Automating algorithmic representations of musical structure using IGME: The Interactive Generative Music Environment

Hunt, Samuel J; Mitchell, Thomas; Nash, Chris

Abstract

In this paper we explore the recreation of existing musical compositions by representing the music as a series of unique musical bars, and other bars that can be replicated through various algorithmic transformations, inside the Interactive Generative Music Environment software, or IGME. This re-composition approach is intended to explore whether the pre-existing music could have been created using the processed based approaches offered by the IGME software. If music can be expressed by algorithmic processes then we propose that original works of music can be expressed or created in the same way. Such a justification can provide a rationale for creating the unique compositional processes and workflows that IGME affords to those looking to compose with generative and algorithmic music techniques, and avoid many of the pitfalls of generative music. Music can be imported into IGME and automatically analysed to find unique bars, and bars that have been transformed from them. The overall timeline can be visualised to quickly demonstrate the structure of the music, using colour to differentiate unique musical ideas, and arrow-arcs to show the relationships between different parts. Such a process reduces the overall entropy of the music data and provides an educational insight into macro level music structures. Each of the techniques are explained and examples given. In addition, data sets have been pre-computed for several genres of music, showcasing the distribution of different types of techniques.

• Presentation Conference Type: Conference Paper (unpublished)
• Conference Name: Innovation In Music 2019
• Start Date: Dec 5, 2019
• End Date: Dec 7, 2019
• Deposit Date: Feb 25, 2020
• Publicly Available Date: Feb 27, 2020
• Public URL: https://uwe-repository.worktribe.com/output/5455331