Pixar's Andrew Dayton
The leading studio's tech director looks at the similarities between games and films
Among some of the interesting speakers delivering talks at next week's Develop Conference in Brighton is Andrew Dayton, technical director at Pixar.
Here, he talks about the similarities between the worlds of games and animation movies, the challenges of story and technology, and why he feels that Pixar and Blizzard are two companies cast from the same mould.
Q: First of all, how do you view the videogames space, through the eyes of somebody working for a company like Pixar?
Andrew Dayton: I honestly see, as the years progress, games are becoming very close to film in the sense that the cost of some of these big games surpass how much a film costs - and also the technology being used... there are times when we, in the film industry, look at some of the techniques that the games industry is using to see if we can utilise it for our processes.
But personally I see a lot of these games becoming epic franchises that are as big, if not larger, than some of the film franchises. You ask many people if they know Shrek or Toy Story, and they will. But at the same time you ask a lot of kids or younger adults if they know World of Warcraft, or Diablo... and they're just as popular.
Typologically-speaking I think we all pretty much swim in the same pool, but we all have different specialities. For film our thing is that we have to get super-detailed, hi-res final renders; on the games side it's the same sort of desire, but they're having to do it on the fly, to create them in real-time.
I think that's where the convergence is between the two worlds, but everything is pretty much analogous - you have to have a good story, a good art department, good artists that are executing these things.
Because the competition is stiff in both worlds - for Pixar, we have tonnes of competition, from all the different studios. Everyone is trying to out-do each other, and it's the same thing with games - each game company has to come up with the latest, greatest game with bells and whistles... on top of having a great story.
Q: You mention story there - that's one of the things that many people would point to as needing to evolve, as creators attempt to embed a wider range of emotions into games.
Andrew Dayton: Absolutely - there will always be a special place in my heart for the first-person shooters where you just go in there and blow other people up. But as I've gotten older I've started to gravitate towards games that are more story-heavy.
Some of the companies that I have a lot of respect for are companies like Bioware - you play Mass Effect or Dragon Age, and those are games that are completely immersive. With a movie you're walking away from a two-hour experience at most - with these games you're walking away with 60-100 hours of immersive gameplay.
I actually think it's harder for game companies - on top of a great story, you have to make it immersive and you have to make it seem like you're not on rails, that you're not just playing the game and following a linear progression - like a film.
You have to have choices, and I think that complexity is where games succeed or fail.
Q: Interactive is the key word, and I don't think anybody underestimates the challenge of presenting an immersive story, plus the illusion of genuine choice. It's almost an impossible task, isn't it?
Andrew Dayton: It is - for films it's hard enough to come up with an 80-minute story, where you're basically following the characters along an arc. That's hard enough as it is, and it's a linear progression - you go from point A to point B to point C, and the audience doesn't get to change or choose that path.
Once you start creating these immersive games you need to be able to give your player unlimited choices, tie them all together - and then if you're crazy enough to add a multi-player online game you then have to interact with every other player who's going along their own storylines, and exist in the same world.
It's massively intricate, and I think that's why games, to a larger degree than films, are very hit-and-miss. For one successful game, how many bad games are there? There are so many places to fail at in the process of creating a game - the story, the interaction or the technical aspects.
Q: One solution to that problem is the path followed by CCP in Eve Online, which is as much sandbox with player-generated content as possible. Maybe that's the best idea?
Andrew Dayton: Right - the thing is that on paper that probably sounds like a great idea, but the actual construction of it is probably extraordinarily hard. Saying: "Okay, let's just let everybody do their own thing!" is fine, but how do you tie that together so that it's seamless?
It's tough, all the way around. If you talk to some game designers or artists, they're always probably butting their heads against a wall trying to figure out how to create a world that's entertaining, but not just for a couple of hours - something that you want to keep going back to again and again.
Look at what Blizzard's done with World of Warcraft... how many years has that been out?
Q: Over five years now.
Andrew Dayton: Over five years - the shelf life of that game is historic.
Q: And the Cataclysm expansion looks likely to reinvent the experience again.
Andrew Dayton: We've met quite a few of the Blizzard people actually - some of the cinematics team. Blizzard and Pixar almost seem like mirror images of one another: The same sort of philosophy; the same sort of work ethic; the same attention to detail.
They're basically the biggest company in the games industry, and we're the biggest company in the 3D film industry - it's interesting talking to them, because we always come across the same sort of problems and it just comes down to the philosophy of what you're trying to do.
Their philosophy has always been trying to make the greatest game they can make, and for Pixar it's about trying to make the greatest film we can make. It's never an easy process - every film has its own slew of nightmares and production problems, and it is almost like giving birth with tonnes of complications in the process.
Nothing is guaranteed, or goes as planned, but if you keep yourself grounded in what you're trying to do - which is make a piece of artwork that people would love to watch - it touches them. And that's the same thing that Blizzard and Bioware do - they're making games that, when you turn the computer off, you're still thinking about the game. There's an important place for that.
Q: What sort of things do you think Pixar can learn from games companies? You mentioned the real-time element, is that a factor that you're maybe a little bit jealous of?
Andrew Dayton: I think the things we look at generally is what we can use real-time graphics for to help us approximate what the renders will be. Can we use stuff from the GPU to give us a real-time sense of these assets, so we can utilise and quickly modify them?
So that when we poke "Render" - and when we do that, we're talking per frame, that can go anywhere between 3-10 hours, with 24 frames per second we have to visualise... So for us we're looking at what gaming is doing, what they use.
Well, they use the GPU - so what can we get from that sort of technology, from offloading some of our work onto the GPU that we've not been able to do traditionally? It's one of those things where we're not using the technology for the final output, but we are looking at game technology to increase production up to that final output.
Q: A little while ago I spoke to EA's Glenn Entis about the progress in bridging The Uncanny Valley. Some of the best animated film experiences are those which haven't really tried to imitate reality completely, but some would say it's not quite happened yet for the full on experience. Can it ever happen?
Andrew Dayton: It depends on what you're looking for. If you're talking about Pixar, we're a different medium - we're not trying to make it completely realistic. If we did, well s**t - shoot it live! But if you look at what James Cameron did with Avatar, whether you like the story or not, I thought what he did in that film was completely revolutionary.
He took snippets of what WETA did for Lord of the Rings, and even then they had The Uncanny Valley on those films. But once they went into the Avatar world, you saw what he did with motion capture, performance capture and the rigs that he created, he proved without a shadow of a doubt that you can get near photo-realism.
Is that something we would do? No, because we're not doing action adventure live-action films. What we do is more like saying: "Well, wouldn't Cinderella have been more awesome if it was photo-real?" No, probably not, it's a different medium.
Because of that difference from reality to that style, you can do certain things in that medium that you couldn't get away with in live action... Avatar's renderer is Pixar's renderer, it's our software it's rendered on. It just shows you what you can do when you're tasked with the impossible.
This technology is something that's available for us now, but it might be available for gaming further down the road. It might not be something that you can render in real-time yet, but it's possible that there are ways of getting the processors there - it's all a matter of number-crunching at this point. We've proven we can do it, it's just who's going to be the one that comes up and has the Avatar-like game, that completely changes the rules.
Andrew Dayton is technical director at Pixar. Interview by Phil Elliott.