Beat and Tempo Fluctuation Analysis with "beat_it" Toolbox

Disclaimer: This project is currently in a very early alpha stage of development. A first alpha version is planned for release later this year.

This repository supports the workshop "How Constant Is Your Beat? Computer-Assisted Analysis of Beat and Tempo Fluctuations from Acousmatic Music to Minimal Techno with 'beat_it' Toolbox," presented at the Rhythm under the Microscope conference (University of Music and Performing Arts Vienna, 25-27 September 2024). It also accompanies the presentation "Understanding and Emulating Time: Analyzing and Simulating Musical Microrhythm Timing with the beat_it Toolbox" at the inmusic25 conference (University of Bath, 20-22 June 2025).

The repository provides Jupyter notebooks that demonstrate techniques for analyzing beat and tempo fluctuations using the beat_it Python toolbox. It also includes all the necessary audio examples and annotations.

Repository Structure

• **/examples/**: Contains the audio example files for analysis.
• **/csv/**: Includes corresponding cue points and annotations in CSV format.
• **/sv/**: Provides annotation layers for import into Sonic Visualizer.
• **/sv_project_files/**: Contains complete Sonic Visualizer project files (.sv).
  • Note: Audio files may need to be manually relinked from the /examples/ directory after opening a project.
• **/slides/**: Slides from presentations at Rhythm under the Microscope and inmusic25

Getting Started

Prerequisites

The code and notebooks in this repository are optimized for Python 3.11. We recommend using a virtual environment to manage dependencies.

Installation

1. Clone the repository:

   git clone <your-repository-link>
   cd <your-repository-name>

2. Set up the environment:

   We suggest using Conda to create and manage the environment.

   # Create and activate a new conda environment
   conda create -n beat_it python=3.11
   conda activate beat_it

3. Install dependencies:

   Install the required packages from the requirements.txt file.

   pip install -r requirements.txt

4. Optional Pitch Detection Dependencies:

   To use the pitch detection notebooks (e.g., parm_pitch.ipynb), install these additional libraries:

   pip install crepe parselmouth tensorflow

Usage

Once you have set up the environment and installed the required packages, you can run the Jupyter notebooks to explore various analyses.

Colab Notebooks

The Jupyter notebooks are also available for use in Google Colab. You can access and run the notebooks directly from the following Colab folder:

Colab Folder for Notebooks

License

This project is distributed under the MIT License.