Uncharted 2: Among Thieves Interview, Part Four
I want to be Nathan Drake, a sniper on a tightrope, fingertips that grip better than steel claws, a modern swashbuckler who's parlayed the half-tuck from design band-aid to video game vogue. He's wittier than I'll ever be, far better looking without a shave, and the playable linchpin in Naughty Dog's Uncharted 2: Among Thieves, the PS3-exclusive sequel to acclaimed 2007 action-adventure Uncharted: Drake's Fortune.
Uncharted 2's out on October 13th and the multiplayer demo launches this week, so we caught up with Naughty Dog co-president Evan Wells and creative director Amy Hennig for an over-the-shoulder peek at the design process.
Game On: You've developed Nathan Drake's acrobatic repertoire in Uncharted 2 substantially. He's much more procedural now, much less robotic or spasmodic when you maneuver him. You look at EA's Madden series, and they've grappled with this issue for years and years…that juncture where your ability to control the character leaves off and motion capture for the sake of realism or visual ballet takes over. Having the visual satisfaction of seeing a tackle or lunge or spin execute realistically often comes at the expense of perfect control of the character. You seem to have the balance just about right here. How'd you pull it off?
Amy Hennig: It's a constant challenge, and I think you expressed it beautifully in that you have to be constantly aware that there's this line between the sort of non-interactive parts of the animation, which are there for beauty and responsiveness, and there's a lot of games that fail at either one aspect or the other. Part of it is they may not have the complex procedural blended animation system that we developed, which is really our saving grace in this case. There are games where the animation is beautiful but you don't actually feel like you're controlling them because you can't tell that your inputs are being respected. Or there's such a delay that it doesn't feel like you're really playing the game.
For us, gameplay always has to come first. If there was something we really wanted to see Nate do, but it was going to feel sluggish or unresponsive, we killed it and came up with a different way to do it. So there are certain cases where we'd love his animation to be even smoother, but it would have sacrificed responsiveness, and that's got to be number one. That's why coming up with this blended animation system was such a priority, and actually an absolute necessity. We couldn't have done the game without it. It's the basis of everything, to be able to say I can be running along and still be loading my gun and still be reacting to the gunfire around me and still sort of launching into my next move. Because some of Nate's moves…you're not aware of it playing the game, but there may be literally 30 animations blended up on top of the one motion he's going through right then and there.
Evan Wells: Yeah, and it was already our number one priority in original Uncharted. To create this animation system that would allow us to blend dozens of animations all at once. We've got these extensive debug options and you can bring up the animation tree on the screen and sometimes you literally can't fit all the animations on one screen. And that's just Drake. All of our characters have the same abilities, so any single frame that you take from the game might have about 100 animations in play. And it's all to honor the gameplay, to make sure the player feels in control the entire time.
One of our imperatives in Uncharted 2 was to create these big over-the-top set piece moments, these action sequences pulled out of a summer blockbuster, like Drake being caught in a building as it's being bombarded by helicopter missiles, and it starts to collapse and fall. We wanted that not to be a cut scene. We wanted you to be in control and playing that moment, and so that required Drake be able to, as he's running, be stumbling, because the building's been rocked by a missile, and yet while you're doing that, you might be diving into cover and reloading your gun and flinching from a gunshot all at once. Our programmers are just out of this world. We're really spoiled by having the best in the industry. Ordinarily a game designer's coming to them asking for a feature or a request to do something crazy in the game, and the programmers cringe and say "No, that's technically impossible." Our programmers love the challenge. They'll dive in and say "Okay, let's do it--it's crazy, but let's do it."
GO: Uncharted 2 has this sense of visual proportionality to it that graphically advanced games like Crysis don't. You can chop down trees in Crysis with bullets, but it doesn't service the gameplay, it's just there to show off the physics. In Uncharted and Uncharted 2, you don't have ridiculously hyper-reflective floors or walls or ground that looks like someone poured lacquer over it. You have realism, but it's obviously stylized.
EW: I think a lot of people would be pretty surprised at the level of research and detail that goes into creating our own authentic look. Our artists conduct tons of research online, looking at real locales and such, and our concept artists take that and build on it, even try to exaggerate a little bit, to make it larger than life, because this is an action-adventure. The meticulous work that goes into recreating those environments is staggering. I'm sometimes shocked at how much detail they put into the levels, and it's the subtle detail, as you're pointing out, it's not the over the top, in-your-face specularity, just to push the graphics chip as far as it can go. It's the subtle stuff, the nuances, and really recreating that authentic look of the environment.
AH: Not only that, it's trying to find the sweet spot that your own particular look. I think you could probably flip through a magazine and spot an Uncharted screenshot, even if it was unmarked. I think there's a look to our games, without being overly stylized, that's distinctively its own. And I think it's because we're starting from a realistic base, then giving the game a slight stylized approach, and I mean the characters as well as the environments.
Some of that's in the way we use color, something that's recently seemed unusual in games, because everybody's been de-saturating for some reason. It's just like a little bit of exaggeration in our game, and I think that in the case of our characters, that's part of what helps us avoid the uncanny valley everybody likes to talk about. If you try to make something too much a simulation of reality, whether it's your characters or the environment itself, there's something about it that's dead. Something that sort of falls flat. You just have to pump it up a bit.
So for instance, I don't know if everybody realizes this, but if you look at our cinematics, yes, everything's performance-capture or motion-capture. But we don't capture face. We have four cameras on the stage when we do our cinematics, one master shot and then a close-up on each of the actors so that we have it for reference, and then the animators do all of that facial animation by hand. The reason is, not only is it letting them use some of their traditional skills, as opposed to just processing motion-capture data, but the effect is better than facial capture. I think you have a lot of people in this industry that are so seduced by the idea that they could literally simulate reality by either a facial capture or scan or something, that they don't realize they're not actually getting an attractive result.
