Microsoft has revealed a new Natal tech demo in advance of the motion controller's E3 PR storm, showing the efforts of Microsoft's Israeli R&D team at its annual Think Next seminar.
The presentation showed attending delegates the way in which Natal "sees" the world, displaying the infra-red "depth map" that adds the third dimension to the conventional 2D webcam image. During the demonstation, Natal locked on to one, then two human skeletons (overlaid over the image), using facial recognition to determine gender before assigning an appropriate Avatar.
The demo itself, captured by Next Gen News in a shaky cam video, was an intriguing mixture of the PrimeSense system that provides the image and depth map, alongside Microsoft's own technology in scanning the depth map and generating the vector skeletons with which the system tracks player movement.
So far, so good, but nothing that we hadn't seen before during our showdown with the system during last year's gamescom. However, the demo itself seemed to be somewhat slicker than what we saw at our presentation, certainly in the way it picked up the arrival of a second skeleton within the Natal viewing radius.
Engadget additionally reports from an anonymous "tipster in the audience" that the lag issues with the system have been "eliminated". However, judging by a frame-by-frame look at the movie linked above, overall response seems to be in the same ballpark as when we played it last year (thanks to the lady in the video for an impromptu rendition of the patented DF flappy-arm latency test). Big-screen displays like the one used at the event have their own lag issues of course, so the jury's well and truly out until E3.