#algorithmiccomposition

Tower of Babel (ToB), as I’ve currently named it, is the project I’ve chosen to dedicate my summer to as part of my Sound Design MSc. Over the course of the programme I’ve been getting hooked on developing my own tools for creating sound, primarily in MaxMSP, and it feels fitting to have my year here culminate in a project exploring this.

Concept

ToB is the continuation of work I started in my second semester ‘Non-Realtime Systems’ module. The concept is a sound generation environment controlled through standard English rather than a programming language such as SonicPi or Supercollider. As such, it would be a fairly expressionistic approach to sound, rather than the highly controllable environments offered by DAWs and coding environments, and therefore it should be stressed that anything I develop here will be developed through the lens of my existing work and approaches to sound.

History

The author as wee boy

Ever since I was an edgy edgy teen, I’ve been making electronic music, and while the early results were fairly atrocious, over the more recent years, I’ve evolved a personal voice, dwelling mostly in the shadow of the mighty oak that is experimental ambient music. As such the sonic results of this project will most likely fit under this description - it’s quite a large and varied field, but it would be unreasonable of you, the reader to expect top 40 hits, or lounge jazz classics to be a likely outcome.

Tower of Babel v1.0 

The first iteration of the code exists as a relatively ugly/scrappy Max patch and is available to download and poke here, with some example outputs available in a separate blog post. It uses a 2013 semantic dictionary to drive the analysis and derives scale intervals from valence. The mapping of analysis to sound is quite simple, and there are some glitches and holes in the analysis engine, but it works as proof of concept.

Going Forward

My main tasks are, as I can see them:

Build a new/expand upon existing semantic dictionaries to produce something suitable for sound

Start building a lexicon of sound for the software to draw upon