Metamorphosis

Real-time audio visualisation with TouchDesigner

A butterfly-shaped skeletal model becomes a continuously generated, disturbed, and returning particle field. Analysed, amplified, and smoothed audio drives turbulence and model scale. A liquid background, feedback, and bloom establish an organic audiovisual relationship in which form changes with sound intensity.

My contribution: Geometry processing, node programming, audio mappings, and visual compositing
Project stage: Geometry, particle generation, audio response, and live output

Sound and transformation

Let sound shape growth, disturbance, and return.

A SOP-to-POP workflow turns the butterfly-shaped skeletal geometry into a particle field. Noise, feedback, and audio intensity move the particles between a recognisable figure and dispersed fragments.

SOP → POP
Geometry into particles
Import and sample the butterfly model, emit particles, and build a feedback loop.
CHOP
Sound becomes control
RMS analysis, gain, and filtering turn live sound into a control signal.
TOP / COMP
Compose audiovisual depth
Displacement, feedback, bloom, and compositing combine the model and liquid backdrop.

Complete original visual combining the liquid background, butterfly, and particle dispersion — Original TouchDesigner output

TouchDesigner operator network

The operator network separates model import, particle generation, audio control, base-model rendering, particle overlays, background geometry, and image distortion. Shared control signals connect the audio and visual branches.

Audio Device In→Analyze / RMS→Math→Filter→Noise / Scale

Audio Device In supplies live sound. Analyze extracts RMS energy, Math applies a gain of 5, and Filter smooths it with a width of 0.22. The processed level controls Noise POP amplitude and the model scale through 1 + abs(audio). A separate Noise CHOP drives Displace TOP for continuous liquid distortion.

Experience the audio interaction

Choose an audio input to explore model scale, particle disturbance, and background response. Try a rhythm, enable the microphone, or upload local audio. Higher intensity produces stronger dispersion and glow.

METAMORPHOSISLIVE AUDIO / PARTICLES

Audio level0.00

SensitivityPlayback volume

Drag to orbit · Microphone and audio are processed on your device

Loads when in view

Analyze

Extract audio energy

RMS

Math

Amplify the control range

gain = 5

Filter

Smooth the response

width = 0.22

Noise POP

Disturb particle positions

amp = audio

Transform

Let the model breathe with sound

scale = 1 + abs(audio)

This browser experience uses the original butterfly model and liquid background to recreate the sound mappings, particle disturbance, and glow. The film below shows the actual output of the TouchDesigner project.

From model to audiovisual system

The development moves from model import through particle generation and feedback to model–particle synthesis, liquid-background integration, and final compositing. A stable outline supports audio-driven variation while the background adds depth.

Complete original five-stage development collage: model import, particles, model synthesis, liquid background, and final composition — Five stages: import → particles → synthesis → liquid background → final composition

Original butterfly geometry and particle-field working view

Live response and moving image

The live image shifts between stability, slight disturbance, and broad dispersion. Feedback retains motion traces, Bloom strengthens local glow, and Composite joins the particles, model, and background.

Complete eight-frame collage of audio-driven dispersion and glow — The complete original arrangement of eight visual states.

Final live output and its operator connections

Connecting sound, geometry, and image in real time

SOP, POP, CHOP, TOP, and COMP collaborate to turn audio signals into continuous visual behaviour. The project demonstrates geometry processing, node-based creative coding, live audio analysis, and audiovisual compositing.

TouchDesigner / SOP / POP / CHOP / TOP / COMP