Eurogamer is reporting that Microsoft has ironed out one of the kinks with Kinect, with its engineers making changes that will allow the peripheral to recognize players who are sitting. The problem stemmed from the method Microsoft used to create the skeletal models within the software. Kinect would set the base node for the skeletal model at the bottom of the spine, thus confusing the system when players were in a seating position with their knees in front of their pelvis. Microsoft recently changed the base node from its previous position to the back of the neck, thus alleviating the problem.
Blitz Games Studios, a developer well acquainted with the accessory, was on hand to explain the changes.
“It means that should the bottom of your torso get confused with the sofa, because your bum and your legs are enveloped inside the sofa, it doesn’t matter because your hands and arms are still working,” Blitz Games Studios CTO Andrew Oliver explained. “A few months ago they changed stuff around. A lot of developers were like, ‘Oh my God! Everything’s broken,’ because all the nodes were moved. But then it was like, ‘Oh, actually, this is more logical.'”
While previously, developers had to code specifically for the possibility of sitting gamers, this new functionality is now built into the system. Unfortunately these changes will miss most of the peripheral’s launch line-up.
Oliver revealed that Blitz had to code for a number of player positions on its own. These changes had to be made to accommodate Blitz’s The Biggest Loser, which required players to do push-ups and sit-ups.
“We were talking to them last February saying, ‘Are you ever going to fix the libraries so it will work on the floor?’ And they went, ‘Oh come on, that’s lying on the floor. That’s so rare. We’ve got other issues we’re dealing with.’ We said, ‘Okay, we’ll go write it ourselves.'”
by M.H. Williams
original post: http://www.industrygamers.com/news/kinect-can-finally-see-sitting-gamers/