Lighting should respond to intent.
We build tools that let designers describe what light should do and play it on the fixtures they already have.
Light is a language.
Lighting designers already give direction in words like “warmer”, “build”, “drop” and “slower”. Lightshaped turns that direction into cues for the mapped rig. You can review and edit each cue before the engine plays it.
Playback has its own engine.
The model writes the show and the engine renders it. Playback runs independently of the model. Run last Friday's show on the same rig and it replays frame for frame.
Verify before going live.
Strobes are capped at 18 Hz. A failed render triggers a blackout. Effects that can trigger photosensitive reactions require safeguards.
Start with the space.
Every installation is different. Fixture placement shapes how a cue looks. The workspace keeps measured, inferred, designed and verified details separate.
Keep the work.
Rigs are JSON. Shows are JSON. Inspect, version and replay them in Lightshaped or in your own pipeline. The work is yours.
Build on the installed rig.
The fixtures, drivers and pixel controllers already installed around the world are the proven last mile. Lightshaped is the control plane above them: it should coexist, earn trust on bounded universes and only own output after commissioning.
Three things finally lined up.
Language can now meet a real rig model.
Lightshaped gives the model explicit geometry, fixture roles, patching and show vocabulary, then turns its output into validated, editable instructions instead of raw playback data.
Addressable lighting is already in place.
LED tape, DALI bridges and pixel nodes make lighting addressable in buildings and venues. That hardware is already in place at many sites.
Sun-aware, software-first venues.
Architectural installations are run 24/7 now, with schedules and sun events. Static cue stacks can't keep up with that surface area. Generative software can.