Art of Record Production Conference
May 17-19, 2019
Berklee College of Music, 921 Boylston Street, Boston, MA 02215
Friday, May 17 • 15:00 - 15:30
Algorithmic Music Scoring: An Approach to Automated Composition for Audiovisuals

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

In this paper, I review the concept of algorithmic composition and discuss its applicability and implementation in audiovisuals. Although the practical replacement of a human composer is not the objective in this approach, it is possible to envision a tool to leverage efficiency and variability on music creation for a wide variety of audiovisuals. Depending on narrative needs, productions may range from a thorough human-made piece, with traditional recording and mixing stages, to a highly variable, computer-produced sequence for multi-linear and interactive environments. Tools as Amper Music®, an AI music composer among others, currently provide a framework for automated music segment production tailored to audiovisuals. Although it features speed, control and royalty-free music output through an online browser interface, its proprietary algorithms are not open to develop a particular composer’s idea or style. It also opens a discussion about formulaic compositional techniques through style templates.

While algorithmic note-by-note generation potentially offers flexibility and infinite diversity, it poses significant challenges such as achieving performance sensibility and producing a distinctive narrative style through program design. Starting by evaluating suitability based on a range of audiovisual categories, I examine possibilities, advantages, and challenges offered by algorithmic composition studies that employ Markov models, a- life/evolutionary music, agents, generative grammars, and artificial neural networks/deep learning. According to those models, I outline rule-based strategies for music transformation in synchrony to emotional cues. Finally, I propose a compositional tool design based in modular instances of algorithmic music generation, featuring stylistic development in connection with an audio engine.

avatar for Alvaro Lopez

Alvaro Lopez

University of California Riverside
Alvaro Lopez is an electronic musician, composer and sound designer, BM in Composition and Production, and MA in Music Technology. He is currently a PhD candidate in Digital Composition at UCR focused on artificial intelligence for music analysis, generation and composition. Lately... Read More →