The FFT (Fast Fourier Transform) converts a time-domain signal into its frequency components. The waveform shows amplitude over time; the spectrum shows amplitude over frequency at a single moment; the spectrogram shows how the spectrum evolves over time — frequency on Y, time on X, amplitude as color.

Time becomes space. The same data, a different dimension. When a waveform is extruded, revolved, or helixed through three-dimensional space, we perceive the temporal structure as a spatial object. Sculptors like Anish Kapoor and architects like Zaha Hadid have used sound waveforms as generative forms.

Michel Chion coined "synchresis" — a portmanteau of synchronism and synthesis — to describe the spontaneous welding of sound to image when they occur simultaneously, regardless of any logical connection. Try pairing a cymbal crash with the door opening, then with the bouncing ball. Both feel "right."

In real GPU shaders, uniforms are global variables passed from the CPU each frame. A live performer routes audio analysis values (bass energy, spectral centroid, RMS amplitude) into shader uniforms to make visuals react in real time. This lab shows the same principle using Canvas2D arithmetic.

Richard D. James (Aphex Twin) embedded a self-portrait in the spectrogram of "Windowlicker" (1999) using a technique where pixel brightness at a given frequency-time coordinate produces sound at that frequency. Drawing on a spectrogram is composing in the frequency domain — you sculpt the harmonic structure of sound visually.

Live AV performers structure sets like a score: audio textures, visual behaviors, and parameter mappings change on cue. This timeline models that paradigm — each lane is a separate dimension of expression that can evolve independently, creating counterpoint between sound, image, and their relationship.