This work is structured as a Jupyter Notebook plus a Processing sketch.
The notebook is used to analyze a score from Roman Haubenstock-Ramati Ludus Musicalis compositions, Punkte and Pizzicati in particular. Using Multi Template Matching the signs in the graphical score are recognized and tracked, and a MIDI rendering and a MusicXML transcription are produced.
Instead of supplying a requirements.txt file to automatically install the dependencies comments in the code are used to provide information about what packages are needed, leaving the possibility to swap or omit them if not needed for a specific application. This has been tested with Python 3.12 on Windows and Mac.
It is possible to analyze both image files or acquire images from a webcam.
For copyright reasons it is not possible to supply the original scores, 2 hand written examples are provided instead.
The processing skecth is a companion material meant to be used during a performance for providing an advancing timeline.
This work hasbeen accepted at the Sound and Music Computing 2025 conference.
Davide Andrea Mauro and Axel Berndt. “Automatic Rendering of Ramati’s Ludus Musicalis Scores”. Accepted at 22nd Sound and Music Computing Conference, July 07-12th, Graz (Austria). 2025.