Midi Miner -- A Python library for tonal tension and track classification
This tool addresses the need for automated music analysis in computational musicology, though it is incremental as it implements existing models.
The authors introduced Midi Miner, a Python library that calculates tonal tension using a spiral array model and classifies MIDI tracks into melody, bass, and harmony, enabling cleaner data for analysis and training.
We present a Python library, called Midi Miner, that can calculate tonal tension and classify different tracks. MIDI (Music Instrument Digital Interface) is a hardware and software standard for communicating musical events between digital music devices. It is often used for tasks such as music representation, communication between devices, and even music generation [5]. Tension is an essential element of the music listening experience, which can come from a number of musical features including timbre, loudness and harmony [3]. Midi Miner provides a Python implementation for the tonal tension model based on the spiral array [1] as presented by Herremans and Chew [4]. Midi Miner also performs key estimation and includes a track classifier that can disentangle melody, bass, and harmony tracks. Even though tracks are often separated in MIDI files, the musical function of each track is not always clear. The track classifier keeps the identified tracks and discards messy tracks, which can enable further analysis and training tasks.