Microsoft: Project Natal 'obviously' works with people of all ethnicities

Or you can be a someone who pulls up wikipedia to find a technology in a vain attempt to try and put someone down rather than actually knowing anything about engineering, product design, or imaging. I like how anyone that brings up any technical problem with the system is attacked. I don’t troll. I’ve discussed Natal with my engineer and robotics institute friends from Carnegie Mellon University. We all agree Natal as Microsoft promoting is mostly smoke and mirrors.

If Microsoft is using a TOF camera with respectable resolution then we are years away from seeing a release date. The cameras that can achieve what people are speculating about are over $9,000.00 a piece. A few corporations have manufactured simpler devices that have far less resolution that are a few hundred dollars. Fine for large gesture control but not for sensing minuscule movement. The cheap sensors have a resolution of a few inches. This is because the LED IR emitters required for millimeter precision are extremely expensive. The sensors for the cheap cameras are meant for gesture controls like the NXE demo where you wave your arm to move the stack. The people making these things hope they will make remote controls obsolete by enabling you to change channels by moving your arm left and right and change volume moving up and down.

TOF sensors are easily overwhelmed by ambient light and heat sources. They also can not provide a multiple user interface like Microsoft is promoting in their product vision demo. While the processing power required to read the data from a time of flight camera has been reduced it will still require more processing than the 360 can offer for the job if it is trying to render a game of any complexity. Natal will have to have its own processing unit.

Any IR sensor will get tripped up when an object crosses over another source of IR radiation. A time of flight camera needs a very high contrast image to accurately track movement. Anything that lowers the contrast level will prevent accurate data from being collected. That is why darker objects will not fare as well as light objects.

The prototype may use this type of camera but the press product and specs were based on a combination of camera technologies to make the end product affordable and feasible. All of the current technical documentation says that the RGB camera is used to track motion and the IR camera is there to track depth. The depth calculations are fairly easy implement but the motion is not. That is why the RGB camera is used for that purpose making Natal nothing more than an advanced Eyetoy with a depth sensor. The Playstation Eye has pattern recognition and could be used for facial and voice recognition, that is just software. Add a depth sensor and you get Natal.

With Natal you will get games like the breakout demo, which was very easily tripped up when you cross your limbs or put a hand in front of your body. There is a reason why everyone was playing the game with their arms and legs spread. The burnout demo was very simplistic. The gas was on or off and wild motions sent you straight into the wall. Truly advanced controls like the vision demo are not possible with the unit Microsoft showed at E3.

SO really Natal has a whole lot of problems and technical issues to overcome before it can see the light of day. Far more than Sony’s controller of the Wiimote. It may be ambitious, which is why it is getting a lot of hype, but the reported possibilities just are not within the capabilities of the proposed time period to bring the product to market.

I did not realize how many "doesn't care about black people" jokes were made in the comments. I apologize for adding to them.