Analysis tool

This section contains a proof-of-concept (PoC) implementation of a pattern detection algorithm for symbolic music, specifically designed to extract frequent similar patterns from musical pieces. It implements limited functionalities derived from a more general, proprietary algorithm.

You cant test the algorithm using this MusicXML sample file or this ABC notation sample file.

Upload
Accepted formats: MusicXML (`.xml`, `.mxl`, `.musicxml`), ABCNotation (`.abc`), MIDI (`.mid`), Kern (`.krn`).
A list of one or more viewpoints to use for representing the music. Choose between 1 and 3.
4
Minimum length (number of notes/events) for a pattern to be considered (`3` to `12`).
8
Maximum length (number of notes/events) for a pattern (`Min length` to `50`).
4
Minimum number of times a pattern (or a similar variation) must occur to be reported.
0
Minimum similarity distance threshold used by the [`Levenshtein` distance function]. The exact interpretation depends on the distance function implementation, but lower values generally mean higher similarity is required. `0` set the algorithm to mine exact matches only.
If enabled, grace notes are kept during the initial cleaning/parsing of the music file. Otherwise, they are removed.