One Tool, Two Programs and Several Ideas for Composition with Spectral Modelling Synthesis

Celso Aguiar

Center of Computer Research in Music and Acoustics (CCRMA)
Depatment of music, Stanford University
Stanford, California 94305-8180
USA

Abstract:

Sound can be approached from complementary perspectives. This paper describes aspects of one of my works, 'Piece of Mind' for tape, where time domain techniques derived from 'musique concréte' are blended to powerful spectral modelling (SMS) in the processing of natural sounds. Programs written in colaboration with Xavier Serra have made SMS available in CLM's synthesis environment. Deterministic components are synthesized by a highly efficient IFFT algorithm that relieves the delay of pleasure composers have historically accumulated with additive synthesis.

Bell sounds and singing voice are hybridized or cross-faded to provoke ambiguity of identity. On the other hand, the way SMS handles its stochastic part exposes the fact that dsp guys and composers may not always speak the same language. Being aware of what synthesis techniques are geared to and the ability to present proper solutions are first requirements. A graphical interface to display SMS information on a single view is introduced.