The VFX supervisor discusses how MPC helped director Gareth Edwards bring the iconic monster to the big screen.
Helmed by ex-VFX artist Gareth Edwards, the newly released Godzilla reboot manages to walk the line between staying true to its original B-movie roots while delivering a stunning $160 million blockbuster. Anyone who has seen the critically acclaimed movie will agree that the human characters are mere decoration compared to the real stars of the show - Godzilla and the two Mutos. There were a number of VFX houses involved bringing these characters, and the destruction they cause, to the screen, but the bulk of the work was handled by MPC. We caught up with visual effects supervisor Guillaume Rocheron, to discuss MPC's role on this movie.
Dan Sarto: Tell us about MPC’s work on this production.
Guillaume Rocheron: We were involved with Godzilla very early on, probably the year before principal photography even began. In 2012 there was a piece presented at Comic-Con which Gareth put together with the team here at MPC. It was a very early teaser designed to show his overall intentions for the movie. It mostly served to get across the mood, the look and feel for the movie. It was quite slow, showing the aftermath of destruction, silhouettes of the Godzilla creature who, at the time, we hadn't even finished the designs for. So that's pretty much how we first got involved with the show.
A year later when principal photography started in Vancouver, that's when we really started to work on the movie. During preproduction MPC was mostly responsible for designing and creating Godzilla and the two Mutos, and then down the line we handled all the shots where Godzilla is fully visible. This consisted of pretty much the entire third act of the movie, where Godzilla and the Mutos arrive in San Francisco, the halo jump and the final showdown where the 3 monsters are fighting it out.
So it was a pretty big undertaking, first realizing all the monsters and then working on all those epic shots with Gareth for the big finale of the film. It was a great project. When you're a visual effects artist it's always so rewarding when you get to work on the title character of the movie, who in this case is Godzilla, the most iconic monster of all time. There was a real excitement within our team that we had the opportunity to bring this creature to life.
DS: How did you approach designing the movie’s overall look and feel from a visual effects standpoint?
GR: It was a very interesting approach. As you know, Gareth was a visual effects artist himself before becoming a director. That meant when it came to thinking about animating the characters and designing visual effects for the movie he wasn't at all intimidated or overwhelmed by the technology. He had very smart opinions.
But in terms of the look and feel, Gareth is a big fan of monster movies, and particularly of early Spielberg movies where they use a lot of small teases to gradually build up tension before unleashing everything on screen. So this was something that we played with a lot. The idea that we were teasing early on in the movie, then in the final showdown we go all out and really show the characters and their personalities.
It was also very important that everything in the movie be grounded in reality. So from a design standpoint, for example, when we were thinking about Godzilla, we studied a lot of lizards and alligators, Komodo Dragons, and such. We want you to believe that Godzilla could have existed and really be there. He has to be rooted in reality. But there is also a big legacy around the character of Godzilla, so we also had pay homage to that. For example, Godzilla always stands upright, he doesn't slip down onto his legs like a lizard. So we really had to find the right balance in terms of how best to treat him, keeping in mind that when you put Godzilla on screen you need to make sure he has a very strong presence and an iconic look.
Another thing Gareth always wanted grounded in reality was the position of the cameras. Whether we're on a roof top, down on the street, or a helicopter, there's always a justification as to where the cameras are and how they behave in terms of the action. He also wanted to make sure every shot was iconic. For Gareth, a good shot is a shot where you can identify what's happening and read the composition very clearly, even if you had your eyes squinted.
DS: With regards to the shots you worked on, how many were fully CG, and how many were live action plates with CG elements?
GR: Well, Gareth spent a long time working on the previs, so by the time he got onset he had a pretty good guide to work from with regards to camera angles, lenses, what needed to be shot in live action and what would be added in later as effects. Most of the shots had some kind of live action plate, with CG environments, characters and effects work. So for example, for the sequence on the Golden Gate Bridge we shot the live action parts in Vancouver then extended the bridge out and added in San Francisco bay.
The final showdown was very similar. It was either green screen plates for foreground elements, or footage shot in Vancouver that we had to extend with CG environments. With the shots that had greenscreen foreground plates we were really following the previs very closely because Gareth had put so much work into it. He spent almost a year working on all the previs for the major set pieces, really paying attention to the finer details. So it was a really good base to make sure that we were capturing the right things onset, and to make sure that we could fit the creatures accordingly into the shots when we came to do the VFX work later.
We decided to go with an all-CG San Francisco for the monster battle as it gave us a lot more control. We ended up mapping out all of the cameras based on the previs. Then, we organized a reference shoot. We sent a team of photographers to San Francisco to take photographs on the rooftops of suitable buildings, using the Envirocam we developed for Man of Steel. This allowed us to capture very high resolution 360 images which we then used to create 2 1/2 D versions of the city. We actually ended up recreating most of downtown San Francisco in CG.
Using the CG city meant that we had control over where we could put things like fire, or dust clouds, because even though the scenes happen at night, we wanted to make sure we had clarity of action. Gareth has a really good eye and wanted to create these iconic silhouettes of the creatures as they battled it out, so it was really important to be able to use the environment to contribute to how we were lighting the shots. To really build an interesting composition, we ended up having to do a lot of virtual cinematography within our CG San Francisco.
DS: What were the main creative and technical challenges you faced?
GR: I think from a technical standpoint, especially in the end sequence, one of the most challenging aspects was just the huge amount of layers required. You have to create the city, which needs to be destroyed, you have all these pockets of fire and other elements. Those shots were incredibly complex, especially when the creatures are kicking a building, or interacting with water, because then we have reference to gravity on a huge scale. For example, on many of these shots we need to produce 5-10 blocks worth of CG dust, so we're producing massive solutions in order to get the level of detail required to give the sense of scale. Constructing all those different elements was extremely time consuming but necessary to make a very cohesive frame.
But what I really loved on Godzilla was because of Gareth's approach to making sure every shot is iconic, we were putting all that technology towards producing a great looking image. For instance, Gareth never asked us to destroy a building just for the sake of doing it. The composition of every shot was carefully considered and everything was there for a reason.
And also just creating the Godzilla character itself was a massive challenge. He is the biggest and most complex character we've ever created at MPC. One of his fingernails, for example, is the size of an SUV, and because a lot of the shots would be seen from a human perspective that meant we needed to make sure we have a very high level of detail, not just in terms of the model and texturing but also things like deformation details. We had to think about skin thickness and scale sliding, to really try to convey the scale and the weight of this creature. We had hundreds of displacement maps over the body, and hundreds of texture maps to really go into the intricate details required for the close ups. So that was quite a big undertaking too.
Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.
Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.