Interview: ‘The Creator’ Director Gareth Edwards and the Oscar-nominated VFX Magicians of Industrial Light & Magic [VIDEO]

---

Amidst a landmark year for cinema that contains many superlatives, in the field of visual effects there’s arguably no film from 2023 that bests The Creator, Gareth Edwards’ odyssey and magnum opus. Set decades into a future where AI has brought out the worst in humanity, John David Washington stars as a jaded husband traversing the globe to hunt down a powerful AI capable of destroying the human race, and perhaps find his wife in the process.

The Creator was crafted through an irregular process: rather than setting up green screens and compositing all of the necessary effects and enhancements in post using fully digital tools, Edwards enlisted Industrial Light & Magic to embark on a daring mission that would allow him to shoot everything they possibly could in-camera, utilizing wholly unique locations across East Asia and physical stand-ins which would then be worked over by the VFX team. That sentence alone is enough to strike fear into the heart of any VFX artist, but, to their credit, ILM rose to the challenge trusting Edwards completely. The end result is some of the grandest, most believable VFX work that live-action sci-fi has ever seen, with arguably not a single frame that jars or falls flat.

In the wake of The Creator receiving two nominations at the 96th Academy Awards – particularly its tremendously well-deserved Best Visual Effects nomination for ILM VFX supervisors Jay Cooper, Ian Comley, Andrew Roberts, and Neil Corbould – we sat down with Edwards and three of the four Oscar-nominated VFX wizards (Cooper, Comley, and Roberts) to discuss their process behind this monolithic work in the sci-fi canon.

Griffin Schiller: Well, I guess the right way to start is to say congrats on the VFX nomination. I wanted to start with your takeaway of being recognized in that capacity for so much groundbreaking stuff in a film like this.

Ian Comley: What’s amazing is we get to enjoy the process, but we are representing such an incredible team of people, a global team of people who came together within ILM and in fact also some other vendors as well who contributed to the movie. Folks poured their hearts into this one. They could see the vision via the stunning photography that Gareth, Greig Fraser, and Oren Soffer had come back with. If you weren’t working on this film at Industrial Light and Magic, you wanted to be. It was that kind of project.

Jay Cooper: The thing that’s most interesting to me is that we’re getting such great feedback for a movie that was so unconventional in its construction. I’m speaking a little bit for Gareth, but he wanted to approach this almost like a student film, to shrink down as small as you could to travel to all these amazing locations. I’m going to let him speak about that. It’s much more fun to hear it from the director’s mouth.

Gareth Edwards: I got very lucky in my career. I did a very guerrilla, small film [2010’s Monsters], that got me the attention that allowed me then to do these big blockbusters. I always felt there’s a sweet spot somewhere between the two. The simplest version of that was, go shoot it like an independent film, and then keep a lot of your money if you have any left, and use it all in post to then create the science fiction. So we didn’t really have a green screen or anything like that. We went to seven different countries. The crew’s small enough, so it’s cheaper to fly anywhere in the world than it is to build a set. We shot everything in-camera. For every single effects shot in the film, there was something in camera that represented it. Even if it was going to be a sci-fi vehicle, there was a crappy little Ford car that all the actors were in, and then it all got digitally redesigned and replaced in post.

Jay Cooper: That’s the trick. I feel like that’s the part that we brought to bear, is how to roll with those amazing locations and to build out this world in a compelling and interesting way. James Klein, our fantastic production designer, did an amazing job of designing everything, the robots, the vehicles, these environments. We wouldn’t be anywhere without his contribution or that of Oren and Greig. It was a really fun but very difficult project.

Griffin Schiller: That was actually where I wanted to go next, because I’ve heard you, Gareth and Greig and Oren talk about crafting such an imperfect look for the film, like in those films from the seventies that you were drawing upon. I can’t imagine that’s very easy. You’re not working from green screens. These are real clean plates, real environments that you’ve got to insert the assets into.

Ian Comley: Totally. Andrew was out there as our sort of one man visual effects crew, gathering as much as he humanly could, and what we got back was gold. But there was also a lot of detective work piecing that together, looking at the plates and trying to fill some of the gaps. Our artists were looking for all the cues and subtleties of the lighting. We’ve got our simulant characters with the negative space and all the mech in their heads, so how can we reconstruct a very localized lighting environment for putting in the effects work? We’re looking at all the subtle bounce, the red off Alphie’s collar and cream and little chrome pings from the badges. We’re trying to think about light that might be coming through the hole in the head, that negative space as well, and trying to construct this environment. In many ways, if we had perhaps just had an HDRi, I don’t think we would’ve ever been lazy, let’s say, but we would’ve leaned on that very heavily. But because we had that sparingly and we’re already in the zone of piecing together this intricate lighting information by eye with very old school methods, we were already in that zone of trying to get a tight integration in our very bespoke setup.

