ah, chears.
Why need willpower and self-controll when you have technology eh?
I share some of your scepticism, but only because it's impossible to tell from pre-recorded video footage of clearly pre-programmed events just how good the technology really is.
The physical interaction such as rippling the water and gesturing to put on the goggles has been done before so we know that works fine.
The voice recognition is interesting and if it really can sense things by just the tone of your voice then that's really cool, however that just sounds like really clever coding (credit to the LH crew ) and nothing that couldn't already be done with existing hardware.
I'm very much undecided. It's something you really need to see and use yourself to be able to make a proper judgement.
And while Natal itself is no doubt a very clever piece of hardware, it couldn't actually do most of the things Sony showed their EyeToy and dildo combination doing and what the Wiimote can already do, without you holding something extra in your hand for it to track. Then the technological gap between it and the other two looks very small indeed so pricing and a killer app will be everything.
Theres much it probably couldn't do that the others can, but thats also quite true vicer-verse.
It will be interesting seeing what tis strengths/weakness's are, and just how good its "resolution" is.
(could it sense figures if they are non-occluded, for example).
I really think Natal would be great at enviromental interaction though, Penumbras basic point/grab physicality of its gameplay would work paticularly well. (dedicting a clasp shouldnt be that hard, Ive seen other demos do it with much more primative set ups)
And yes, its of course hard for us to tell how much of that demo was real.
--
Its interesting to note, Nintendo once patented a speach-to-text system where the tone of your voice would effect the font. They never did anything with it though beyond the concept. Whether thats because they couldn't do it , or because their paranoia over online meant they wouldn't, I dont know.
Id guess senseing volume per-word is easy enough, and that could correlate in some way to the intensity of the emotion. Other ones would be much harder to pick up though.