In our ongoing work on evolutionary music composition, we explore linear genetic programming as a method of creating a virtual music composer. This process hinges on viewing the composer as a Turing-complete virtual register machine that outputs pieces of music. In this paper we compare different designs for the virtual machine, exploring various instruction sets and memory architectures; analysing their ability to create music statistically similar to that of a given corpus. We also explore different genotype sizes to see how much memory the virtual machine needs to converge to an acceptable result.
This content is only available as a PDF.
© 2017 Massachusetts Institute of Technology Published under a Creative Commons Attribution 4.0 International (CC BY 4.0) license
This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.