01
Song Painting
RHYTHMENERGYSPECTRUMHARMONY
Generative Music Analysis
Turn any song into visual DNA.
MusicDNA turns a track into a 3D sound genome, painted in real time from rhythm, energy, spectrum, and structure. A 2D mode is also available for clean share cards.
No random dataAnalyzed in the browserNo account
Preview: Sound Genome 3D
Demo preview. After choosing a track, the genome is painted in real time from its rhythm, energy, spectrum, and structure.
Analysis station
Choose a track to analyze
Place audio files in the music folder, then select one to generate its visual DNA.
Want your own track analyzed or described in more detail? Contact us.
🎵
No tracks found.
Add an audio file to /public/music and update manifest.json to start.
- 1Add an .mp3, .wav, .ogg, or .m4a file to /public/music
- 2Add an entry to /public/music/manifest.json
- 3Reload the app
- 4Select the track and click Analyze
New
Compare two tracks
Analyze two tracks side by side and compare BPM, energy, brightness, dynamics, structure, mood, and key moments. MusicDNA helps you understand how two pieces of audio differ, locally in your browser, without uploading your files.
A
Track A
Track A
Start from the selected track
vs
B
Track B
Track B
Pick or upload a second track
- BPM, energy, brightness and dynamics, side by side
- Normalized 0-100% timeline alignment
- Real similarities and differences from the audio
01 · Pipeline
How MusicDNA works
Four steps from sound to signature.
- 1
Choose a track
Pick any audio file from your music folder.
- 2
Decode the audio
We decode the file with the Web Audio API in your browser.
- 3
Extract real features
Rhythm, energy, spectrum, harmony, and structure are extracted from the signal.
- 4
Generate visual DNA
Deterministic visualizations are rendered from the real analysis.
Visualizations are deterministic. The same file always produces the same DNA.
02 · Signal
What MusicDNA reads
Every visual is rooted in measured signal features.
Waveform
Downsampled peaks and RMS envelope across the track.
Rhythm
Onset strength, beat times, and BPM estimate via autocorrelation.
Spectrum
FFT frames, spectral centroid, flux, and low/mid/high balance.
Harmony
Chroma vector and key estimate via Krumhansl-Schmuckler profiles.
Structure
Section boundaries from spectral and energy novelty.
Dynamics
Loudness curve, dynamic range, loudest and quietest moments.
Not every feature is available for every track. Uncertain values are clearly labelled as estimates.
03 · Gallery
Seven visualization modes
Each mode is driven by real, deterministic analysis.
02
Sound Genome
RHYTHMSTRUCTUREENERGY
03
Waveform Landscape
DYNAMICSSTRUCTURE
04
Rhythm Constellation
MOTIFSRHYTHM
05
Spectral Aurora
TIMBRESPECTRUM
06
Harmonic Portrait
TONALITYHARMONY
07
Track City
STRUCTURESONG
Why
Why visual DNA matters
What music looks like changes how you listen.
See the structure
Discover where a track builds tension, slows down, and reaches its peak.
Compare sounds
Spot differences between versions, mixes, and tracks at a glance.
Create visual artifacts
Generate cards, posters, and DNA images from the songs you love.
Analyze without guessing
Every metric is derived from the real audio signal, not invented.
04 · Trust
About MusicDNA
MusicDNA turns sound into a visual fingerprint, built entirely from real audio analysis on your device.
No account
No sign-up. No tracking. No upload step.
No random data
Every metric is derived from the real signal.
Analyzed in the browser
Decoding and feature extraction happen on this device.
Export on your terms
Share cards and JSON only when you choose to.
All metrics come from analysis of the selected audio file. Estimated values are labelled. If a value cannot be computed reliably, we do not show it as a fact.