An Automated Music Improviser Using a Genetic Algorithm Driven Synthesis Engine

This is a single authored chapter in Applications of Evolutionary Computing. The volume is part of the series Lecture Notes in Computer Science, a peer-reviewed publication that began in 1973.

The origins of the chapter began in a paper delivered at EvoMusart 2007, the Fifth European Workshop on Evolutionary Music and Art held in Valencia, Spain. The Workshop represents leading edge developments in the application of biologically-inspired computer science to creative practice in the arts. Biologically-inspired computer science includes artificial intelligence, ‘swarm’ behaviour and evolutionary computation and it was the latter that the researcher employed to produce new improvising sound software.

The chapter describes the technical solution and aesthetic strategies deployed in developing software that enables a computer to be able to respond a stream of sound that it is presented with. The researcher is an improvising musician of long-standing and he was able to rely on this experience to influence his programming of the software so that it was able to analyse the sound that it encountered for the acoustic variables of frequency and amplitude and translate those into the musical equivalents of pitch and loudness.

The chapter compliments its interpretation of the software’s sensitivity to sound with a corresponding examination of how the software generates its own responses to that sound. In this way, the software emulates an artificial parallel of the processes that are said to distinguish the collective improvisational musician – a receptivity to the contributions of other improvisers and the ability to offer new acoustic events to the emerging collaborative composition.

The chapter contextualises the researcher’s own efforts in terms of both the broad tradition of improvisational musicianship and the practice of equipping computers with the ability to evolve in response to a changing environment.

• http://portal.acm.org/citation.cfm?id=15...
• http://www.lcc.arts.ac.uk/41749.htm