What is missing in deep music generation? A study of repetition and structure in popular music

TL;DR

This paper investigates the role of repetition and hierarchical structure in popular music, revealing insights that highlight gaps in current deep learning music generation systems and proposing new evaluation criteria.

Contribution

It provides a detailed analysis of music structure and repetition patterns across datasets, offering new perspectives for improving deep music generation models.

Findings

Music structure exists at multiple hierarchical levels

Songs use repetition and limited vocabulary, diverging from general collection statistics

Repetition patterns follow a general trend related to predictability

Abstract

Structure is one of the most essential aspects of music, and music structure is commonly indicated through repetition. However, the nature of repetition and structure in music is still not well understood, especially in the context of music generation, and much remains to be explored with Music Information Retrieval (MIR) techniques. Analyses of two popular music datasets (Chinese and American) illustrate important music construction principles: (1) structure exists at multiple hierarchical levels, (2) songs use repetition and limited vocabulary so that individual songs do not follow general statistics of song collections, (3) structure interacts with rhythm, melody, harmony, and predictability, and (4) over the course of a song, repetition is not random, but follows a general trend as revealed by cross-entropy. These and other findings offer challenges as well as opportunities for deep-learning music generation and suggest new formal music criteria and evaluation methods. Music from recent music generation systems is analyzed and compared to human-composed music in our datasets, often revealing striking differences from a structural perspective.