Digital music provided the tools to a revolution in the understanding of sound and music. We are particularly interested in extracting information that even a human with a well trained ear cannot do, or at least express.

By music analysis, we mean the discovery of patterns in data represented in musical form. For example, midi/musicXML files, music scores and tablatures, and midi/OSC events captured in live performances. We performed a wide range of researches in this field, from functional harmony analysis to the discovery of microtiming and microdynamics patterns, from modeling rhythm and variation, to visualisation of common harmonic paths by composers.

By sound analysis, we mean the extraction of information directly from raw data, such as wav/mp3 files or microphone audio streams. This involves the development of new technology to create games, the automatic detection of dangerous sounds (like gunshots and explosions) for public security, and music information retrieval (MIR) applications (such as automatic chord recognition).

One of the main results of this research is our contribution to projects at the MusiGames Studio and the Audio Alerta innitiative.