Search form

Roaring Through the Previs on ‘Godzilla’

The Third Floor supervisor Eric Carney walks us through the previs on Gareth Edwards’ hit reboot of the Japanese classic.

Final production image from 'Godzilla.' All images © 2014 Warner Bros. Pictures. Courtesy of Warner Bros. Pictures.

Gareth Edwards' visual effects laden Godzilla was brought to the screen by artists from a range of VFX houses, including MPC, Weta Digital and Double Negative. But supporting this process, getting involved well before the movie’s pre-production began, were the artists at The Third Floor. We caught up with previs supervisor Eric Carney to discuss his involvement with this film, the scope of the work involved, and the role previs can play in big CGI-heavy action films.

Dan Sarto: How did The Third Floor first became involved with this project?

Eric Carney: It would have been around April 2012 that we first started. We were initially contacted by Legendary Pictures, who had a director and a script that they were really excited about. At that stage they wanted to get their partners at Warner Bros. really excited about the project too. So we were asked to work on a kind of show and tell for the studio, which would involve things like artwork and the original Comic-Con trailer. One of the things they wanted to include as part of this was a previs of one sequence from the film.

So we started working with Gareth Edwards [the film’s director] to previsualize the Hawaii battle sequence. This takes place around the middle of the film, and it is the first time we really see Godzilla. Up until this point, he is sort of like the shark in Jaws. We get glimpses of him but we never really see the full creature. So this scene was chosen to really show off the mood and tone of the movie, and how it would be different from the previous U.S. incarnation of Godzilla.

Gareth, who was still in London at this point, was working remotely for about five or six weeks with me and several artists on my team. In the end the sequence we put together ran for about seven minutes. Then when the film moved forward into production, we came back onto the project in November 2012 to previs the entire third act of the movie.

DS: So would you consider that first piece of work a pitchvis piece?

EC: Technically yes, but they were really keen for it to be a sequence from the film. Legendary was very confident that they were making this movie, so they didn't want the pitchvis to be throwaway material. A lot of the work we did in that first sequence made it into the finished film.

DS: Did that initial sequence have finished VFX in it?

EC: No. The visual effects weren't added until a year later. It was initially shown to Warner Bros. just in its previs state. But even in the previs state we were editing the sequence and adding music and sound effects to try and set the mood, tone and excitement of the overall scene. We were bouncing edits back and forth with Gareth in London, who was also making changes and adding sound FX and other elements. In the end, Legendary had a sound designer who came in and did some more work on the sequence, refining the sound design and adding some music to create a really exciting sequence.

DS: And at this point are you lighting and adding textures to the previs?

EC: Yes, absolutely. The previs is fully textured in Maya.  We approximated the lighting and did a sort of hardware render. Then we brought it into After Effects where we added explosions, water effects, lens flares and glows. We didn’t want people pulled out of the emotional experience so we tried to make the previs look good.

DS: In terms of the third act, what were the biggest challenges you faced in with the previs?

EC: The sequence consists of Godzilla, who is fighting two other creatures, the Mutos, in San Francisco. We've seen Godzilla before, and he's kind of like a big lizard, so we had a fairly good idea of how he might move. But one of the biggest challenges was figuring out how to animate the Mutos. When we were given the designs of the creatures, they were still in development actually. There was a lot of exploration work with Gareth in order to figure how the Mutos would fight and how Godzilla would fight them.

We started to look at footage of animals - Komodo Dragons, bears, even some winged creatures. Gareth sent us a lot of reference material of real animals fighting. So we started out doing a lot of fight tests and sending them over to Gareth, figuring out what he liked and didn't like and developing it further. He had a vision for the film, that even though it's a monster movie, it would feel very grounded in reality. So with that in mind you have to think of these monsters just as big animals, and research how big animals fight each other. A lot of the things you might think would be cool for a monster fight aren't the kind of things real animals do. So there was a lot of exploration in finding the right tone.

