Computational Creativity: Compose the Music for a Movie using only its Automatically Extracted Brightness Curve

TL;DR

This paper presents a method where a computer automatically extracts a brightness curve from a silent movie and uses it to compose music that synchronizes with and aesthetically complements the film, demonstrating a novel approach to computational musical creativity.

Contribution

It introduces a novel process of automatic brightness curve extraction and rule-based composition to generate film music, showcasing a new tool for creative collaboration between AI and human artists.

Findings

The composed music synchronized well with the film's visual content.

The process produced aesthetically pleasing and surprising musical pieces.

The method enabled a degree of creative freedom in composition.

Abstract

Since its conception, the computer has found applications to accompany human creativity. Today, the debate about computers and creativity involves several challenges, such as understanding human creativity, modeling the creative process, and programming the computer to exhibit behavior that appears to be creative to some extent. In this paper, we are interested in how the computer can be used as a tool to promote creativity in a musical composition. We automatically extracted the brightness curve from a silent movie and then used it to compose a piece of music to accompany the movie. We extracted several parameters from the brightness curve, and applied compositional rules from these parameters to write the instrumental music for the film. The final composition has a synchronicity and aesthetic fit with the film that are surprising. This compositional process also allowed for a degree of aesthetic freedom that would otherwise have been impossible.