Jay Cooper: We’ve become pretty good cheaters is the short version of it. This is a movie, we’re trying to create compelling images, so sometimes not having exactly the right lighting reference frees us in a way that gets better images, which is counterintuitive.

Griffin Schiller: And I imagine just by nature of the way that you guys shot the film, because you were relying on such minimal outside light or just your rotating setup with the LEDs, that it allowed you to get in there and place lights in a certain way or manipulate light in a certain way that perhaps you wouldn’t have been able to had it been a larger lighting setup that was a bit more staged.

Gareth Edwards: There was a very brave, insane choice that was made early on, and I’m amazed ILM agreed to this. But it’s a robot movie and we never agreed on who would be a robot until halfway through the edit. So when we’re on set, basically on all these movies, what happens is you go, “They’re a robot, they’re a robot, they’re a robot,” and they have to put those silly pajamas on with all the dots on, and then you’re stuck with it. You cut the scene together and suddenly you go, “Uh-oh, the guy that’s going to cost thousands of dollars to fit into the robot is out of focus and hardly noticeable.” You had an amazing performance of this guy as a robot, but it’s not going to be in there now, and you get trapped in this bad result. So we were like, “Look, the big downside is we won’t have all this data and all these tracking markers. The massive upside is that in post-production, we can just cherry-pick wherever your eye goes in the frame or whoever’s doing something super interesting. We can turn them into a robot.” So we had this beautiful, kind of kid in a candy store day halfway through the edit where we just watched the film together with freeze frames being like, “This guy is this robot.” And then it’d be like, “Okay, next one, this woman is this robot.” And we only had seven different robot heads, and every physical body of all the robots in the movie is the same body, but your brain doesn’t notice. You get away with that kind of stuff.

Jay Cooper: That was the day that we had to buy a lot of beers for our tracking department, because we had to console them in some way of how hard it was going to be.

Griffin Schiller: I’ve also heard about the shoot you did after the shoot, where you basically went out with just a small crew and John David Washington, and you were shooting these more intimate naturalistic moments. That presents a new set of challenges, because these are things that are unplanned. So Andrew, I don’t know if you were there for those moments, but when Gareth would go in there and do something that’s more instinctive and then you would have to go collect the data afterwards, you’re kind of relying on your initial instincts in that moment and your knowledge of the process too, right?

Andrew Roberts: That’s right. I started as a 3D generalist and then would often be the artist that receives material from on set and then tries to figure out how to construct and build a scene from that. So that would be one thing that I would refer to is, if I can only get one thing from here, what would it be? And then I did my best to capture that. And then whether it was an HDRi, or a clean plate, or just tons of photo reference, I would try and capture that. I think Gareth’s familiarity with visual effects allowed him to be more bold with his framing and with his choices, knowing that, okay, we know that ILM and the rest of the vendors are going to be able to execute these shots. So there was a level of trust there in how we approached it. Typically you would know, you are framing this way. This is your actor, that’s your background, you know what the design is, and then you’re going to do four or five takes until you’re happy, and then you do the next setup and the camera’s pointing in this other direction, and then you sort of rinse and repeat. Pretty early I realized that Gareth doesn’t like to say cut, and so he will reset and have his actors just take it from the top again. But he’s moving around and he’s hunting for that perfect angle to capture that moment between them. So we might start off just looking over Alphie’s shoulder towards John David, but then as he resets, Gareth’s wandering around, and now they’re in profile, and now we’re behind John David looking at Alphie, and now it’s a wider framing. So the typical approach of, okay, this is my background and these are the things I’m going to capture, I had to set aside and just try and follow Gareth and take note of the areas where he was shooting, or when he was done, try and shoot some clean plates or try and do a 360 to try and get main coverage of what he shot. But I think that not being beholden to a specific angle, not being restricted, really allows us to have some beautiful imagery.

Gareth Edwards: This Gareth guy sounds like a nightmare.

