COMPARING ARTIFICIAL INTELLIGENCE MUSIC AND HUMAN MUSIC: A CASE STUDY OF PROF. DAVID COPE’S EMMY SYSTEM

I. Ren

Abstract:This is a pilot study analysing the difference between algorithmic compositions and human music, and subsequently evaluating the authenticity of music-composing algorithms. In this project, we use tools from information theory to study the music pieces created by a well-known system, “Emmy”, built by Prof. David Cope. We make the comparison between Emmy’s compositions in the style of J.S. Bach’s chorales [1] and actual chorales pieces written by Bach [2]. The main metric we use is the Shannon entropy (shortened as entropy in the paper), which measures the unpredictability of information content. We found that the overall entropy of Emmy’s pieces are in general higher than the real Bach’s pieces, indicating that the AI music are less predictable and are short of repeated notes. We also design the tool of entropy profile and the entropy window to examine the evolution of the entropy within the piece. It turns out that Emmy’s pieces have higher and smoother pitch entropy profiles, relatively authentic duration entropy profile, which gives clues on how to improve the composition algorithm: reduce the introduction of new pitch patterns. Taking both the pitch and duration into consideration, Emmy’s music have higher entropy profiles throughout, which confirms the observations and conclusions from the whole-piece entropy perspective.

Computer Science,Art