Bringing to life the vision for Disney Animation’s “Big Hero 6” required technology that didn’t exist.
Not a problem. Ever since Pixar legends John Lasseter and Ed Catmull took the reigns at Disney Animation in 2006, reaching for the impossible has been the mandate. For “Big Hero 6,” the Disney team made a leap of faith in a new computer software called Hyperion, named for the street in Los Angeles where Walt and Roy Disney founded their animation studio.
Hyperion, which was developed in house as a “science project,” wasn’t ready when production on the film began. TheWrap caught up with Disney Animation’s Chief Technology Officer Andy Hendrickson and technical supervisor for “Big Hero 6” Hank Driskill for a behind-the-scenes look at the gamble they took to bring the robot Baymax and his world to life.
What is Hyperion?
Andy Hendrickson: It’s a whole new way to think about the lighting problem called global illumination. which starts to simulate the physics of how light behaves in the real world, rather than just painting by light, which is what we used to do before.
Hank Driskill: Hyperion was a science project here. We had been having all kinds of conversations in various forms, including a group that Andy pulls together every couple of weeks called Tech Trust. We kept being more and more ambitious with what we were putting on screen. We were starting to feel constrained by our tool set.
Andy greenlit an experiment. It was missing a lot of bells and whistles that you actually need in a production renderer, but it was just to showcase the idea and what it might be capable of. It came out of that science experiment mode in early 2013. And it was around that phase we were in the middle of ramping up ‘Big Hero 6.’
AH: [With Hyperion] we simulate the physics of light at a base level and we use that to actually to do the first pass of a scene. And it gives you a nice consistent look from a base level. And then you can tweak it and practice photographic artistry on it to make it really, really great. Similar to how you’d set up on a motion picture live action set. You set up a few lights to illuminate the area nicely and then you use some specials in key and fill areas to really set the mood, or the tone or the dynamic of the scene. Practicing more the artistry of lighting than the mechanics of lighting.
HD: We were using RenderMan on previous shows and that effectively was our Plan B. That was going to put limitations on what we were going to be capable of in the film. It was going to change conversations with the directors because just the scope of what they wanted to pursue. It was really ambitious what we all wanted to pursue for the film. But pretty much from January of 2013 to June of 2013 we were operating with a Plan A and a Plan B. Plan A was to continue R&D on what became Hyperion during that time and pursue the possibility of using it for the film. But we were also doing asset production during that time. Every rock, every tree, every bush, every bystander, every in a CG movie is crafted by people. We were building the world piece by piece. And we were actually building the city in this case because we also had this very ambitious desire to build the entire city.
With Hyperion wasn’t not quite ready to go when production began on “Big Hero 6,” talk about the moment you decided to scrap Plan B and commit completely to this new software.
AH: What we looked at was what the promise of what the world could be. We just looked at and said if we have to go that way [Plan B], we’ll just have to scale back what’s in the world. Nobody wants to do that. We want to put something on the screen that makes everyone go ‘wow, I’ve never seen something like that before.’ But there’s realities. Could we do it, could we not? I’d love to say it was coldly calculated and that there was no doubt at all, but its not true. We weren’t sure we could pull it off.
HD: Some people were really excited by the promise. Some people were really anxious about something new and different because it was going to be painful. That anxiety was all the way up and down from the artists on the floor all the way up to the executives. There was anxiety about what this was going to be capable of, whether it would deliver on time, were we making a horrible choice that was going to impact the ability to deliver the movie. There was a big leap of faith on a whole lot of people’s parts to be okay with us trying to pull this off. It was a little bit crazy.
AH: There was a lot of adrenaline associated with that. Basically we got to the point and some of the test images that we had created, we just looked at them. We were so enamored with what we saw in the visuals that we were creating, we said ‘oh my god we have to do this.’
What kind of culture have John Lasseter and Ed Catmull brought to Disney Animation and how did it make taking a risk with Hyperion possible?
AH: It’s great working with those two at the top of our studio. One of the big mantras is ‘everyone’s a filmmaker’ and that’s been really, really positive. We each have a piece to play in this large presentation that we call a motion picture. So there’s story, there’s art, there’s the technology that fuels the art. There’s the art that provides a challenge to the technology. All these are kind of studio philosophies given to us by Ed and John. We really practice them, they’re not just some words on a page. We actually do practice a lot of these things every day. We’re constantly being challenged by ‘hey what if we could do this in the art.’ And everybody says ‘there’s no possible way to do that today.’ Ok, well we don’t care about today, how are we going to do it tomorrow? So that’s kind of a thing we do everyday, figuring out new ways to do things that quite possibly weren’t not possible five years ago.
It’s Ed and John’s license and mandate to invent the future. That’s what we’re doing here. And we’re doing it every week. Every day of every week, trying to invent what the future is in animation and in motion picture. And it’s really great to have them here driving our studio.
HD: I’ve been here 17 years, so I certainly know before they came and once they came and the differences. As do we all. They’ve created a director-driven studio where we really do rally behind the directors and their vision. It isn’t as encumbered by the voices of outside forces who drag the movie away from what the director might want to make. That’s a very wonderful thing from a creative standpoint. From a technical standpoint the biggest thing is that they really push us to be a culture of innovation. They push us to be trying all the time to think new thoughts and think of better ways to do things. They encourage us both down here and at Pixar to chase ideas even if those ideas go in two different directions. We have a really nice communication path between the two studios, but we’re not obliged to follow the same path. It let’s us experiment and let’s us try things like Hyperion. And worse case maybe it fell on its face and we learned a thing or two. Best case it came out this really cool idea that turned into something amazing. They really encourage us to experiment.
Ed was one of the people who created RenderMan, but we wanted to try something different. It wasn’t necessarily a gate, we didn’t feel that if we didn’t get Ed’s buyoff that we didn’t get to do it because he just isn’t that sort of person. He encourages us to try things out.
Obviously the experiment worked, your film made it to theaters on time. What will audiences notice about this film that wasn’t possible before?
AH: I’ve got to say that there’s two things that really stand out. One is there’s this sequence we call first flight when Hiro first on the back of Baymax and they go flying around the city of San Fransokyo. In motion picture movie terms that’s actually happening at the magic hour, which is right at a sunset, which is some of the more dramatic lighting that comes in live action films. We did the same thing with this particular sequence. And it is mindboggling. So that’s one thing that I think everybody when they see it they’ll recognize it.
AH: There’s another part that I think is really interesting which is Baymax himself. Baymax being made of translucent vinyl, he looks like he’s made of translucent vinyl. That’s not an easy thing to achieve. Because if you can imagine the light bounces around inside that translucent vinyl volume, inside the balloon which is the outer skin of Baymax. And bounces around in there an average of eight to 10 times before it comes back out. We had to simulate that to get him to really be vinyl instead of hard plastic.
HD: The second shot is flying over the port of San Fransokyo sign. It’s the classic port of San Francisco sign re-purposed. You’re flying over this big plaza and there’s over 1,000 cars driving on the roads there. There’s 6,000 crowd characters just in that one shot as well as downtown buildings. Everything is reflecting off everything else.
What’s the next advance in technology you’re working on?
AH: I can’t really say too much, but I can say that there’s this movie called ‘Zootopia.’ It has thousands of characters in it and they’re all furry creatures. So that’s the next challenge in rendering problems. It’s got furry creatures and it’s got a lot of them.