Cross-library duplicate detection across Rekordbox, Serato, and VirtualDJ. Confidence-tiered groups, per-copy quality scores, and a recommended keeper — with every playlist re-pointed before a single file moves to Trash.
Years of DJing leaves duplicate tracks scattered across multiple folders and drives — same song ripped at different bitrates, bought from different stores with different tags, synced between machines with different filenames. Rekordbox, Serato, and VirtualDJ don't have built-in dedupe that understands both *exact* copies and *near*-duplicates — and even if they did, they wouldn't protect you from breaking playlists when you delete the wrong copy. The other failure mode is worse: blunt dedupe tools that group two *different* songs together and let you trash one with a single careless click.
Supported today
Rekordbox · Serato DJ · VirtualDJ (incl. Favorite Folders) on Windows 10+ and macOS (Apple Silicon + Intel).
Most dedupe tools only look at filenames, so they miss the real problem: the same song stored as 'Track.mp3' at 320 kbps and 'track(1).mp3' at 192 kbps. MLD layers three signals — file hashing for exact copies, metadata tiers for near-duplicates, and opt-in acoustic fingerprinting for everything tags can't see — and shows you *why* each group matched with a confidence number instead of a take-it-or-leave-it verdict. Just as deliberately, there is no blind 'delete all duplicates' button: cleanup happens group by group, with quality scores and audio preview at hand, and playlists are re-pointed to the kept copy BEFORE any file is moved. That review-first design is why nothing breaks.
A content-based hash: two files with different tags, bitrates, or filenames match if they contain the same audio. Useful when you have the same song from multiple stores with different metadata. In MLD it's an opt-in second scan — strong matches group automatically, borderline ones are flagged ACOUSTIC REVIEW for human ears.
Each group recommends a keeper — typically the largest, highest-quality copy — and shows every copy's bitrate, duration, file size, BPM, and key next to it. The per-group Quality Check adds an objective 0–10 score from the actual audio spectrum, so a low-quality re-encode can't hide behind a healthy-looking bitrate tag. The final call is always yours: one click moves the KEEP marker.
No — that's the core guarantee. MLD rewrites every Rekordbox, Serato, and VirtualDJ playlist to reference the surviving copy *before* moving the losing copies. Playlist membership is preserved end-to-end.
Losers go to macOS Trash / Windows Recycle Bin. Fully reversible until you empty the bin. MLD never hard-deletes.
Yes — every group is reviewable. Pick a different winner with one click, preview the actual audio of any copy, or run the group's Quality Check and let the scores settle the argument. Groups you're not sure about you simply leave alone.
Yes. Scanning is read-only. Nothing moves without your per-group confirmation, files only go to Trash/Recycle Bin, and playlists are re-pointed first so references are never orphaned. At least one copy in every group is always kept, and any borderline match lands in a separate review list MLD never removes on its own.
No. Folder Library mode runs the same duplicate detection — including the acoustic scan and quality scores — on any music folder. Plex, Navidrome, and plain-folder collections are first-class citizens.
Every copy carries its own match-score badge — an exact match or just a close one — next to its bitrate and quality. You spot the true twin at a glance instead of trusting a single verdict for the whole group.
Free tier covers scanning and detection. Pro is a one-time lifetime license — paid once, no subscription. Current founding pricing is on the homepage.
