Turn any recording into accurate sheet music

ML-powered music transcription with per-instrument stem separation, per-note confidence scoring, and automatic tablature generation. Built for musicians, producers, and music educators.

Processing Pipeline

1Audio Input
2Stem Separation
3Pitch Detection
4Note Quantization
5Confidence Scoring
6Sheet Music / Tab

6

Core Features

5

Input Formats

4

Output Formats

6

Instruments

Core Features

Stem Separation

Isolate vocals, guitar, bass, drums, keys, and other instruments from any audio recording using state-of-the-art source separation models.

Per-Note Confidence

Every transcribed note includes a confidence score so you know exactly which sections need manual review.

Auto Tablature

Generates guitar, bass, and ukulele tablature from the transcription. Handles alternate tunings and capo positions.

Pitch Detection

Frame-level pitch analysis with sub-semitone accuracy. Handles polyphonic material and overlapping notes.

Note Quantization

Intelligent rhythmic quantization that preserves musical intent. Handles triplets, swing, and rubato passages.

Multi-Format Export

Export to MusicXML, MIDI, PDF sheet music, or Guitar Pro. Import into any DAW or notation software.

Supported Formats

Input

MP3WAVFLACOGGM4A

Output

MusicXMLMIDIPDFGuitar Pro

Instruments

GuitarBassPianoDrumsVocalsStrings

Tunings

StandardDrop DDADGADOpen GCustom
In Development

ML pipeline is functional

Stem separation and pitch detection are working. Tablature generation and confidence scoring are in progress.

Follow Progress