Building Uncharted 3
Richard Lemarchand describes the delicate balance of technology and creativity behind Naughty Dog's latest
As 2009 drew to close, Naughty Dog was faced with a problem.
Not a harsh problem, or a disturbing problem. Not the sort of problem that causes a company to come crashing down, leading to cancelled projects and heavy redundancies. No, this was a rare sort of problem that only a small handful of studios ever have the right mix of talent and good fortune to tackle: Uncharted 2: Among Thieves had been released to a shrill cacophony of critical adulation and ringing cash registers. It was being lauded as a new milestone in interactive spectacle and storytelling, even as it was selling millions of units to adoring PlayStation fans.
The problem? Naughty Dog had to make another one.
So, yes, a problem. An enviable, privileged problem, and one that Naughty Dog's gregarious lead designer Richard Lemarchand doesn't mind pondering one little bit.
"The success of Uncharted 2 was something that was unparalleled in any of our experience," he says, smiling, a compelling fusion of utter sincerity and stating-the-bleeding-obvious.
"I think really each one of us has reacted to it differently. We all felt the pressure to rise to the challenge of making something as good as or better than Uncharted 2. But my philosophy has been that if you spend your time doing whatever kind of work you do, really focused [and] not working towards any kind of secondary goals like praise or awards or anything, but just trying to do the best that you can in your craft, that is the real key to success."
We came to it with what Zen Buddhists call the 'Beginner's Mind' - not resting on any laurels, not making any assumptions
According to Lemarchand, this was the approach that allowed Naughty Dog to make the quantum leap from the polished and likeable Uncharted: Drake's Fortune to the frequently astounding Among Thieves.
"We just set to it and applied ourselves with gusto. We came to it with what Zen Buddhists call the 'Beginner's Mind' - not resting on any laurels, not making any assumptions."
The suggestion that there was no room for improvement goes too far - melee combat and multiplayer were just two of a number of key areas that Naughty Dog immediately highlighted for refinement - but so too does the notion that Uncharted 3: Drake's Deception required another giant step forward. Nobody could reasonably expect another revolution, and, excessive hand-wringing over review scores aside, nobody who was swept along by Uncharted 2's artful rollercoaster ride would have wanted it anyway.
The fact is that Drake's Deception is the product of a studio at the very top of its game: pushing the limits of familiar hardware, finding new depths in fully realised characters, finding innovative applications for techniques that it invented and very nearly perfected in the space of a single game. Naughty Dog is bigger now than at the start of Among Thieves, but it is also better prepared, and has more resources at its disposal.
Take its "lovely, beautifully designed" new offices, complete with custom motion-capture studio, for example. The bulk of the motion-capture work is still completed on the Sony lot in Culver City - "It's actually the same lot where Gone with the Wind and the Wizard of Oz were filmed" - but Drake's Deception's animation team is now able to try out new ideas and "experiment" with more freedom.
The evidence of this improved attention to detail is apparent in the game, with long, varied contextual animations streamed directly from the disc.
"We have idle animations in Uncharted 3 that are over 12 minutes now," says Lemarchand. "We can stream them, we don't need lots of memory, we just need the disc access time. We also just put in a lot of extra polish onto Drake and other characters. He'll reach out and touch walls as he passes them, which is something that we tried not to overdo, but it does further the naturalistic look his animation has and sell him as a real human being as you steer him through the environments."
Indeed, one of the more widely reported problems with Among Thieves was a knock-on effect of its lush environments: in situations where players had to identify the correct route for Drake to traverse the sheer amount of detail proved confusing, drawing the eye to random objects and causing stretches of trial-and-error experimentation and frustrating dead-ends. This is exactly the sort of minor flaws that Naughty Dog has defined and eradicated for Drake's Deception.
"I'm not quite sure who it was - either our lead gameplay programmer Travis McIntosh or it might have been one of our lead artists, Teagan Morrison - they came up with the idea for a system whereby we could find out from the many play-tests we conducted for Uncharted 3 exactly where people were mistaking a piece of environmental detail for a ledge or surface that they could climb," says Lemarchand. "Every time that someone tried to jump up to something and didn't succeed, we recorded what we called a 'bad jump point' and recorded it in space with exact detail. We can tell where the player is and their progression when they made these bad jumps. Then we could see those bad jumps in the game engine."
