Music nerds
with too many tracks.
Timbre is made by a small crew. For years we have been digging for music, DJing and tinkering with tracks. And we kept running into the same problem: the bigger the library, the harder it gets to find the one track that fits right now. Timbre is our answer to that.
Timbre estimates what makes a track tick – and makes it visible with one yardstick.
Why Timbre
Somewhere in your library sits the track that holds your set together and keeps the flow going – Timbre helps you find that missing piece. Every DJ app shows BPM and key. What’s missing is what actually defines a track: how much energy is in there, the genre mix, how the track sounds, what fits next. Timbre estimates exactly that, with the same yardstick for every track in your library – and shows you the result instead of a gut feeling.
- Tempo
- 126 BPM
- Key
- 1m A minor · mixes with 12m, 2m, 1d
- Energy
- ++
Genre mix
Sound profile – rank within the library, 0 to 100
How we work
Your music stays yours.
On its own, Timbre only reads from your online storage: measure, forget. From it we keep measurements, fingerprints and the names of your tracks – no music.
Built for the booth, not the lab.
Every feature has to hold up in real sets, not demo data.
Genres are blends.
Hardly any track is just techno or just house; Timbre shows the mix instead of a box.
We love music and burn for it. And anyone with a lot of music runs into the same thing again and again: keeping it in order. When you play or produce, flow is what counts. Timbre takes the searching off your mind. And leaves you more time for listening.
What Timbre builds on
From the community, for the community.
Timbre uses these open-source libraries and this open research – built over many years by people who share what they know. Thank you. Our own work (taxonomy, energy, catalog, display) builds on top.
Essentia and the Music Technology Group models
Genre mix, style, mood and vocals – with Essentia and the pre-trained models of the Music Technology Group at Universitat Pompeu Fabra, Barcelona.
Essentia code: AGPL-3.0 · Models: CC BY-NC-SA 4.0 / BY-NC-ND 4.0
AudioMuse-AI
Open-source software behind sound search and similar-sounding tracks.
AGPL-3.0
librosa
Tempo, key and many signal measurements.
ISC
Chromaprint
The acoustic fingerprint that lets the catalog recognize tracks.
LGPL-2.1
FFmpeg
Reads almost every audio format so we can measure it.
LGPL-2.1
Running alongside: FastAPI, PostgreSQL, Authentik and many more open components.
Papers, BibTeX and what comes from where: Research
Write to us
Feedback, ideas, bugs, a track Timbre gets wrong: hello@timbremusic.app. We read everything.