DS: At this point are you still working remotely with Gareth?

EC: No, in December we moved up to Vancouver. We were there five months before they started production, and we hired some local artists to come and work with us. We ended up with seven previs artists working in Vancouver, as well as our previs editor. Then we had four or five artists back in LA. We'd be sitting down with Gareth several times per week, which was great. He always came in with a lot of ideas but was also open to trying different things.

DS: Were you provided with assets to work with, or were you creating everything you used in the previs from scratch?

EC: We were provided with a number of assets including the Godzilla model itself. We were also provided with models for the two Mutos that we modified, and then we were given a model of San Francisco from MPC. Everything else we had to build. For a lot of things we were building based on what the art department had already designed, and for other things we might be building from a real location such as the Golden Gate Bridge. So there was a lot of asset building that we had to do, but also we were able to leverage material from our archive, which is our repository of assets we've accumulated over the many projects we've worked on. So if we needed something like a battleship, or an aircraft carrier, we'd just pull those out of our library if we already had them.

DS: Did you do any techvis with regards to planning how some of the action scenes were actually going to be shot?

EC: We didn't really do extensive techvis, primarily because at a certain point, most of the shots we prevised were going to end up being fully CG. Also our main focus was on the story, composition and tone, and helping create blueprints for the production team to work from.

In some scenes we tried to work within the constraints that the crew was going to have, to make the previs more useful when shooting. For example, at one point there's a Muto cave, and we had plans from the art department on how they were going to build the set. So we knew which sections were going to be open, and what the proportions were all going to be, so we could stage the scene in a way that would be achievable for them to shoot on the day. But at the same time we weren’t telling them exactly where the camera needed to go.

DS: Was there a distinct point where you finished the previs and moved into postvis?

EC: Absolutely. We finished the previs around March 2013, and they finished shooting in July 2013, and came back to LA for post. We started postvis in September 2013 and worked until around January. That portion was handled by one of our postvis supervisors, Scott Hankel.  My understanding was it involved a lot of the work we did for the third act, and integrating that work into the plates they had actually shot. And they were able to of course leverage of all the previs, since they used it as reference and adhered to it closely, which meant you could often slip elements right into the comp without changing too much.

DS: Any highlights from your work on this project?

EC: It was a great project. In the end we had about 40 minutes of previs, once everything had been edited down. So there was a large amount of footage. The halo jump was definitely a highlight, and a lot of the imagery in the final sequence, the trailer, and the poster are close to the previs we did. Working with Gareth too was really a fantastic experience. He was a very collaborative director and a lot of the suggestions we put forward made it into the movie. Seeing the final product was great!

DS: In terms of previs, postvis, techvis…where do you see all this going with regards to the way the studios are using these processes?

EC: I'd like to see it heading to a point where studios will previsualize the whole film. We actually did this for a movie called Paradise Lost, which ultimately didn't go into production. But we prevised the whole film, and when you do that you realize that it isn't just useful for figuring out the complex things, like the VFX questions and what you're going to need to build. You can also get a good sense of what is and isn't working story-wise. On this particular project, after seeing the film laid out end-to-end, the director realized that there was a scene they didn't need. Had the movie gone into production, this could potentially have saved a couple of million dollars!

That's where I would love for it to go, and I think that directors would find it to be an exciting process. They can experiment and try different things the risk of being stuck with something that isn't working. Of course it's nice to be able to be out onset and discover something organically and change course because of that. But the reality is that the studios are spending $150-200 million on these movies. They don't want to leave too much to chance. They don't want to get into editorial and have to fix major problems already locked into a release date, which they aren't going to want to change.

I think the previs society motto is "fix it in pre," as opposed to the old adage of "fix it post." If you spend the cycles in previs to really figure it all out, and you have the quality level and enough polish, you're not leaving things to chance. There are costs in doing so, but I'm a firm believer that it will pay off down the line.


Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.

Dan Sarto's picture

Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.