% open dj-dedupe
DJ Dedupe
Finds duplicate tracks in your DJ library and removes them without breaking a single playlist.
PROBLEM
DJs download the same track twice all the time, a different night, a slightly different rip, and over years it quietly eats the drive. You can't just delete them either: rekordbox tracks every file by its location, so delete one in Finder and the track goes red and any playlist using it breaks. So the dupes just sit there.
WHAT I BUILT
A Mac app that finds the real duplicates, including the same song saved at different quality, lets you pick which copy to keep, then removes the rest safely. It relinks every playlist to the kept copy, quarantines the loser instead of deleting it, and backs the database up first. Nothing ever shows up missing. Download it, drag it to Applications, hit Scan: one-click undo on everything.
HOW IT WORKS
rekordbox keeps its whole library in an encrypted database. The tool opens it, fingerprints the actual audio so a 320 MP3 and a WAV of the same track match but two different songs never do, and groups the dupes. On go it adds the keeper into each playlist where the duplicate sat, removes the duplicate, deletes its library entry, and moves the file to quarantine. Every run is backed up and reversible.
WHERE IT'S AT
Live and packaged as a Mac app: download the DMG, drag it to Applications, done (Apple Silicon, rekordbox 6 or 7). I ran it on my own library first, 13,000 tracks, 882 duplicate groups, about 9.5GB to reclaim, and tested every database write on a copy (delete, reopen, integrity check) before it ever touched the real collection.
WHAT I LEARNED
The nervous part was writing back into rekordbox's database, where a wrong move corrupts the whole collection. So I tested every write on a copy first, deleted a track, reopened it, ran an integrity check, confirmed playlists still pointed at the right place, and only then let it near the real library.
STACK
Python · pyrekordbox · Chromaprint · FastAPI · PyInstaller
% cd ..