Forme Cangianti is an experimental live-coded audiovisual composition. This piece is part of my ongoing artistic research aimed at developing a visual identity for computer-generated music. My aim is to distribute the ranking of the visual and auditory components by hierarchically placing the two components on the same level. All (major) sonic events must result in a perceivable visual change and vice versa. To level the music and visual component hierarchy, I grounded my composition on a third paradigm that drives both components: mathematical functions. The performance is structured into two stages. In the first stage, I pre-program a number of mathematical functions that inform the behaviours of both the auditory and the visual components. The auditory component is programmed in SuperCollider, and variables of these functions are mapped into acoustic features of a dozen SynthDefs. The visual component is programmed using GLSL shaders, and the functions’ variables determine shapes’ presence, properties, and relation among them. The second stage occurs at performance time. By live-coding in SuperCollider, I progressively intervene in the underlying mathematical functions by changing some of their variables in real time and enact changes in music and the visual (via OSC messages).