The Algorithmic Composition Explorer Using a learning environment to enable users to experience intrinsic feedback on musical algorithms

Abstract

This thesis proposes a novel design for an interactive system that introduces people to algorithmic composition. There are many libraries and environments for algorithmic music, with little consistency in terms of programming language and software framework. For those that have little or no programming knowledge, or those that just want to get a tactile understanding and quick feel of a selection of different algorithmic approaches, there are few options. To use most existing algorithmic composition tools users are still required to carry out some implementation or integration work, which also require some expertise in computer music frameworks. This thesis proposes a system that attempts to fill that gap. The primary aim is to build a system that takes a learning through practice approach, allowing learners to get an understanding of how a particular algorithm works by experimenting with its parameters and listening to the results. If a lack of technical knowledge should not be a barrier to the use of the tool, the system needs to be widely available, and not requiring of any complex installation or setup. Finally, the system is designed such that it can be built upon and expanded, and released as open-source software. There are limits on what can be built as part of a thesis such as this, so the design must provide a framework for future expansion.