Title of thesis: An Entrainment Based Approach to Generative Music
Supervisor: Dr Tillman Weyde and Dr Newton Armstrong
Research Group: Music Informatics Research Group
Overview and research interests
Entrainment, the co-ordination of temporarily structured events through interaction, is a fundamental trait of all musical behaviour. Ethnomusicologists are increasingly becoming aware of its importance as it offers a new approach to understanding music making and music perception as an integrated, embodied and interactive process.
Generative musical models have thus far failed to consider entrainment, instead opting for a probabilistic, symbolic approach to music generation. In this thesis, an entrainment based model of musical behaviour is developed to contribute to this understanding.
Drawing upon Synergetics, the theory that is concerned with how the cooperation of individual components of a complex system leads to the formation of new macroscopic spatial, temporal and functional structures. A network of coupled oscillators, which have been shown to be able to perform complex musical tasks such as beat induction, forms the base of the model. The oscillators are locally coupled and spatially embedded, allowing them to be entrained to an external or self-driven beat, in frequency synchrony, phase-locked synchrony, or harmonic relationships. These networks are then used as part of a deep network to produce rhythmical structured patterns.
- Generative music
- Evolutionary musicology
- Non-linear Dynamics
- Neural Networks.
Higher educational background
MSc Creative Systems, Sussex University
BA (hons) European Theatre Arts, Rose Bruford College
Lambert, A. (2012) 'A Stigmergic Model for Oscillator Synchronisation and its Application in Music Systems', Proceedings International Computer Music Conference, Lubjana.