Papers
Topics
Authors
Recent
Search
2000 character limit reached

Discovering novel computer music techniques by exploring the space of short computer programs

Published 6 Dec 2011 in cs.SD | (1112.1368v1)

Abstract: Very short computer programs, sometimes consisting of as few as three arithmetic operations in an infinite loop, can generate data that sounds like music when output as raw PCM audio. The space of such programs was recently explored by dozens of individuals within various on-line communities. This paper discusses the programs resulting from this exploratory work and highlights some rather unusual methods they use for synthesizing sound and generating musical structure.

Authors (1)
Citations (8)

Summary

  • The paper explores generating computer music by analyzing the output of very short C programs (bytebeat), particularly focusing on the use of time-based arithmetic expressions.
  • Key findings include the identification of novel music synthesis techniques heavily utilizing bitwise operations and bit shifts, creating unique structures like "Sierpinski harmonies."
  • The research demonstrates how methods like modular arithmetic and controlled overflow in these minimal programs can yield unexpected musical results and suggests future avenues for automated discovery.

Discovering Novel Computer Music Techniques through Bytebeat Exploration

The paper by Ville-Matias Heikkilä explores the innovative field of computer music, specifically through the exploration of the so-called "bytebeat" phenomenon. This exploration is characterized by the generation of sound and musical structures from very short computer programs written in C. These programs are capable of creating complex musical outputs reminiscent of established artistic styles like demoscene, glitch art, and chip music.

Technical Framework and Methodology

Heikkilä presents a framework where an infinite loop in a C program outputs raw PCM audio, mapped to unsigned 8-bit samples at a rate of 8000 samples per second. Within this simple structure, the variation in musical output is introduced through different arithmetic expressions applied to a time counter variable t. The foundational template for these programs is succinct and relies heavily on bitwise operations and modular arithmetic, serving as the core computational tools for generating audio signals.

The exploration process was greatly facilitated by online tools that allowed real-time generation and modification of musical expressions. These tools employed JavaScript and ActionScript, providing a platform for wide-ranging experimental contributions from an online community.

Findings and Analysis

The research identified 71 formulas, with 58 adhering to the constrained subset of C expressions that exclude function calls and other complex operations. The findings reveal frequent use of bitwise arithmetic and bit shifts, which are unconventional in traditional music synthesis methodologies. Specifically, expressions like t&t>\>8—referred to as Sierpinski harmonies—demonstrate unique binary manipulations to create multitonal melodies with distinct octave intervals reminiscent of Sierpinski triangles in their amplitude progression.

The study uncovers methodologies where pitch values are derived from bitwise arithmetic, such as in the expression t*(42&t>\>10). This technique involves using integer bitwise operations to control the frequency of waveforms, resulting in novel melodic structures not commonly produced by standard musical synthesis techniques.

Moreover, the study highlights expressions leveraging modular arithmetic to synthesize musical elements, emphasizing how the overflow behavior, typically considered undesirable, can be harnessed creatively in sound design. Such expressions introduce new ways to conceive rhythmic and percussive components by manipulating amplitude values.

Implications and Future Work

Heikkilä's work indicates that explorations into the space of short, computationally simple programs yield innovative approaches to music synthesis. It suggests potential avenues for further research, including automated expression generation via genetic algorithms and enhancing exploration platforms to encourage further discovery. The advent of bytebeat has already inspired software initiatives seeking to extend its capabilities, indicating a fertile ground for both artistic and computational exploration.

Future advancements could involve expanding the programmable syntax used in these expressions or integrating more robust computational methods to discover emergent musical patterns. The documented methodologies in this paper present a compelling intersection of computer science and art, offering new paradigms for algorithmic compositions.

In conclusion, Heikkilä's research underscores the potential of minimalistic programming paradigms in creating complex musical outputs and opens up new discussions on the applications of computational creativity in the field of digital music. As this area continues to evolve, it is likely to incorporate more sophisticated programming constructs, facilitating the discovery of even more musically engaging concepts.

Paper to Video (Beta)

No one has generated a video about this paper yet.

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Continue Learning

We haven't generated follow-up questions for this paper yet.

Collections

Sign up for free to add this paper to one or more collections.