Neurythmic: A Rhythm Creation Tool Based on Central Pattern Generators by Daniel Bennett, Peter Bennett & Anne Roudaut
We describe the development of Neurythmic: an interactive system for the creation and performance of fluid, expressive musical rhythms using Central Pattern Generators (CPGs). CPGs are neural networks which generate adaptive rhythmic signals. They simulate structures in animals which underly behaviours such as heartbeat, gut peristalsis and complex motor control. Neurythmic is the first such system to use CPGs for interactive rhythm creation. We discuss how Neurythmic uses the entrainment behaviour of these networks to support the creation of rhythms while avoiding the rigidity of grid quantisation approaches. As well as discussing the development, design and evaluation of Neurythmic, we discuss relevant properties of the CPG networks used (Matsuoka's Neural Oscillator), and describe methods for their control. Evaluation with expert and professional musicians shows that Neurythmic is a versatile tool, adapting well to a range of quite different musical approaches.
Music: Central Pattern Generator Neural Networks, Nonlinear Dynamic Systems, Feedback, NoiseOther research: HCI, effect of computer systems on autonomy
