This project analyzes a musical score (encoded as XML, MEI, MusicXML, etc.) by passing a sliding window over the surface and applying the discrete Fourier transform (DFT) to the pitch classes within the window. The result is a numerical representation of the harmonic qualia that can be viewed as tables or as graphical visualizations.
Download or clone all files. In addition to Python 3.8, the following Python packages are required: music21, numpy, pandas, plotly, and tkinter. Use the notebook DFT_Main to run the program and generate visualizations. A small corpus is included, but additional files can be added in DFT_Corpus. Visualizations are interactive plots that are saved as html. The save location can be edited in DFT_Graphing.