Towards a gestural control of environmental sound texture synthesis
Aymeric Masurelle · 2011
Through this project, a development of a methodology to guide the creation of intuitive and pertinent mappings between one-hand gesture and environmental sound texture (rain, wind, wave) descriptors is shown. The proposed methodology is composed of two complementary key stages. First, a perceptual investigation on sound descriptors controlling a corpus-based concatenative synthesis process is carried out to figure out the relevant ones. Afterwards an experiment where participants perform gestures while listening to sound texture morphologies is achieved to create a sound and gesture-related descriptor dataset. Then by applying a canonical correlation analysis on those sound and gesture-related data, one is able to determine pertinent gesture features and their correspondences with the retrieved relevant sound features. Thus pertinent clues to create gesture-to-sound mappings are provided by this method.