EPISODE · Aug 19, 2025 · 11 MIN
FIREFLY / STARLING Project Proposal:—Algorithmic ‘SWARM’ spatial composition played LIVE!
from Cliff Crego · host Cliff Crego
FIREFLY/STARLING Project Proposal: One of the species of music I do is algorithmic composition. The intention is to side-step traditional metric notation altogether, and generate electronic sounds directly from their mathematical description. TWO projects I’m working on involve the movement of sounds in objective or real space. Inspired by the Varèse, Xenakis, and Le Corbusier 1958 World's Fair Philips Pavilion—we did a reconstruction of this with the ASKO SCHOENBERG Ensemble years ago—I am hoping to implement a grid of a 40-voice sound source distributed over 400 light/speakers. The simplest and most flexible way to do this would be an APP that audience members could download in situ that flashes colored lights on and off, while at the same time playing pitched percussive sound. This means that there are three versions of a piece: light only, lights + sound, and sound only. All versions could be performed in- or out-of-doors. The first two pieces I’d like to realize both use simple swarm algorithms abstracted from Nature: (1) FIREFLY (2) STARLING. FIREFLY: The primary parameter is how the 40 relatively autonomous voices—simple percussive, say, woodblock sounds—move in and out of synchronicity using complex, non-periodic rhythms. Secondary parameters are random spatial and random pitch distribution (a simple 4 octave C2 to C6 1/4-tone field). Other variables are total duration and max/min temporal density limits. STARLING: Here, there are two equal primary parameters, with an one octave ‘swarm’ of 1/4-tone pitches at high to low density over a six octave field—always maintaining coherence with the same simple percussive woodblock sound with clear spatial transients—that ‘flies’ around in pitch space. WORK PLAN: Start simple with laptop models: scale down to 4 voices in monophonic rendering; then, stepwise, scale up to 40 voices. Once optimal temporal and pitch densities are worked out, scale to spatial distribution adding a dynamic / amplitude 7-step scale ppp=far away // fff=very near, as well as stereo headphones output; from there, scale up to the next level with simple single-purpose smartphone wi-fi/cellular APP. That’s it. Each piece would last about 10 minutes. They wouldn’t be written down, but generated anew each time. If you’d like take part the FIREFLY/STARLING Project (pro bono) in any way—whether as coder, mathematician, AI specialist, or APP designer— or offer support in any way, you can email me at [email protected] AVIAN soloist: An Oregon Rock Wren, recorded in MAY IMAGE: Whitebark pine atop EASGLE CAP (Lost) GLACIER PEAK in background
What this episode covers
FIREFLY/STARLING Project Proposal: One of the species of music I do is algorithmic composition. The intention is to side-step traditional metric notation altogether, and generate electronic sounds directly from their mathematical description. TWO projects I’m working on involve the movement of sounds in objective or real space. Inspired by the Varèse, Xenakis, and Le Corbusier 1958 World's Fair Philips Pavilion—we did a reconstruction of this with the ASKO SCHOENBERG Ensemble years ago—I am hoping to implement a grid of a 40-voice sound source distributed over 400 light/speakers. The simplest and most flexible way to do this would be an APP that audience members could download in situ that flashes colored lights on and off, while at the same time playing pitched percussive sound. This means that there are three versions of a piece: light only, lights + sound, and sound only. All versions could be performed in- or out-of-doors. The first two pieces I’d like to realize both use simple swarm algorithms abstracted from Nature: (1) FIREFLY (2) STARLING. FIREFLY: The primary parameter is how the 40 relatively autonomous voices—simple percussive, say, woodblock sounds—move in and out of synchronicity using complex, non-periodic rhythms. Secondary parameters are random spatial and random pitch distribution (a simple 4 octave C2 to C6 1/4-tone field). Other variables are total duration and max/min temporal density limits. STARLING: Here, there are two equal primary parameters, with an one octave ‘swarm’ of 1/4-tone pitches at high to low density over a six octave field—always maintaining coherence with the same simple percussive woodblock sound with clear spatial transients—that ‘flies’ around in pitch space. WORK PLAN: Start simple with laptop models: scale down to 4 voices in monophonic rendering; then, stepwise, scale up to 40 voices. Once optimal temporal and pitch densities are worked out, scale to spatial distribution adding a dynamic / amplitude 7-step scale ppp=far away // fff=very near, as well as stereo headphones output; from there, scale up to the next level with simple single-purpose smartphone wi-fi/cellular APP. That’s it. Each piece would last about 10 minutes. They wouldn’t be written down, but generated anew each time. If you’d like take part the FIREFLY/STARLING Project (pro bono) in any way—whether as coder, mathematician, AI specialist, or APP designer— or offer support in any way, you can email me at [email protected] AVIAN soloist: An Oregon Rock Wren, recorded in MAY IMAGE: Whitebark pine atop EASGLE CAP (Lost) GLACIER PEAK in background
NOW PLAYING
FIREFLY / STARLING Project Proposal:—Algorithmic ‘SWARM’ spatial composition played LIVE!
No transcript for this episode yet
Similar Episodes
Jul 17, 2026 ·30m
Jul 10, 2026 ·26m
Jul 6, 2026 ·18m
Jul 3, 2026 ·58m