An Interface for Sequencing with Concatenative Sound Synthesis

Although there has been a good deal of research in concatenative synthesis, most existing approaches tend to focus on offline composition. The realtime systems that have been developed place emphasis on source pruning and on exploration rather than composition. This paper presents an interface for sequencing with concatenative synthesis with particular emphasis on realtime performance. The interface is designed to address some weaknesses of previous realtime sequencing systems, by allowing the user to seamlessly work at the local level of individual beats as well as the global level of musical passages. The interface deals mostly with the target side of concatenative synthesis by using sound classes, numerical descriptions of sound that group source sound units. The interface exploits the fact that sound classes are purely numerical and provides mechanisms for interpolation, modulation, and stochastic variation to aid in creating and modifying compositions in a performance context.