Editors’ Note: Deadline’s Reopening Hollywood series focuses on the complicated effort to get the industry back on its feet while ensuring the safety of everyone involved. Our goal is to examine numerous sides of the business and provide a forum for leaders in Hollywood who have a vision for how production could safely restart in the era of coronavirus.
Mark Sawicki is a Clio-winning VFX and opticals artist whose credits include HBO’s Tom Hanks miniseries From the Earth to the Moon, The Dark Knight Rises, Bullet to the Head, Tropic Thunder, 3:10 to Yuma and X-Men among several films, and he’s the co-author with Juniko Moody of the upcoming book Filming the Fantastic With Virtual Technology: Filmmaking on the Digital Backlot. As film and TV productions grapple with new guidelines for social distancing, and the re-assessment of crowd scenes, Sawicki offers his thoughts on how VFX can help solve many of the challenges arising as crews look to return to work. Sawicki, who has taught at several colleges and schools including The Stan Winston School of Character Arts and the Global Cinematography Institute, is currently developing TV content based on his Filming the Fantastic books.
DEADLINE: As production comes back, there have been concerns about staging crowd scenes, and the possibility of scaling back on background actors on set. Those in production will increasingly look to VFX to solve the challenging situation of creating a crowd scene. Let’s talk about this. When it came to creating realistic crowd scenes, I always go back to what Ridley Scott and his Oscar-winning VFX team pulled off for Gladiator in creating Coliseum masses by replicating a sampling of actors.
MARK SAWICKI: Creating crowds from a few people is a common practice that goes all the way back before digital. The closer people are and recognizable, it’s much better to shoot people and duplicate them or replicate them.
A typical process would go like this: We’ve got to fill a theater of 100 seats, but we’ve only got 10 people. Obviously, the way that was done before is you’d clump them all together, and of course, we can’t do that. So I’m going to talk about a technique that was used on the 2005 Heath Ledger movie Casanova by (VFX supervisor) Bill Taylor called the paper doll cutout technique.
Essentially, let’s say you have 10 actors and you space them apart not 6 feet, but 10 feet, and you put them in front of a green screen, and they’re standing there, and they’re cheering, and then you do that take. Then they walk away. They switch costumes, put on different hats, rearrange their order, and now they go back to the same place, but they move over 5 feet.
Now you shoot take 2. So now you’ve got 20 people that are 5 feet apart, right? Repeat the process, and you’ve got 40 people. Now you’ve got a whole line of people. In that same line of people, put them behind themselves and behind themselves, and very quickly, you can exponentially get a large crowd.
Now, techniques like the software Massive which uses CGI people, that was used for Lord of the Rings, and it’s fantastic. But if you were to examine one of those little CGI puppets up close, they’d look pretty goofy, right? In order to make it work, you have to have them be very simple, very low poly count, so to speak, so that if you look at them, they would look pretty gamey. They wouldn’t look real. So the trick would be to use real people when they’re close to the camera and then throw in the digital actors to continue with them if you need a ridiculous number of them.
DEADLINE: Some background actors are worried that their jobs are potentially in peril in the post-COVID environment. But we’re always going to need background actors, even to do digital shots.
SAWICKI: I think background actors are very important. You know, this is a moving target as we’re adapting. I mean, obviously, they’re going to get hit really big with the social distancing. So a producer might want to hire 100 people, but just because of safety, they can only hire 10, and then you have to rely on these tricks to multiply them. I think people will move over to using fewer people and just duplicating them because I don’t really see a workaround with the social distancing.
I think your excellent news site has touted all the notable problems, the legal issues, the medical issues which will be very complex. The other thing I can say about people that are afraid of losing their jobs, like background actors, all of this is very well founded, and I don’t think there’s any other group. VFX was a prime example of having to adapt. When I started, people would do matte paintings, oil on glass, really traditional, and then Photoshop came in.
So what happened? What happened was you had some matte painters adapted, and it was really hard to go from holding a paintbrush to holding a mouse. Totally different. Imagine sketching with a pencil on paper, and now someone gives you a 5-foot-long pencil and asks you to do the same thing. It was very similar to that. The technological hurdle was enormous. Some adapted. Others felt that, you know, this ain’t for me.
So I certainly don’t want to be a fear-monger and say, oh, it’s all going to go this way or that way. It will go this way or that way, and some people will find workarounds for it and find other ways of dealing with it.
DEADLINE: This sequence you sent us of the VFX from the ABC series Pan Am, created by Stargate Studios and Zoic Studios, illustrates how we can do less with more in creating crowd scenes. As safety restrictions are implemented post COVID-19, do you think that a greater distance will be required between actors in creating a VFX-heavy show?
SAWICKI: The Pan-Am process entailed bringing all the technology to the point of the shoot where, when you move a camera, its virtual twin moves in exactly the same way within the computer, but you still need to shoot a bunch of people and partial sets.
But my concept is, let’s really take it to the extreme. Nobody can touch nobody, right? And you still want to put them in a big space. So my thought was this, using the paper doll cutout I mentioned, you get yourself a truck or a motor home or something like that, that has a compartment that’s cleared out that becomes a pre-lit green screen stage, and it’s all sterilized.