"So let's say that there's a pipe on a part of the wall that isn't a climbing point forward and there's a cluster of red dots underneath; we would know to raise that pipe up out of reach or take it out altogether or model it in some different way. It was such a simple thing to do...but it made a positive difference to the quality of the game."
This sort of subtle innovation typifies Naughty Dog's approach to Drake's Deception. The studio made so many significant technological advances for Among Thieves - from rendering and post-processing effects to the Dynamic Object Traversal System that gave rise to its collapsing hotel and cargo train set-pieces - that a single game couldn't possibly exhaust their potential.
And as one person has an idea that demands a new system or the improvement of an existing one, those changes will inspire yet another daring idea in someone else. Lemarchand describes it as a "leapfrogging effect," and it's how many of Drake's Deception's most significant challenges were born, from its desert setting to a brilliant sequence onboard a capsizing cruise-liner.
Naughty Dog extended the capability of its Dynamic Object Traversal System beyond animated objects to objects driven by physics. "This created a whole bunch of new possibilities," says Lemarchand.
Drake's Deception's cruise-ship level is the manifestation of several of those possibilities: Nathan Drake on the deck of a vast boat, reacting dynamically to its movements, which are in turn dynamic reactions to the movements of the roiling, storm-lashed ocean. The deck is scattered with physics objects that slip and slide across its surface. There are secondary bodies of water on the ship itself that move independently from the surrounding ocean. At first sight, it takes the brain a few minutes to process what Naughty Dog has achieved.
They committed to do it and they did do it. It's amazing. I can say this because it's not my level, but I think it's a landmark
"That level was designed by my friend of co-lead designer Jacob Minkoff, who told me last week that when he came up with the idea for the cruise ship and when we pulled together a group of programmers and technical artists to pitch the idea to them, they initially looked at him as though he'd gone stark raving mad," says Lemarchand.
"Perhaps he had! You have to be seized by a little bit of madness to be so audacious as to suggest that in 2009 - when he first came up with the idea - that there's a ship floating around on a virtual ocean whose hold would then be compromised and which would flood with water which would cause the ship to tip onto its side and begin to sink."
"Of course, we have to run that dynamic water system inside the ship as well as around it, and Jacob says - and I think this is to his very great credit - that having looked at him like he was mad, the programmers then proceeded to talk together for 20 minutes and came back to him and said, 'All right, let's go for it.'"
"They committed to do it and they did do it. It's amazing. I can say this because it's not my level, but I think it's a landmark."
Despite the "pain" of creating Among Thieves' barrage of bravura set-pieces, and the tireless efforts of "scripting geniuses" like Kurt Margenau and Jacob Minkoff, Lemarchand believes that Naughty Dog has "grasped the nettle and gripped it even harder" for Drake's Fortune. The Uncharted team understands that audacious sequences like the collapsing hotel, the cargo train, and the cruise ship are why people queue up at midnight to buy their games. They are as much a part of the series' identity as the dusty relics and smart-mouthed dialogue.
As handy as it must be to have a motion-capture studio in the next room, Naughty Dog's best work is far more impressive than the sum of its considerable resources. It exists at the point where technology and creativity intersect; it would not be possible without the former, but the importance of the latter cannot be understated - or, to the chagrin of its competitors, imitated.
Lemarchand believes that, in the future, that balance will tip towards creativity, and the only limit games will face is the far reaches of human imagination. For now, though, mastery of technology is a core part of Naughty Dog's identity, and there is still plenty of room for progress.
"I think we have a ways to go in terms of graphical technology," Lemarchand says, "in terms of more sophisticated skin-shaders and shaders for other kinds of material. Translucent materials are still proving to be a problem for us. It's very difficult for us to make something that looks distinctively like a piece of glass rather than a piece of clear plastic, for instance."
"In terms of physical, dynamic simulations, we can keep grinding that out for quite some time. We do a very good job of making great-looking fire and smoke in Uncharted 3 - and I hope the visual effects guys won't mind me saying this, because it's amazing how good the world looks in Uncharted 3 - but I'm sure they have a million and one new ideas for making dynamic fire simulations that'll be even more impressive than we've already done, that would be even more processor intensive than what we've already done."
"I think that we are rapidly approaching the point where ideas are more important than the technology but I don't think we're quite there yet."