/// Mocta-X Camera · Concept study

A camera built for motion. Someday.

The Mocta-X Camera is a design study for dedicated capture hardware — purpose-built optics, on-device inference, one mount per room. No unit exists and none is for sale. The software it would run is shipping today as Mocta-X Dashboard.

Mocta-X Camera, a hardware capture camera concept
/// Design targets

What a capture camera would need.

Targets from the design study — none of them measured, because nothing has been built.

Sensor — target

High-frame-rate global-shutter capture, so fast movements are not smeared. Not yet built.

Field of view — target

Wide enough to cover a room from one mount. Not yet built.

Latency — target

Pose extraction on the device itself, with no upload. Not yet built.

Compute — target

An on-board neural accelerator running the same 33-landmark pose model as the software. Not yet built.

The idea

Mount it once. Capture everything.

A fixed camera that covers a room from one mount, so a clinic or an erg room never has to set up a phone. The concept renders and the 3D model on this page are design studies.

Mount it once. Capture everything. screenshot
What it would run

The same pipeline, at a higher frame rate.

The software already grades from a phone clip. Purpose-built optics and a faster shutter would remove the two things a phone does worst: motion blur on fast limbs, and a frame rate the timing metrics cannot always trust.

The same pipeline, at a higher frame rate. screenshot
Where it started

The 3D Lab is the software version.

Two phones and a laptop already reconstruct joints in 3D in the dashboard's 3D Lab (beta, 6.2 mm RMSE on a simulated rig). The Mocta-X Camera is the question of what a dedicated capture device would add. It is research, not a product.

The 3D Lab is the software version. screenshot
/// Explore the design study

Every angle. Every component.

Drag to rotate. An industrial-design study from the Conrad Challenge pitch — sensor placement, enclosure geometry and mount profile are sketches, not a build.

prototype · rev 3
Drag to rotate
Neural compute module

Target: on-device inference with the same 33-landmark pose model, no upload.

150° optical stack

Target: a wide-field global shutter covering a room from one mount.

Single-mount install

Target: wall or ceiling mount, calibrated once.

/// Status

A concept, not a product.

The camera started as the hardware pitch behind Mocta-X. The software outran it: the dashboard and the iPhone app do from a phone what the camera was meant to do, and the 3D Lab covers multi-camera capture. We keep the page as a record of where the idea stands. If you run a clinic or a lab and want to talk about dedicated capture, write to us.

Open the dashboard. Run your first session.

No install, no credit card. The free plan grades clips every day — five images and two videos — and the AI coach is included.

Open Dashboard