I've been fascinated by Project Natal since I first saw the videos from E3 and as a card carrying geek, I've been combing through the videos looking for clues on how the system works.
The marketing videos make a lot of promises: facial recognition, speech recognition, image scanning, object scanning, motion capture, and a slew of other "out there" technologies.
It's easy to stage a demo where you can control a lot of factors such as lighting and even wardrobe, but how would this work in consumers' hands in the real world?
Based on my experiments with the Touchless SDK, the Project Natal had their work cut out for them.
It's not just what the Project Natal device can do, but where it will be expected to do it.
From dimly lit dorm rooms to bright and cheery family rooms, consumers are going to expect the technology to "just work."
As Seen on TV
Right now, it's very hard to separate fact from fiction, hype from FUD about how ready the technology is to come to market, let alone for people to start guessing a release date.
There's a lot of debate on how well the technology works. Project Natal even landed a guest spot on Late Night with Jimmy Fallon, where the technology showed some cracks.
First off, what's with the red jumpsuits? I take it that's not the normal attire for appearing on Jimmy Fallon.
Based on my own experiments with the Touchless SDK, this was probably done to increase the contrast of the person against the background.
This raises some interesting questions (aside from the obvious "how ready is this?")
For one, if you can't control lighting conditions to set up the ideal environment in a TV studio, where the heck can you? Secondly, will the Project Natal documentation recommend you wear something bright and paint your room a nice neutral earthtone?
Sean Malsrom points out a less obvious oddity:
It is clear it doesn’t yet work properly. I have yet to see a Burnout video where people are not driving 50 mph when they should be driving 200 mph, and they’re still crashing into walls all the time.
He has a good point and if you follow the player's movements, sometimes they match up, sometimes they don't.
Sean also has a lot of other commentary and analysis on the buzz around Project Natal.
Now wonder then, that the "breakout demo" has been making the rounds, it's easiest to capture the location of four limbs pointing out in different directions.
So, the real question is now what?
When will this be out? How much will this cost? Will it connect to the XBOX 360 by USB?
If so, could you plug it into your PC can capture the data from it? Where's the SDK? Will there be integration with XNA Game Studio?
There are a lot more questions than answers and I suspect it will be like that for a time to come.