The specs are listed here:
http://ipisoft.com/products.phpShort answer: PS3 Eye.
And now the long answer:
The Eye camera is video based (RGB) so technically speaking, the range is as far as it can 'see'. In practical terms, the capture space is 20 x 20 ft or 7 x 7 m. The higher frames rate also helps with precision tracking.
Kinect does not use video, it uses infrared light (IR). The advantage with this is that it can capture actual 3D depth data from a single camera; with two you get 3 dimensional data from two vantage points, making it possible to get full 360 degree coverage with only two cameras. The downside is that the IR range is very short by comparison making the effective capture space for Kinect only about 7 x 7 ft or 2 x 2 m. Also, because Kinect has a lower framerate (30fps), you cannot capture really fast motions as accurately as PS3 Eye (60fps).
I like both systems but prefer Kinect most of the time because it's much easier to set up (plug-and-play) and I can use it in my living room without the need of special lighting or clothing. To use Kinect, I simply set up sticks in the living room, plug the devices into my computer and I'm ready to go.
However, I also keep PS3 Eye available to capture very complex or 'big' motions if I need it. With up to six cameras, you can capture a lot of data for more precise volume reconstruction in Studio. I've seen some iPi users using this setup for very elaborate actions stunts and fight scenes.
However, to use PS3 Eye I would have to shoot outside of my house because it requires a larger environment. Additionally, it requires special clothing, lighting considerations, and moving my computer equipment out of the house, all of which increases the 'hassle' factor quite a bit.
When working on our projects, I try to storyboard most of the shots around what I know we can do with Kinect, and save the tricky stuff for a single 'special' capture sessions if it becomes necessary.
So far, with our Little Green Dog productions, we've been able to capture everything we need with dual Kinect. For you, this depends entirely on how you intend to use the captured data of course.
G.