Ian Comley: It’s worth calling out the effort to take these plates without the full complement of data, and get it to a level where we’ve hopefully got a really tight integration. There was a lot to do there. But on many projects, I think it’s fair to say, having gone to those lengths and having put energies into an asset, you want to show it off. You want 15 shots of that thing, you want to fly through all angles, and it really wasn’t like that here. To Gareth’s credit and the strength of the film, you get these glimpses of the world, you get moments of robots only ever seen once, and it builds up this kind of residual impression that, wherever you could point a camera there would be things enriching this world. It just so happens we are on this particular narrative train, and that’s taking us down this path. It didn’t necessarily make things easier for us, but I think it really helped the audience.

Gareth Edwards: One of the big conversations you get into a lot when you’re turning over shots to become visual effects is, everyone wants to understand what it is that’s there. I think if you really went to the future, like 2070 or something, and shot a film and then came back in time and edited it, there’d be so many things visually you’d see and you wouldn’t have a clue what they were. There’d be giant buildings and little devices and vehicles, and people would say, “What is that?” And you go, “I don’t know. We didn’t have time to ask.” But it was there. To me, if someone from 70 years ago came today, they wouldn’t understand a lot of the things they’re seeing. The first question from ILM is, “Okay, so what is that building and what does it do?” And you’d be like, “I have no idea.” And that was a common answer. And it was like, “No, no, no, we’re not supposed to have an idea.” That makes it real. You know what I mean? It just looks like that. That was the beauty of Star Wars, as there were so many throwaway things that all could be their own little movie if you cared.

Jay Cooper: The other thing that I thought was really helpful,  is finding something that was only the spine of a location, and shooting that as if it were already completed. So whether you shot interiors of airplane hangars or inside of greenery, things that have a very rich textural component to start with, and then took our ability to elevate those with visual effects, that was really cool. And I feel like we ended up in a better place than if we had commissioned that as a set, where we had made a bunch of choices about, okay, this is what the ground is going to be, this is what those walls are going to be. The fact that you start with something that you don’t have, there’s all of this naturalism too. I think it gets us to a much better place, and it allows us to find imagery that we haven’t seen before. Because what happens again is, and this is not to bag on set building, once you say, “All right, it’s going to exist in a stage,” you say, “It’s going to have flat ground and it’s going to be built with plywood, and then we’re going to dress on top of it,” whereas the more locations that you go to, the more interesting things there are to start with that you would never have thought about. That location that you guys shot in Thailand, what’s the name of it, Andrew? The Super Collider, the–

Andrew Roberts: Oh, synchrotron, yeah.

Jay Cooper: Where they turned it into the AI lab, it’s an amazing location. And all of the catwalks and the stuff that’s already there, the metal tubes that I have no idea what they’re used for, are very scientific. And then our ability in visual effects to extend that world out and build forms around it, and layer in hanging arms, that’s an amazing marriage of production and location work with visual effects. It’s also something I don’t think you see a lot of. We’re charting our own lane in that way.

Andrew Roberts: Eight countries, 80 locations. At times it was exhausting, but I think it was absolutely worth it. The production value is up there on the screen.

Gareth Edwards: Everywhere in Thailand, we were like, “Where is there a place in Thailand that looks like this,” and I’d show some pictures, “Or like this?” And one of them was obviously some futuristic sci-fi-y lab, “Not really anywhere. We don’t really know,” “But there must be somewhere.” And eventually they were like, “Well, there is a particle accelerator, but you can’t film there.” “Can we just go visit it?” “Yeah, but they’re not going to let you film.” “Can we look around just for inspiration?” “Okay, okay.” So we go and they were really kind to us and they showed us all around, and we were like, “Is there any way we could ever film here?” And they’re like, “Well, what is it you want to do? It’s a multimillion dollar facility.” And I was like, “Well, we’re going to have guys shooting guns, explosions.” And they were like, “No, no, we’re really sorry.” And then as we were about to leave or something, they were like, “What other films have you done, out of curiosity?” And I ended up going, “Rogue One.” And they were like, “Okay, okay, you can film here on one condition.” And it was like, “What?” And they said, “We have to be in the movie.” And so everybody in white coats that you see in that whole sequence, they’re nuclear physicists for real. Star Wars opens a lot of doors in Asia, it turns out.

Griffin Schiller: I do want to ask because you brought it up, specifically about your design ethos when it comes to coming up with these ships, the buildings, the costumes. Obviously this is a love letter to the science fiction films that you grew up with, but how do you design a world that feels familiar, but also original and not just imitating existing sci-fi, something that we haven’t necessarily seen before?

