Gene-synth: An Intelligent Synthesiser for Musicians

Sebastian Beswick, University of Tasmania Part of CW12

I’m currently studying for a Bachelor of Computing with Honours at Utas, and as a keen amateur musician, I have a special interest in Computer Sound – after all, most modern music relies heavily on computer generated synthesised sounds.

Now, Wouldn’t it be cool if you could sit down in front of a computer and, after showing it what you liked, have it intelligently generate you unique works of art, (in this case synthesised sounds), based on your own personal aesthetic preferences?

My honours project Gene-synth leverages the incredible power of modern evolutionary computing and artificial intelligence techniques to deliver a solution that would allow all composers – regardless of their level of technical ability – to create beautiful original sounds. Gene-Synth works by automatically learning and modelling the musician’s sound preferences, and using Darwinian techniques (such as survival of the fittest) to breed generations of high-quality sounds for immediate use in their production.