending the best copy to keep.

Supported today

Rekordbox · Serato DJ · VirtualDJ (incl. Favorite Folders) on Windows 10+ and macOS (Apple Silicon + Intel).

Why native integration matters

The power of acoustic fingerprinting is invariance: the fingerprint stays nearly the same when you re-encode, re-tag, or rename a file, because the audio itself stays nearly the same. That makes it the only reliable way to catch cross-format duplicates and re-encoded copies of tracks you've already had for years. MLD layers it with the Group Scorer so you don't have to manually compare every group — the best copy is highlighted automatically.

Frequently asked questions

How is this different from a regular file hash?

A file hash (MD5, SHA256) changes when even one bit changes — different encoder, different tag, different bitrate all produce different hashes. An acoustic fingerprint is stable across those changes because it summarizes the audio content, not the file bytes.

Will it match a remix as a duplicate of the original?

No. The fingerprint reflects the actual audio — different mixes, different masters, and remixes all produce different fingerprints. Only acoustic-twin copies match.

Does Chromaprint know what song this is?

Chromaprint produces a fingerprint; matching it to a specific song name requires a database lookup (MusicBrainz, AcoustID). MLD uses Chromaprint locally for duplicate detection without lookups — the fingerprint comparison is what matters for dedup. Online identification is a separate optional feature.

How fast is the fingerprinting?

Roughly 5–10 seconds per track on a modern machine, since the audio has to be decoded. A 10,000-track library finishes in 10–30 minutes; results are cached so subsequent scans are instant.

Does anything leave my machine?

No. Fingerprints are computed locally and stay local. MLD never uploads audio or fingerprints.

Get your library in shape in minutes

Free tier covers detection and viewing. Lifetime access is $49 — $19 for the first 100 DJs.

Related guides