Gareth Edwards: Well, I have to give James Clyne a lot of credit for this and his team, but we got into this habit of taking silhouettes. Because most things in a shot, when it’s all said and done, with lighting and a bit of atmosphere, it becomes a silhouette. It becomes just a simple shape. If we were doing a tank, for instance, we would take the shape of a tank and then start removing things. We would create a negative space and delete parts of it, rub them out and go, “do you still understand it’s a tank, but yet something weirdly is missing?” We did that with vehicles, with tanks, with jet-copters. Then we learned a little trick. We wanted to go back in time and copy the Sony Walkman. And there’s something about the product design of the eighties and early nineties that has character, but you can also understand instantly what the job of that device is. The problem is when everything’s a clean slab like Apple today, what does it do? And when you look at those product designs, there’s divides and breaks in them that are not where you expect. They’re not like, oh, it cuts across the microphone and then there’s a cut across the speaker. They sort of merge together, and then the cut is somewhere really odd, where the battery would go. So we would take these silhouettes of tanks and vehicles and then find even stranger places to do the divides that are not where you expect the difference of mechanics to be. And it’s all subconscious, but it kind of works. And it kind of screams eighties and nineties product design, and all the cool sci-fi stuff we grew up loving did a lot of those tricks too.

Jay Cooper: We had a great library of reference images for a number of eighties technology pieces. There were toasters and coffee makers and all kinds of stuff where the pieces fit together in a specific way. And that extended, I think, to the surfacing. Alphie herself has this interesting mixture of a number of different plastics and metals, three or four different types that are going into it. We’re always trying to find ways to connect these things. Gareth would say multiple times, “Product design, product design, product design,” and we’re looking for opportunities where, when you’re doing modeling, you start with just a big blob and you’re constantly carving in pieces. But in this case, it was, okay, this is a removable part, it has to feel like it fits in, which means that the edges around all of it are going to have a three millimeter soft radius. And that extended to a ton of stuff. There was, I think we called it the Gareth curvature, which would show up in a lot of different things, which was imagery that we had almost in the first weeks of hockey sticks with a soft radius. Gareth knows exactly what I’m talking about. We saw it in buildings, we saw it in robots, we saw it in environments over and over again. And that just became part of the design technique of this world.

Ian Comley: You guys have already mentioned James Clyne as being fundamental, but what’s really unusual is he stayed on pretty much to the very last shot delivered in the movie. So from our point of view, we were already working within the framework of the design language that Gareth and James had been putting together, but we could keep asking questions throughout. Our artists were reaching out directly through us to questions for Gareth, but equally, they were reaching out to James and having these great conversations, and James would do a quick paint over here and had an artist doing a model there. That was enormous just for detailing and fleshing out this world around the framework.

Jay Cooper: That’s another place where I think our movie’s a little bit different from the way that visual effects are typically run, where it’s very much like a segmented system, pre-production, production, and then post. In this case, we’re kind of all on top of each other, where we were doing a tremendous amount of what would traditionally be production work in visual effects. We would model something, and Gareth and James would take a look at it and say, “Well, it’s kind of right from this angle, but it’s wrong from this one.” And then James would take that image and do a paint-over on top. We’d take it back, do it again, and back and forth. That shuttling between things is typically very difficult just because you know that every time you do this back and forth, it’s additional layers of work. But we knew at the very onset that this was something that we wanted to do. So we built systems where there were art assets that were created in the art department that became the consoles of cars, for example, or became parts of guns. And it’s completely different from other ways that production usually works, where we actually would leverage assets that were built in the art department and we would dress them up in the most modest ways possible to try and get them into shots.

Gareth Edwards: Normally what happens on a movie is you have production, where you go shoot the film, and then everyone, nearly everybody that was involved in that goes, “Bye, see you in a year or so, nice to know you, blah, blah.” And they all go and do the next gig and then post-production starts, and it’s a complete brand new nightmare for a whole group of new people. So my producer Jim Spencer, I was like, “Jim, you’re staying on post.” Because what happens on every movie is they whack up a green screen, and basically all they’re doing is they solve a problem on set by just putting a green screen there, and now we don’t have to worry about it. It’s like, no, that becomes someone’s nightmare six months from now when you’ve all gone home, and so it’s going to be your nightmare. We’re not going to do that, right? It really helped, having continuity from production to post. No one could really escape these issues, and therefore, I think we went about it a lot more efficiently, where it was the right use of money in the right areas.

The Creator is nominated for two Oscars at the 96th Academy Awards: Best Sound and Best Visual Effects.

This written interview has been edited and condensed for clarity.