Multidisciplinary project call from Collegium Musicæ, Musicology and Computer Science. Enhancing the MORFOS software to study polyphonic works and orchestration from a perceptual perspective, through an in-depth analysis of multi-scale formal diagrams.
The research conducted in this post-doctorate focuses on the study of polyphony in musical works within the framework of the MORFOS algorithm. Currently, the software can only analyze homophonic works by creating a multi-scale data structure called the Multi-Scale Oracle.
The goal is to enhance this data structure to handle multiple musical voices simultaneously in a given work. Beyond adapting the underlying data structures, the challenge of separating voices from audio files arises. Extending the software in this direction would enable the study of polyphonic works and orchestration from a perceptual standpoint, through the in-depth analysis of multi-scale formal diagrams.
The scientific objective of such a project is to assist musicologists in better understanding cognitive phenomena during musical listening, particularly in the context of orchestration. This is achieved by generating representations via the MORFOS software, called multi-scale formal diagrams. These representations take polyphonic works into account and allow musicologists analyzing them to reconstruct the different musical voices and instrumentations that contribute to the orchestration of a given musical work.