Characterization of musical performance using physical sound synthesis models

Abstract

Sound synthesis can be considered as a tool to characterize a musical performance together with the performer himself. Indeed, the association of real-time synthesis algorithms based on an accurate description of the physical behavior of the instrument and its control parameters, and gestures capture devices the output parameters of which can be recorded simultaneously with the synthesised sound, can be considered as a "spying chain" allowing the study of a musical performance. In this paper, we present a clarinet synthesis model the parameters and controls of which are fully and explicitely related to the physics, and we use this model as a starting point to make a link between the playing and the sound. © Springer-Verlag Berlin Heidelberg 2004.