Building a tool to auto-tag and audiocheck tracks. What is your workflow? Wanna test?

DJ Software & Apps
1

Hey everyone,

I’ve been building a tool to automate the annoying admin work between downloading tracks and importing them into Rekordbox.

I got tired of:

  • fixing filenames
  • correcting inconsistent/missing ID3 tags
  • manually looking up metadata
  • adding artwork/BPM/genre/release dates
  • sorting files into folders
  • discovering “320kbps” files that were clearly transcoded/upscaled

Especially for tracks coming from pools, promos, edits, Bandcamp downloads and other less-standardised sources.

The workflow for many DJs seems to be something like:

download → rename → Mp3tag → metadata lookup → copy/paste tags → move files → import into RB

So what my tool does:

  • monitors an inbox folder
  • identifies tracks via Beatport
  • writes proper metadata automatically
  • renames/organises files based on your preferences
  • detects likely fake-320/transcoded files using spectral analysis
  • more checks/features to come
    Depending on your feedback I’ll probably add more sources for the metadata, audio quality checks and other features too :slight_smile:

Tech stack is mainly Python + Beatport v4 API + FFT-based audio analysis.

Before pushing this further, I’m trying to better understand how other DJs handle library management and metadata cleanup today.

So I made a short survey (~5 min):

(Beta invites are at the end of the survey.)

But also genuinely curious about the workflows here:

  • Do you clean/fix ID3 tags? If yes: manually, Mp3tag, OneTagger, custom scripts, something else?
  • Do you standardise filenames/folder structure?
  • Do you verify audio quality? Spectral check, Spek/Fakin’ The Funk, critical listening, something else?
  • What’s the most annoying/friction-heavy part between download and Rekordbox import?
2

Your read on the standard workflow is right — download → rename → Mp3tag → RB is exactly the grind. What changed it for me was narrowing the problem to “only trust what you can verify cheaply”:

Filenames: treat them as disposable. Standardize on Artist - Title (Mix), strip the pool noise, and never rely on the filename for metadata after import.

Tags: enforce a small schema — genre from your own list (not the pool’s), energy 1–10 you assign once per track, key normalized to a single notation. The payoff isn’t the batch tagging, it’s that six months later you can actually filter by it.

Fake 320s: the cheap reliable test is spectral — a real 320 has content above ~16kHz, transcodes show a cliff there. Anything borderline goes to a “quarantine” folder instead of being deleted, so you can check it later without regret.

Sorting: folder structure matters less than smart playlists if your software supports them — but that only works after tags are clean. That dependency is the part nobody mentions.

Curious how you’re handling the Beatport lookups — rate limits are usually where these tools get fragile in daily use. Happy to give it a spin on Windows when there’s a build.