Understanding the Language of Music : From an Etude Generation Perspective

Abstract

In recent years there has been a focus on the accessibility and equity of music instruction in under-served communities. The value of private/guided instruction is clear, but the question of reaching students from lower economic classes presents challenges our current educational system struggles to address. The Blugold Computational Music Suite is a system that supports novice to intermediate musicians in their musical goals by giving them access to computer generated lessons that mimic what a private instructor would assign to them. This work investigates methods for automatically generating musical etudes that are derived from a given musical selection. We compare various methods implemented from the current literature by quantifying the similarities between the generated music and the original selection. Using the similarity results, we make recommendations for which methods should be considered when generating etudes for musical instruction.