Darren Aronofsky’s Noah generated a tremendous amount of discussion, and in some circles controversy, in its depiction of the Biblical story of Noah’s Ark and God’s decision to erase his mistakes and cleanse the Earth before starting over again. Biblical action-adventure epic? Psychological thriller? Horror film? We’ll leave those questions to critics, scholars and cranks. Our concern is more earthly, less divine – how was the film made?
Working alongside main visual effects house ILM was LOOK Effects, known for their close relationships with directors such as Aronofsky and Wes Anderson (most recently doing vfx on The Grand Budapest Hotel). Helming their team was veteran vfx supervisor Dan Schrecker, who heads the studio’s New York office. I recently had a chance to talk to him about LOOK’s efforts on the film, from flocking birds to the forbidden fruit at the center of the story of Adam and Eve.
Dan Sarto: So tell me about your work on Noah...the overall project scope, the main sequences and some of the key areas of production.
Dan Schrecker: We produced over 300 shots on the film. At our peak we had over 40 people working on the production. The biggest sequences and shots we did as far as man hours and complexity were all the birds. Using Houdini, we built a new feather system from scratch as well as a new flocking system. We also did environment and set extension work, matte paintings throughout the film as well as a bunch of work in the Garden of Eden.
In particular we did lots of environmental work for the flood. Everything before the flood was shot in Iceland. We needed to make the Earth look a little more scarred and sort of destroyed.
It worked out great, working with ILM, figuring out what work could be broken off from some of the bigger elements. They worked on things like the Watchers, the flood itself and the big battle scenes.
DSa: Tell me about the new feather and flocking systems.
DSc: Basically, we built out a new feather system in Houdini. We did our animation in Maya. We worked back and forth between the two programs, building and caching the animation in Maya, sending that into Houdini to get feathered and then rendering the feathered birds out of Houdini.
It was the same thing for the flocking system, which we also created from scratch in Houdini. Because we had a huge number of birds flying into the ark, we had to figure out a thousand ways for them to avoid the geometry of the ark without having to animate everything by hand. There were thousands and thousands of birds that we shot. We built a finite amount of birds just in terms of the models and then extrapolated out a number of variations, randomly assigning those to create different bird types to give the appearance of an infinite number of birds. Actually, there were a pretty finite number of birds that were modeled in full scale and color.
DSa: So you used a sim system to multiply out the flocking?
DSc: Well, we did for most of the flocking. There are certain shots where we have flocking birds and then the hero bird, which we had to hand animate. So for hero birds where we needed to have as much control as possible over their behavior, we didn’t use a sim. Our animation team ended up doing close to 3800 hand animated pairs of birds in the course of the film.
On the Noah movie website, there is a feature called Ark Experience where you can go inside the ark and see some of the things going on. If you go into the avian deck you can see some of these bird shots and get a sense of the scope of the work we did…forty second long panning shots of flocks of thousands of birds. It was a big task for us.
DSa: Did you spend time onset or just work from plates you were sent?
DSc: No, I was onset. I didn’t go to Iceland, but went onset when the crew got back to New York and the Long Island set. I was onset a bit when they were shooting the exterior of the ark before the flood. Once they started shooting inside the ark I was onset pretty much the whole time while Ben Snow from ILM was running second unit.
So, the exterior battle shots, the heavy mud and Watcher shots that ILM was doing, Ben was outside supervising those while I was inside the ark supervising onset for shots we were doing. It actually worked out really well in that regard.
DSa: Tell me about how you manage the onset dynamic, trying to ensure you get the shots you know you’ll need later for the visual effects, while minimizing the impact you have on the rest of the shoot?
DSc: Well, it’s good working with this crew because I have such a history with them, with Darren [Aronofsky, the director], with Matty [Libatique, the DP] and with Mark [Friedberg, the production designer]. We’ve worked together now on a number of films. So, I can sort of read when it’s okay for me to push and say, “Hey, I need two minutes here guys…I need four minutes, I’ve got to shoot this, can you clear the set…” We do that as little as possible.
You know it’s always sort of a balancing act. How much do I absolutely need this? Because any time I slow down production it’s obviously valuable time for everyone else. So I really have to weigh how crucial my needs are. A lot of times, my onset team is shooting HDR throughout the process. We try to clear the set as fast as we can. We’ve got it down to shooting HDR and getting any other reference footage pretty quickly, where we can do it all in probably under a minute after a take. So the DP would give us a minute or two to shoot what we needed to get done.
It is always a challenge because people are……they finish the shot and they’re ready to move on to the next set up. They want to start taking down lights. They want to start moving set pieces. But they usually let us do our thing. You get into a rhythm where once you’re working with the crew long enough, they understand generally what is needed and they’ll usually acquiesce. And at other times, they won’t.
The AD sometimes will say, “Guys, you can’t do that.” It’s always a balancing act. It varies from director to director, AD to AD and project to project. In fact, on the big effects shows, there are times where you have lots of time to get everything you need and there are other times where the expense of running that day is so great that you don’t everything you need and you just have to wing it. We generally do a pretty good job of knowing what we need and being able to kind of eek it out during the day.
DSa: Did you guys do any of your own previs? Was there any previs done for your shots?
DSc: There was previs done, but we didn’t do it. It was handled by a company called Blind Squirrel. I think that was overseen by Andy Fowler [visual effect supervisor] in the visual effects department. The editor, Andy Weisblum worked on the previs and then it got handed off to us. We used the previs to help us bid stuff out and get a sense what we were doing. The whole bird sequence was prevised out and that was good.
DSa: I would assume that previs is pretty important, from helping to bid the work to helping understand what the director wants for any particular sequence.
