OScofo

OScofo: OpenScore Follower

OScofo is an open-source project designed to provide score following capabilities for contemporary music applications. Originally developed as a PureData (Pd) object, OScofo has now been expanded into a versatile C++ library that integrates seamlessly with multiple environments, including a Max Object, Python package, and others comming. Currently under development, OScofo is already functional and serves as a valuable tool for researchers and musicians.

Goal

The aim of OScofo is to provide a straightforward and accessible tool for real-time score following. By keeping the software lightweight, it can run seamlessly on the web through the pd4web platform, thanks to the ability to use PureData directly in web browsers. With pd4web and OScofo will be possible to use the software in rehearsals with just a single click, eliminating the need for external libraries, compatibility issues, or complex installations -- ultimately facilitating the sharing and performance of contemporary music.

Collaboration and Contribution

I invite composers, researchers and developers to contribute to the OScofo project. Not just with code, but with theory, math, etc. I am trying to make OScofo acessible via a Python implementation, to test it should be easy. By sharing the source code, I am trying to provide access to the theories and mathematical formulas that drive the software, all this come from the amazing research work of Arshia Cont and Philippe Cuvillier at IRCAM.

Technical Foundations

OScofo uses several concepts developed by many researches (with focus on the research of Cont and Cuvillier).

• Pitch Comparison: Utilizes the Kullback-Leibler (KL) Divergence method for pitch comparison as presented by Christopher Raphael (2006), Arshia Cont in 2008 and 2010.
• Rhythm Synchronization: Integrates theories of rhythm synchronization developed by Edward Large and Mari Riess Jones (1999) and Edward Large and Caroline Palmer (2002), as presented for Cont (2010).
• Forward Algorithm: For now, OScofo uses the equation presented by Arshia Cont (2010) and developed by Yann Guédon (2005).
• Score Language: Based on the scofo (by Miller Puckette) and antescofo~ (by Arshia Cont, Philippe Cuvillier, and others) language.

Building

Requirements

• cmake and ninja (pip install cmake ninja)
• boost (brew install boost, apt install boost, pacman -S mingw-w64-x86_64-boost)
• On Windows, you need mingw64.

Optional

• treesitter (npm install tree-sitter) (If you want to change/update score syntax)
• pybind11 (optional): To build Python package.
• PureData (optional): To build the Pd Object.

Building Options

• BUILD_ALL: Build all OScofo modules (Python, Pd, Max).

• BUILD_PY_MODULE: Build or not the OScofo python module.

• BUILD_PD_OBJECT: Build or not the Pd Object.

• BUILD_MAX_OBJECT: Build or not the Max Object.

• PDLIBDIR: Where the Pd object will be installed

Building Steps

git clone https://github.com/charlesneimog/OScofo --recursive
cmake . -B build -DBUILD_ALL=ON -G Ninja 
cmake --build build

To install use cmake --install build.