It's like what you were saying, Matt. It's like yeah, it's technically efficient and impressive, but it leaves you feeling cold. I think you need that little bit of exaggeration which the characters or the animation or the artwork or the colors or something that says "That's of this world," which is slightly more fantastical than our world.
GO: You know Introversion? The guys that made Darwinia? They did this Wargames movie homage called DEFCON that used stylized vector graphics and that looked a bit like the sequence in the movie where the computer's simulating global thermonuclear war on these huge screens. The game was stunning beautiful, and without using a single texture or shader. I raise the point to suggest we've got visual appreciation in gaming largely backwards.
AH: Well like you said, it's intellectually impressive, right? But you're not emotionally moved. You're not aesthetically moved at all. I guess, and I've been saying this for awhile, I think there's a seductive aspect in our industry to saying, "How close can we come to absolutely simulating reality?" You even see this in threads on message boards, or people throw up different screenshots of games and say "Well no, this is more impressive." Why everything has to be a versus argument I have no idea, but I think we're just moving past that now. I think we've gotten to a point with the hardware, that I think the argument about "How real is real?" is becoming irrelevant. I think people are more excited to see what's being done artistically in the medium.
I mean this has been going on for a while now, right? We've had the last couple hardware generations with people saying "Well look at this water," or "No, look at this grass," and yeah, there's a certain amount of excitement to saying "Wow, look how we pulled that off," and "Did you see how we did the specularity and subsurface?" That's great, but not if it doesn't contribute to some overall goal. I think the challenge is to say, "Why are you doing those things?" You know, why are you having your programmers working for months on something? What's the end goal? For us, with the blended animation system, the goal wasn't to say look how clever we are, it was to say this is why Nate's going to be a relatable character. This is why you're going to feel for him. It's because he's more human and flawed and vulnerable than most video game characters.
We spent months, for instance, trying to figure out how we were going to render the snow, but because it crucially informs the experience. It's because we want you to believe Nate's really there, and cold, and vulnerable.
GO: Since you mention the snow, I have to backstep slightly on everything I just said about indifference to graphics, because when I walked through the snow at the beginning of Uncharted 2 and saw it was actually piling up around Nate's legs and leaving a trail, I had my little geek moment. Everyone else does snow like it's just shiny hard-pack.
AH: Yeah, and we do spend a bunch of time on that stuff, like worrying in the first game about how wet Nate could get, and would the water only go up to his knees, and how would we do that. Or in Uncharted 2, he's marching through the snow, so it should gather more around his ankles, and if he rolls through it, you want to see the accretion of the frost on his clothing. I mean yeah, we geek out on that stuff too because it's fun, but the whole point is to say how authentic does this feel? Emotionally authentic, rather than just visually authentic so you're not drawn out of the experience. We always say we've done our job right if you're not aware how hard it was and how many months we spent on a certain thing. If we've done it right, it's invisible, and you're not pulled out of the experience even subliminally.
GO: Let's hit social networking. In Uncharted 2 you can send progress updates from the game to your Twitter account. Things like chapters completed, bonuses discovered, and so forth. In the last day or so, however, you've said you're going to disable those Twitter updates in response to concerns about their frequency. It occurred to me that it's also partially the limits of Twitter, right? It's essentially brute force broadcasting--all or nothing.
AH: Yeah, it's just one big pipe with everybody's stuff coming through.
EW: I do think it was an oversight on our part though. We have several events that we allow people to mark for Twitter updates, and on all except the chapter updates, we have frequency limiters so that we wouldn't end up spamming people's feeds. So yeah, it was an oversight on our part, so we took it down and now we're working on a patch to put a frequency limiter on the chapter updates just like the others. An update or two a day is bearable, as opposed to a chapter every hour or two.
AH: Especially when you've got a game that we keep hearing people just want to sit down and play through in one long stretch. That could get ridiculous when the game comes out. Imagine Facebook becoming "So and so just finished chapter one."
GO: It is a little intimidating though, because you're looking at some of these updates, and it's like there's 10 minutes between a chapter, and you're thinking "Wow, this guy's really booking." I'm a dawdler when I play. I take hours to go through even short chapters, just to poke my nose in every corner and see what's there, or how it was done.
AH: I think the other thing is, it could give people the impression that the game is short, when it's not at all. I don't know whether it's a fluke of the way the game sends reports, but even I was looking at some of the gamer reviewers' Twitter feeds and thinking "There's no way this person just finished that." We have to get it fixed, so we don't give people the wrong impression.
EW: That player could've been playing on easy. You never know. [Laughs]
AH: [Laughs] Very easy.
GO: Okay, last question. Uncharted for the PSP?
EW: It's certainly something I'd like to see, given the fact that we're encouraging the franchise to grow beyond the PS3. We've already got the movie deal in production, so yeah, it'd be nice to see someone do a PSP version. I can say straight up that we aren't developing for the PSP, we're focusing entirely on the PS3 right now, but that's not to exclude the possibility of finding a partner to work with somewhere in the future.
GO: Really-actually last question. Without giving anything away, Uncharted 2 leaves the door open, right? It's not the end of the franchise?
EW: Each episode is meant to be standalone, and I think there's lots of different adventures and places we could still send Drake.
AH: Yeah, absolutely. We knew setting off with this franchise that we didn't necessarily want to visualize it as a fixed number of episodes, and then it'd be done. As Evan said, they're standalone adventures, but with enough connection between them that for people who are playing all the games, they can enjoy the extra background. But each can be played individually. You don't have to play the first before the second to fully enjoy either.
GO: Thanks Evan and Amy.
Follow me on Twitter @game_on