bad-ASCII

What is bad-ASCII?

bad-ASCII is a browser-based live visual synthesizer designed for VJs and digital artists. It renders live camera or imported video through typography — treating text not as a medium for reading, but as a physical, malleable material that reacts to sound, motion, and human presence in real time.

Live Performance Features

Character rendering styles including prose, ASCII, Chinese characters, binary, halftone dots, and brutalist blocks. Audio-reactive typography that physically responds to bass, mid, and high frequency bands. Optical flow that orients characters in the direction of motion. Velocity physics that makes the entire typographic grid bounce and ring on bass hits. Chromatic aberration that splits RGB channels per audio band. Temporal echo slit-scan. Simplex noise warp fields. Gray-Scott reaction-diffusion seeded by human silhouettes. Matrix rain with video color blending. Concentric wave displacement. Kaleidoscopic mirror with rotational momentum.

Collaborative Multiplayer

Built-in WebSocket relay for real-time collaborative performance. One machine processes the video pipeline while multiple remote collaborators join via a room code and manipulate the visual state from any device, including mobile. Collaborative pad locking prevents parameter conflicts during live sets.

Projector Output

Dedicated projector output tears the canvas into a clean, borderless popup window for venue projectors and second monitors. The control surface stays on your primary display. Canvas recording exports performance as WebM video with audio.

ML-Powered Segmentation

MediaPipe selfie segmentation separates the human subject from the background in real time, enabling spatial masking where the performer is rendered in one typographic style and the background in another. The gradient zone at the subject's edge is a creative parameter, not just a threshold.

Built By

Developed by Aalap Davjekar — new media artist and creative technologist based in Goa, India.