Musical Synthesis by Means of Cellular Automata and Gestalt Patterns.

In order to emulate a composer-based approach to musical writing, the paper aims to to create of a new musical synthesis model based on gestalt matrices. The approach is described in three different levels: - a sound model: all instruments have their own interface and have specific roles within a musical composition. Every instrument typology is unique, and composers write for specific instrumental characteristics; - an action model: every performer has different ways to interact with the instrument itself. Each instrument is a powerful sound provider. The action performed on the instrument has the function of "choosing" which particular sound model to play; - a score model: every musical staff has specific functions in the score and represents a certain contribution to the outcome which is related to macro-structure analysis. A self-generating matrix, described by use of patterns, is applied to these basic concepts, thus giving them personal stochastic behaviors and creating a comparison system to reduce data processing. [ABSTRACT FROM AUTHOR]