How it works
CoMotionPlay turns a room full of people into a single game controller. One camera watches the audience, the software reads how the crowd is moving or holding itself, and that becomes one clean input driving a game on the big screen.
Your audience installs nothing. No app, no QR code, no phone out of a pocket. They just move. You run a macOS application on one Mac — no server, no rack, no infrastructure beyond the machine you already bring to a show.
Two detection modes
A still crowd and a moving crowd are different problems, so there are two ways of reading the room.
Movement
Measures frame-to-frame change. If the room is jumping, waving or dancing, movement mode reads it instantly and responds hard. Its limitation is honest: a crowd that goes still reads as nothing at all. That is fine for a game about energy, and useless for a game about holding a position.
Posture
Subtracts a captured reference of the empty room, so what remains is the shape of the people in it. Because it measures a state rather than a change, it holds while the crowd holds a pose. That is what makes stand-and-crouch work at all, and it is the mode most titles use.
Calibration is captured, not typed
Every room is a different shape, at a different distance, with a different camera. So the range is measured from the actual people who are actually there.
Ask the room to crouch, and capture. Ask them to stand, and capture. That is the calibration. There are manual sliders as an override, but the captured range is what you should be using — a number typed into a box is a guess about a room you are standing in.
Lock the exposure
This is the single most important thing on the page, and the one most likely to catch you out.
Both detection modes assume the only thing changing in the picture is the crowd. If the camera’s exposure is on automatic, then every time a lighting desk pushes a look, the whole frame changes brightness — and the software reads that as the audience moving. In practice, that means the lighting operator is playing your game.
Lock exposure, white balance and focus before you calibrate. If your camera cannot lock exposure, that camera is not suitable.
Camera placement, honestly
Placement matters more than camera quality, and we would rather tell you where the limits are than let you find them at a show.
The reliable setup is a camera mounted high and reasonably level with the crowd, seeing the whole seating area, with the analysis cropped to where people actually are.
A low camera, close to the front row, is the hard case. From down there, a crouching person and a standing person occupy a surprisingly similar amount of frame — in our own testing the two states came out only about 28% apart. A window that narrow makes calibration touchy: a small movement swings the game across its full range, and no amount of filtering fixes a signal that was never wide enough to begin with. Get the camera up, and the problem largely goes away.
Response
One dial controls responsiveness, driving both the camera’s adaptive filter and the physics of the thing on screen. Turn it up and the game snaps to the room, which feels immediate but amplifies noise. Turn it down and it glides, which looks smoother but can feel like the room is not being heard.
For a first show, start in the middle and adjust once you have seen the room react. The right value is a property of the room, not a default.
What you need
- A Mac to run the software
- A camera that can lock exposure, mounted high with a clear view of the audience
- A screen or projector the audience can see
- Enough light on the crowd for the camera to see shapes — this does not work in the dark
That is the whole rig. If you would rather not own it, talk to us — we can run it at your event, and equipment hire goes through Red Globe.