DM235 said:
RGB camera uses the visible spectrum of light. The Natal "depth" camera uses infrared light. Both cameras would use light intensity and contrast to determine where an object is in "3D" space. Whereas the RGB camera can have trouble dealing with variable room lighting, an infrared camera would have trouble with variable infrared lighting (candles, fireplaces, and potentinally you if you work up a sweat). The E3 videos show some lag with Natal, but it is still a beta product. We'll just have to wait and see what Microsoft can do before it launches. |
Actually RGB cameras rely mainly on edge detection for depth, it has nothing to do with light intensity other than the fact that the better the light, the easier it is to detect edges. This can be very difficult to do and can only provide relational depth information. Such as the person got bigger, so they must have moved forward some, now use an algorithm to come up with some distance that was just covered. While each pixel in Natal's system has an associated depth to go with it, and all objects(such as other players) can be exactly related in 3D space to each other. No guess work involved at all.







