grilld wrote:
So there will be no issue when using it for mocap?
That's right, there is no depth channel with PS3 Eye, and it's not needed to capture motion. With PS3 Eye cameras, Mocap Studio creates the 3D data by differencing and comparing the RGB data from each camera. This is why it's critical to have good, even lighting on the performer and to have good background separation with this option.
Kinect sensors, on the other hand, use
only the depth channel to create 3D data, and in this case Mocap Studio doesn't use the RGB channel at all. Generally speaking, lighting and color separation is meaningless for Kinect. (That's not entirely true but lighting considerations for Kinect is different from PS3 Eye, mainly having to do with reflective surfaces abd IR interference from natural sunlight.)
Either method generates a similar point cloud that Mocap Studio uses to animate the skeleton. The quality of the data, with PS3 Eye or Kinect, depends on getting good calibration and optimal device positioning to minimize occlusion problems. (That is, apart from the special requirements for each device type, like clothing, lighting, etc. See the Wiki for more information about this.)
