VideoGamer.com: Kameo is still considered one of the best looking 360 games, despite the fact it was a launch game.
NB: Yeah. We always said with PD and Kameo that they weren't first generation games, they were second generation in the first generation because as games they were finished, but the architecture and the engines were such, and the support from our shared technology group was such that we could just concentrate in that year on really upping the ante. So like the battlefield, George Andreas, our lead designer on Kameo, he'd always wanted right from the GameCube days, this big expansive battlefield. I don't know if you remember the first ever E3 demo of Kameo on the GameCube, where you could get in this thing and fly over this big battlefield. They said they wanted this big war with thousands of guys and it just wasn't possible. And then George was going, oh I want to do this battlefield if we can, and the art guys said yeah we can do that, it's this ZBrush thing, as far as an art asset's concerned it's big but it's fairly easy. We got all the parallax mapping stuff sorted with the help of ATI, and it was looking really good. We said can we do it? Can we put loads of guys on it?
I said well look, there's this thing we can do now with the Xbox 360, especially because it's got this very fast video memory, it's like an instancing and posturing system where you render something real time, convert it into a sprite and you do that quite a few times, and then munch those sprites up in shaders so they all look quite different. And then, I bet we can render a few hundred of these guys. I got the code going in about a week for this. It's just teapots as you do as a graphics programmer floating around in one of our test levels. And George comes in and there was a cube with loads of these, he said, oh, how many of those are in there? Oh, there's 500 there. He's like, fantastic! So can we have 500 trolls in the battlefield? So I went and talked to Steve Horsburgh, our AI guy, and we worked out how we'd choreograph it from his point of view. Yeah!
So we up it a bit and all right we can do a bit more, we can have 1000. Fantastic! So we put a 1000 trolls down. That's a lot of characters. Anyway, a couple of weeks went by and I was just doing something completely different and I happened to notice in one of our debugging tools I'd made a real nasty bug, in a good way, in the render for these. It meant that it was actually rendering it three times and not once. So there were really 3000 of them but there were two rendered on top of one you could already see. I said George, there are actually 3000 here, and I just looked at the performance and we can have a couple more thousand if we want as well. In the end they said let's leave it at 4000, we'll have 3000 trolls and 1000 guys.
We got these GPU particle systems where the CPU wasn't involved at all. How many particles have you put in that fountain? 50,000. Why? Because I typed it in in the editor. Because we had such strong art and a strong art style, and very colourful... I love blue skies in games, me. I still like Gears and things like that but I'm not too keen on brown games. I like Outrun! To have a nice colourful game like that, I really like that.
We really felt like we were pushing beyond probably what was expected. That was probably a canny decision by Microsoft, to try and choose projects that were fairly far on for Xbox, like PGR3, so they got a lot of bang for buck out of the starting gate. But yeah, I still think Kameo looks great, from an aesthetic point of view as much as from a technical one. But for first generation it was technically impressive. It was fun!
Check back tomorrow for part two of our mammoth interview with Nick Burton, where he talks about the power of the next generation consoles and just exactly what he's working on right now!