Patent-Pending Segment Analysis

Your music has a
harmonic fingerprint.
We reveal it.

Most tools give you one key per song. Music doesn't work that way. Parrser analyzes every section — detecting the keys, transitions, and harmonic patterns that make a track unique.

Get Early Access

Two views. One analysis.

Every track gets two outputs: a DNA Card that captures what your music is, and a Leyline that shows where it goes. Same data, two perspectives.

X-Ray — Structure View
01 — DNA Card

What your music is

This is Gorillaz — 5 tracks, 84 segments, 8 distinct keys. Every color, strand, and attractor maps to real harmonic data: A# Minor dominance at 25%, restless movement between tonal centers, bright D# Major departures. The art isn’t random. It’s generated directly from the analysis.

5 Tracks 84 Segments 8 Keys Deterministic
02 — Leyline

Where your music goes

Dan Terminus — “Noir 01.” The Leyline maps every key change across the timeline. C# Major anchors the first minute, pivots to C# Minor at 1:15, then late departures to Ab Major add tension. DJs get Camelot codes. Producers see structure. Sync teams see harmonic complexity at a glance.

Camelot Codes Timeline View Key Changes Mix Points

Built different. On purpose.

Parrser is a self-contained analysis engine. No Spotify API. No third-party dependencies. Music theory and signal processing, running on our own infrastructure.

Segment

Analysis beats full-track detection

Patent-pending segment-based harmonic profiling significantly outperforms single-key detection on complex music — catching key changes, modal shifts, and tonal movements that full-track averaging misses entirely.

0

External APIs required

Spotify killed their Audio Features API in Nov 2024. Every tool that depended on it broke. Parrser runs on Essentia, self-hosted. We don’t depend on anyone.

First

Per-section harmonic journey

No consumer tool shows how a song moves through keys over time. Mixed In Key, Tunebat, Rekordbox, Traktor — they all return one label. We show the full journey.

40D

Feature vector per track

Key distribution, harmonic journey, spectral centroid, onset density, beat strength, MFCCs, tonnetz, chroma — the full harmonic and timbral fingerprint.

Benchmarking in progress across a larger corpus — quantitative results to be published. Your music stays yours — analysis requires no account and no data leaves our infrastructure.
Read our research →
Essentia 2.2 Segment Analysis Confidence-Weighted Deterministic Art Self-Hosted Patent Pending

Early Access

See what your music looks like