BS: How precise can it actually be? Because I know that even on the game development side, the cameras that can do motion capture of live actors that don't have nodes all over their body -- that's still very much in a developing stage. So to premier this at a consumer level seems like you may have to sacrifice some precision to get it out there.

SK: It's amazing fidelity that we already have. Hopefully, you got a sense of that up on the stage. I mean, with Ricochet, Abby doing the Ricochet demo. I mean, her avatar is responding with very little to no latency, right? And you're able to play that... That's a game that even I can do, get in there any enjoy right away. We should try to let you guys get some hands on with Natal as possible, but you'll see for yourself that it's very, very real.

Again, it would be easier if we were just talking about two points of motion, but we're doing full-body, full-scale tracking. And the power of the software that goes behind that, you're absolutely right to point out that it's a challenge, but that's the advantage we have as a software company.

KG: I heard that you guys are demoing Burnout Paradise with it.

SK: We have that hooked up, and I think that that's part of what we have here. By the way, Kris, that was the first time I interacted with Natal. And, you know, for me, driving that game was as natural as using a steering wheel or a control.

KG: So it drove like an invisible car, right? [laughs]

SK: I tell you what, it's less intimidating for some people to be out there like this than holding a controller in their hands, too.

KG: Sometimes, I don't buy that people are just incredibly intimidated by controllers. I mean, obviously, it's more accessible to punch, and stuff like that.

SK: No, but this is more than just motion control, too; it's voice. So, it's very natural... It's obviously pretty natural to speak, to speak commands, and things like that, to have a larger vocabulary than just be limited to... You know, remember Mass Effect, right? People thought the dialogue system there was pretty incredible. We just expanded it, we simplified it, and it was pretty incredible.

But this goes well behind that, where you're having a natural language conversation with a character or giving commands to something, speaking to a pet, you know, right? It's not just about the motions -- because I do think that that's a big deal, especially the fact that we can track the entire body. But the fact that now I can go in, and it will recognize who I am, and I can speak; that's as natural as it gets.

KG: Did that voice recognition tech, that didn't come as part of the claimed 3DV acquisition, right? That's separate, internal...

SK: You know, again, being part of Microsoft is a big competitive advantage for us, because we can take advantage of all that R&D work that's happening at other parts of the company, including Microsoft Research. We've had a long-standing focus on natural user interface, including voice recognition, so, yeah, we get to integrate all of that stuff and take advantage of it. That's why we can get further ahead faster.

BS: How's it going to deal with multiple bodies? Because that's really tough.

SK: Again, that's obviously part of the work that we still have to do, but we can already support multiple people today. That's, again, all about the software. The software, the processing software, how we can extrapolate the data, so you hand can cross in front of each other, go behind him. So, we're extrapolating your body. We know where your shoulder is, we know where your hand is, where your elbow is. That's just math.