Much has been made of Kinect’s inability to recognise players sitting down. That was just the core Microsoft-supplied detection code though. Developers have been able to code their own detection routines so that Kinect will work whether you are standing, sitting or even laying down.
Now according to Blitz Games Studios co-founder and Chief Technical Officer Andrew Oliver, Microsoft have changed the skeletal mapping of the data from Kinect’s sensor so that it will no longer matter if it cannot tell the difference between your ass and a seat cushion.
Speaking to Eurogamer he has revealed that Microsoft have moved the skeletal model’s “base node” from the base of the spine to the back of the neck. The base node is the key reference point for Kinect’s skeletal model so when it became difficult for the system to ‘see’ the base of your spine when sitting because it was lost among cushions or behind your knees that is when Kinect struggled with a seated player.
Now that Kinect’s base node is not obscured when you sit down, “your hands and arms are still working,” says Oliver. “You can do most actions.”
It is a recent change to the development software that Microsoft provide, one that was made too late for the launch titles. However, Oliver says that “games going forward won’t have a problem”.
So has that addressed your main concern with Kinect? Now it only needs to see your upper body, for some applications will you be able to be closer to the TV than four or five feet? Will Kinect Adventures 2 allow you to sit in your plastic boat peripheral and paddle?