Pretty much, it’s just a green room with lights and a chair and a static camera. This is what’s important, a static camera. The guest goes in there all by themselves, and the operator stays in the cab and runs all the technology. They essentially have a Zoom meeting with somebody across the world, across town who’s got the same arrangement. So each of them are looking at a laptop with their face on it, but the camera is recording them in front of green.
Once the simultaneous interview is done, each element is brought in. The green goes away, and they’re placed within the dimensional CGI set. Let’s say you’re interviewing Francis Ford Coppola about Dracula, you put him in a big Dracula set. It’s all synthetic, and then because each of them are shot separately, you can place them as close as you want. So it looks like they’re sitting 3 feet away as a normal interview.
The two would be staring at each other, so it would almost be the same type of thing as Jack Black in Tropic Thunder playing all these different characters — one actor playing six different roles. They sit there, and they’re performing in front of a stand-in or no one or their playback. In this case, they’re actually talking live to someone, as they would with Zoom. So if someone says something funny, the other person’s going to laugh at exactly the right time because it’s live. So they’re put together, and basically, if you do a push-in or a move, that can be done in post.
So, in other words, you’ve got a cameraperson sitting at home, and they’ve got their little camera or iPad or something where they actually see the composite of the people in the CG set. They move forward. Camera says, “Oh, you just moved forward 3 feet, so I’m going to move the virtual camera forward 3 feet.” It’s almost as if you’re replicating what you would’ve done on set, except separate and in post.
DEADLINE: To address the concerns of social distancing on set, you mentioned a concept to me that involved isolated green stages sanitized with UV light that could be occupied by one or two sole performers with robot cameras operated remotely. Coming back to production post-COVID, how can we make it safer, especially with visual effects?
SAWICKI: What I’ve been kind of experimenting with was just a proof of concept here alone in my house with my bedroom and one green screen and a cell phone camera, is how easy it is to integrate a person within the set. What if you have to have two people kiss? Now we get this intimacy question, and my thought on that is sort of like a continuation of what’s already been done. You get films like The Avengers, right, and you have these actors that are interacting with characters like The Hulk that are 10 feet tall, and they’re huge, and if they touch The Hulk, they got to touch something that’s bigger than the actor, so they have the other actor wearing a big suit or something.
Leonardo DiCaprio gets tossed around by a guy in a green suit who’s replaced by that amazing grizzly bear in The Revenant. I’m thinking that same thing can be done by just modifying these green-suited people by not only wearing a green suit, but also wearing the same protective gear that a doctor would wear if they’re working with a patient, let’s say. So you get someone in a green suit, but they’re on top of it wearing the costume of the other actor, and their hands are free, right? They got clean hands, but their face is covered with some real protective mask or something.
And like we’ve done many times before, we do face replacements, because you got an actor. Very popular. Can’t play the piano to save their life, so you have someone else play the piano, and you stick their face on that other person. So it would be very much like that, and we can achieve intimacy and safety I think in that manner.
DEADLINE: How can the LiDAR technology help when production resumes?
SAWICKI: All of these empty locations need to be photographed and turned into a dimensional model. Elliot Mack, who invented the Lightcraft Technology, went to St. Mark’s Square in Venice early one morning before the sun came up. He shot a thousand or more still photographs of every pillar, the floor, the walls, everything, and then painstakingly, over a month, he made a photo-accurate replica of St. Mark’s Square. Another example of that: There was this tomb that was opened in Egypt. They sent one photographer who shot the walls, all these hieroglyphics, maybe a foot away, high-detailed photographs. And then they re-created this tomb in a photo-real computer graphic set, essentially, and then they have it in a museum. People put on their goggles, and they’re actually walking the space of the tomb. They can go right up next to the wall and practically smell the paint, and it’s highly detailed, and this is an amazing opportunity to have people do the process of photogrammetry, and what I mean by that, let’s say you’ve got a pillar. Now, at St. Mark’s Place, every single pillar is actually unique to itself. What Elliot did was take his still photographs, and he walked all the way around the pillar, 360 degrees, and was able to make a dimensional thing out of all those still photographs. The physics behind that go way back I think to the 19th century, but computers now, with their power, can assemble all of these and come up with a photo-real duplicate of whatever you’re photographing.
DEADLINE: So there’s a huge opportunity here, because it’s painstaking and expensive to erase people digitally when you’re trying to create a clean location, right?
SAWICKI: Exactly right, especially if the camera’s moving and all of that. Like, we had to do that for Eternal Sunshine of the Spotless Mind at Custom Film Effects. They shot in Grand Central Station. We aren’t going to close that place down, right? And the two actors are walking through, and everyone’s disappearing around them. Usually when that’s done is that you shoot a clean place.
The camera’s locked down, and you absolutely must put a beauty light near the actor to make them look great, but you can’t do it without seeing the beauty light. So you go ahead and shoot it with the beauty light, and then you tell the actor to get out of the way. You pull the beauty light, and you shoot the set by itself. Then it makes it easy to erase the beauty light. But in Grand Central Station, you’ve got to beg and borrow from areas around people in order to erase them. You would take the empty spot where the actor was standing and then use that to put it on top of them when they are standing there and vice versa. So you’re chasing it around to try to use some material to put on top of them to make them vanish. If you don’t shoot an empty room, you have to do that thing, and it’s quite time consuming.