Speak Like Kandinsky
Speak Like Kandinsky is a generative art experience that transforms live voice or uploaded audio into dynamic abstract paintings inspired by Wassily Kandinsky. Users can speak into their microphone or upload audio files and watch a Kandinsky-style composition emerge in real time. Tracks pitch to position shapes vertically, maps volume to shape size, reads timbre to choose between circles and triangles, and responds to speech rate and emotional tone to shift the color palette. Generates museum-quality compositions using circles, concentric rings, arcs, triangles, and geometric lines. Freezes the artwork when the voice goes silent and exports it as a high-resolution PNG.
Created by Tiffany Wang
