Good Kinect article and analysis - haven't seen it posted yet.

[mike_intellivision]

Reasonable said:

ArnoldRimmer said:

The first two pages were just information most people probably already know, but I think page three was indeed interesting. I didn't know Kinect's depth sensor has only 320x240 resolution.

What I found quite ironic by the way was this:

"and giving us enough technical info to banish the somewhat unkind "EyeToy HD" talk that has dogged the internet since E3: Kinect is a state-of-the-art consumer-level piece of motion capture equipment with voice recognition and biometric ID capabilities and Microsoft wants you to know that."

What really sets Kinect apart from the EyeToy-camera is the depth sensor, and the depth sensor is not even used or necessary for voice regognition or biometric ID capabilities. These two could be done with an EyeToy-camera just as well. That's just a matter of software, so using those for proving that Kinect is not just an improved EyeToy is really quite stupid. That's like saying "PS3 is not at all like the 360, because it's got HD graphics!"...

Yes the third page is the most interesting, the first two just give the background detail on the tech.

I'd focused more on the Kinect tech than the software and the SDK so I was most interested to find out about the skeletal system being so tightly coupled to Avatar's and how most titles not using them were in fact writing their own routines for processing the input.

For some reason the article really made me consider how dependant (probably) Kinect is going to be on developers thinking outside the box and specific to the device itself.  I get the feeling MS are in some ways putting more effort into titles for Kinect than they have so far for the 360, which got plenty of titles without much effort from MS themselves.  And if it's going to succeed they're clearly going to have to do so.  Kinect simply isn't going to pick up easy titles the way it looks like Move will judging by the Tiger Woods title.

I have gone from being interested in Kinect to a less interested as the specs have changed and the focus has switched to the titles it has, but the feeling I got was the tech does work within it's own limitations and that so long as developers produce decent code (or design their titles to align to MS SDK) then it should deliver well enough for the titles that suit it - and in fact that's going to be the key to Kinect.  Not providing a better version of the Wiimote (which is really what Move is) nor an alternative approach to existing genres, but something different.

But it's clearly more limited than initially implied, particularly before they dropped the specs, losing a lot of fine guesture sensing capabilities as a result.

If Kinect does find an audience though, then I suspect we'll see Kinect 2 aiming for the specs promised this time round, hooked up to whatever console MS deliver after the 360.

Te text I highlighted above  is a really good point. Since some third-parties have done a pretty good job on the Wii and some cannot even properly tell when you are moving the Wii Remote.  It could end up being the same on Kinect -- and the casuals quickly learn what works and what does not.

 

Mike from Morgantown