AN EXTENSIBLE TOOL FOR AUTOMATED MUSIC GENERATION

Algorithmic composition is the usage of mathematical processes to produce music. In this study an extensible programmatical toolkit for automated music generation is developed and used to produce and analyze algorithmic music examples. This work will also provide an overview of current research in multi-ple computer science areas connected to algorithmic music. Several techniques are implemented including genetic algorithms, Markov models, Lindenmayer systems and used to generate experimental data for both quantitative (through general statistics) and qualitative (through human feedback) analysis. The work seeks to demonstrate that algorithms are a capable means of producing useful musical results, either influenced by human input or stand-alone. While the interpretation of output will vary, and perhaps it is difficult to define the boundary for what should be considered music and what should not, the computer’s natural ability to structure and manipulate musical data structures make it a good candidate for a virtual composition platform