VisionMOD

Vision MOD is a real-time visual synthesizer for macOS. It takes the Complex Oscillator idea from modular synthesis into the visual domain: one source supplies the motion, another supplies the image, and one modulates the other into something new.

SIGNAL FLOW

The picture is a chain of modules. Fundamental Vision sits at the head as the source: a built-in or external camera or a video file, with background removal. After it you add, reorder, bypass and remove modules freely. The same module can be added more than once, each instance with its own settings. Every module has a mix control that blends it with the signal coming in.

MODULES

Stable Diffusion: image-to-image generation on the live picture, running entirely on your Mac. Stable Diffusion 1.5 with ControlNet keeps the output attached to the contours of the camera image. A Structure control sets how strictly it follows them. No cloud, no account, no upload.

Chromascope: hue-driven colour separation and recombination.

Minimalize: rebuilds the picture from a reduced set of shapes, in three tessellation modes.

3D Object: signed-distance-field shapes rendered by raymarching, with the picture mapped onto the surface.

3D Mesh: 100 built-in shapes and OBJ import, rotated, scaled and displaced by the picture.

Particle: the picture dissolved into a 3D particle field driven by audio and motion.

Flow Warp: optical-flow displacement.

Feedback: the output fed back into itself.

Discovery signals