beat-guard fingerprints your beats, timestamps them on Bitcoin, and finds them inside other songs, even with vocals rapped over the top.
Bigger artists take beats from smaller, up-and-coming producers all the time, and the producer usually can't prove a thing.
I'm a producer. I built beat-guard so producers like me have real evidence: proof of when we made a beat, and proof of exactly where it shows up in someone else's song.
The same core idea Shazam uses, pointed at a different problem.
The loudest frequency peaks become a "constellation" that survives vocals, compression and noise. register beat.wav
Your file's SHA-256 hash is anchored on the Bitcoin blockchain with OpenTimestamps. Only the hash leaves your computer, never the audio.
Thousands of matching points that all line up at the same moment can't be a coincidence. It also catches beats sped up or slowed to hide them. check song.mp3
A clean report shows where your beat appears, which part was used, the speed change, and your proof of creation.
Tested on beats with vocals on top, every one run through MP3 compression like a real release.
A real report from the built-in demo: a beat sped up 3% with vocals on top, found after an 8-second intro.
Runs on your own computer with Python 3.9+. Your beats never get uploaded anywhere.
# install git clone https://github.com/jeffreyesdavid/beat-guard.git cd beat-guard python3 -m venv .venv && source .venv/bin/activate pip install -r requirements.txt # register your beats (export from your DAW as WAV or MP3) python -m beatguard register ~/Music/beats/*.wav # check a song you're suspicious about python -m beatguard check suspect_song.mp3 --report report.html
I build this in my own time so producers don't have to pay to protect their work. If it helped you, a coffee keeps it going.