pyAMPACT
pyAMPACT (Python-based Automatic Music Performance Analysis and Comparison Toolkit) is a python package that links symbolic and audio music representations to facilitate score-informed estimation of performance data in audio as well as general linking of symbolic and audio music representations with a variety of annotations. pyAMPACT can read a range of symbolic formats and can output note-linked audio descriptors/performance data into MEI and Humdrum kern files. The audio analysis uses score alignment to calculate time-frequency regions of importance for each note in the symbolic representation from which to estimate a range of parameters. These include tuning-, dynamics-, and timbre-related performance descriptors, while timing-related information is available from the score alignment. Beyond performance data estimation, pyAMPACT also facilitates multi-modal investigations through its robust infrastructure for linking symbolic representations and annotations to audio.
Welcome to pyAMPACT’s documentation!
Getting Started