DSc: Right. Exactly. But shots still change. You have previs and then you get onset. All of a sudden the shot that had been prevised is set aside. Previs comes in one of two forms. You can do previs that is 100% technically accurate, so you can figure out if that camera will fit through that door on that set. How fast can I move that camera? Then you have the more creative aspect, which is mostly an expanded storyboard where you’re blocking stuff out to figure out your shots. That’s what we did on this film. There were shots that were blocked out in previs, but when we got onset, those shots no longer translated. So they had to get tweaked in different ways. They were always a moving target. But having previs is truly invaluable, certainly for budgeting, planning the scope of shots, things like that.
DSa: What was the scope of the work you did in The Garden of Eden?
DSc: Overall, we did the Adam and Even glowing effect, a matte painting for the Tree of Knowledge, sky replacements to give the sky a celestial vibe and a CG apple that looks like a cross between a pomegranate and a beating heart. We created a CG version of it using references of how a heart beats, the different ventricles, how they pump.
We also created the snake, which was actually two snakes. We did a green snake that shed and revealed the four eyed black snake. We did both of those in Maya with a full CG environment including rocks, macro-fragment soil and grass simulation to interact with the snake as it slithered towards the camera. We modeled the base snake geometry in Maya and did the sculpting and texturing in Mudbox. Look development was done in V-ray for Maya, using vector displacement for scales and surface detail. We created a fast subsurface scattering shade for the skin and used two-sided materials for the peeling skin. For the eyes, we got the black, inky look Darren wanted by combining glossy, refractive shaders with additional subsurface materials.
It’s interesting to look at the whole religious aspect of the film and how that affected our work. The Garden of Eden sequence in the film is where Noah told his children the story of Genesis. Design-wise it was very intense. You’re dealing with Adam and Eve. What did Adam and Eve look like? We had a lot of really interesting design conversations about this sequence, as well as the look of the Garden of Eden and Tree of Life. It’s a cool sequence. It was very different from a lot of what went into the rest of the film.
Getting into some of these religious questions was interesting. We didn’t delve too deeply into it and essentially we followed Darren’s and Ary’s [Handel, Darren’s co-writer and executive producer] lead. That whole side of following the religious story throughout the making of the film was interesting.
DSa: Well, let’s talk about the religious aspect of the film. You knew this was going to be a controversial film just because of the nature of the material. There was no way that everybody was going to be happy and as these things tend to, they become lighting rods for people’s political agendas, good, bad or otherwise.
But from the standpoint of the designs and look development of fundamental images from the Bible such as Adam and Eve, the Garden of Eden, Noah’s Ark, etc., how were those handled? Were you guys working with an art director? Were you working from concept art and other designs? Did you guys have to design this from scratch? What was your role in the visualization?
DSc: There was various concept art that we worked with throughout the project. We would take some concept art, do some work and show it to Darren. Mark Friedberg was also involved in the process. Even with concept art, there are always going to be conversations regarding issues both creative and financial to determine what type of film you want to make. One main idea was depicting a sort of see through Adam and Eve, where they have this translucent quality and magical glow. We were able to discuss those sorts of questions up front, which I think is smart.
The concept art showed different looks of “glowing human forms.” We shot an actor and actress as Adam and Eve and used a 2D effect to give them this otherworldly yellow glow that thematically matched up with the light being emitted from the Watchers’ eyes, which we called the “Tzohar glow.” This is the glowing material that appears in different places within the film including inside the Watchers’ eyes. It provided a sort of divine glow for Adam and Eve.
So we created a number of different versions of what that color would be. For example, the hue of the yellow. How gold would it be? How bright lemony yellow would it be? We knew it was going to be a 2D effect. So, working from the concept art, we tried several variations until we nailed it, going back and forth with Darren until he found something he liked.
Race is always an issue and a question when dealing with Adam and Eve. It was a bit of an issue as well when talking about the film. We wanted to make sure Adam and Eve didn’t read as any particular race. That’s one of the reasons we went with this yellow-gold look.
DSa: I’ve talked to Henrik [Fett, partner and LOOK co-founder] and Mark [Driscoll, president and LOOK co-founder] before about the overall visual effects business climate. This is a big film for you guys. The business climate is tough right now and has been for some time. You guys have always been known as a top boutique studio in a sense, often working on numerous project with the same sets of directors. But this project pushed you not only creatively but as far as the brawn needed to bring to bear on the production.
DSc: Yah. Absolutely. It’s daunting and especially challenging for companies our size, operating in this sort of middle range. Big companies who have done this type scale work before have a great understanding of the process and they have huge machines and facilities to do work like this. Like ILM, who did the principle work on this film.
And then you’ve got smaller teams, four to ten people, who have been operating in small and efficient ways. Companies in the middle like ours, in order to make that jump up, have to take a huge leap up, taking big risks and thinking of new ways to get to that level, which is what we did on this film. We did it quite successfully but it was quite a challenge and a tremendous learning process for all of us.
DSa: Henrik has told me many times, you make your bid, you win your bid, then you put the bid in the drawer and do whatever is needed to get the film done. But on a show like this where you need to do a fair amount of R&D, such as for the feather and flocking systems, you really don’t necessarily know upfront how long that will take to develop. That’s an investment you wouldn’t normally make except in response to the needs of a project.
DSa: You took a big risk.
DSc: Yah, absolutely. Especially if we don’t have more bird projects lined up [laughs].
DSa: [Laughs] Right. We are the bird studio now!
DSc: Yep, we need more bird shots. So, yah, it’s a big investment and not just for the feather system but for the pipeline and everything else. But it was for a great show.
Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.