early access opens july 1
compose with images
A non-destructive image studio. Layered effect stacks, a full keyframe timeline, and audio that can drive any parameter — pixel sort along vector fields, reaction–diffusion, halftone, sonification, and 35 effect families more, sequenced or animated in any order.
100 founders seats · $149/yr for the first year · waitlist gets first invites
one stack: image, motion, signal
image
Non-destructive layer stacks with groups, blend modes, and masks that are themselves effect stacks. Brush, shape, and text layers. Halftone, risograph, dither, gradient maps, LUTs — finish work, not filter work.
motion
Keyframe any parameter. Bezier easing, one-click motion presets, LFO modulators, seamless loops. Video sits in the stack like any other layer. Export MP4, MOV, WebM, or GIF — the live preview is the export, byte for byte.
signal
Audio drives anything — level, frequency band, or beats, from a file, a video's own soundtrack, or a microphone. Sonification runs the image itself through audio DSP. Generative layers and reaction–diffusion compose from nothing at all.
35 effect families · 140+ sub-modes · gpu-native · macOS + windows
join the waitlist