Visualization supervisor Tefft Smith talks previs, techvis, and postvis on the production of Akiva Schaffer’s hybrid live-action/CG animated action-comedy, starring John Mulaney and Andy Samberg, now streaming on Disney+.
If you needed a couple of pint-sized detectives to investigate the disappearance of a friend who’s in debt to a local gang because of his uncontrollable cheese addiction, you could do worse than to engage Chip and Dale, the chipmunk duo whose enthusiasm and resourcefulness are the stuff of legend.
And if you wanted to hire someone to help the director visualize the best version of their story – before they begin shooting – so that together with the editor, they can then explore any number of versions of their rodent action adventure, which includes a dizzying number of live-action, 2D, and 3D characters interacting in environments ranging from fan conventions to the palm-lined streets of Los Angeles, you might turn to veteran visual effects practitioner Tefft Smith.
As a visualization supervisor on Chip ‘n Dale: Rescue Rangers, now streaming on Disney+, Smith handled previs, techvis, and postvis duties, helping to shepherd the film from its earliest stages through its almost-final form.
Picking up 30 years after their TV series of the same name, the movie finds Chip (John Mulaney) and Dale (Andy Samberg) living not quite their best lives in modern-day L.A., where Chip works as an insurance salesman and Dale, who has had CGI surgery, works the nostalgia convention circuit. When former cast mate Monterey Jack (the one with the cheese habit) is kidnapped, Chip and Dale reboot their broken friendship and take on their Rescue Rangers detective personas once again to save their friend’s life.
Smith, whose credits include Stranger Things, The Mandalorian, The Orville, Kong: Skull Island and Tomorrowland, worked closely with director Akiva Schaffer on Chip ‘n Dale: Rescue Rangers, an experience which he ranks among the most enjoyable and satisfying of his long career. AWN recently spoke to Smith about his time on the film.
AWN: You were involved in all the “vizing,” which means that you got in early and stayed late. Can you take us through the arc of your responsibilities?
Tefft Smith: It was an interesting dynamic. I was brought on by one of the producers of the film as an animation supervisor to help the supervisor, Chris Beatty. He and I started working really well together – then COVID hit and production was shut down. Everyone went in different directions and, when Disney reached out to start back up on production, Chris was unavailable. So they reached out to me to supervise previs for the film. We were in previs for about six months. Previs trickled over a little bit with the start of postvis, and we were doing both for a while. Also, early on, while we were still tackling some of the previs, we were doing techvis and helping [cinematographer] Larry Fong with figuring out certain cameras, how to shoot it on set, especially considering that we were going to be shooting live-action plates with one-foot characters.
AWN: What was the main focus of the previs work that you were involved with?
TS: I can tell you that we prevised over 90% of the film, mostly due to the fact that every single shot had Chip, Dale, or some type of CG character involved. And Akiva really wanted to make sure that this film landed comically, so he wanted to see the performance. It was great because he was really hands-on with the voiceovers, and we constantly were getting audio tracks to line up, and working with the editor to lay everything out so that we knew how the film was going to play before they actually shot anything. It also gave us an opportunity to test certain characters that he really wanted in the film that he didn't know if he could get in the film.
AWN: So, basically, you prototyped the entire film before you put anything before the camera.
TS: Yeah. And it was also, for me, one of the first films where the DP was so hands-on. This really was a collaboration, where Akiva and Larry worked with us on the shots in Maya. At that point we were all on Zoom, so we were screen-sharing and looking at stuff and figuring everything out. And Larry and I could go to a Volume, where he would have digital cameras and he could physically move it as I was animating and setting up scenes. So we really were using all the strengths of all the filmmaker for the previs, figuring out every single shot.
Obviously, there were times when our numbers were a little bit off. But, for the most part, Larry felt confident that he was going in with a clear understanding of what he would need. He was actually able to test cameras before selecting the exact cameras he wanted on the shoot. And he even had the opportunity to figure out, do we really need cranes? Do we need jibs? What is it that we need for the shoot, so that he could help the producers stay within the budget.
AWN: You mentioned that he had access to a Volume, where he was playing with cameras. What other tools did you have at your disposal in putting the previs together?
TS: Before COVID, the idea was we were all going to be on location, and Akiva would be able to sit there with Larry and figure out every single shot. And then the world shut down. So Akiva would look at stuff with us remotely, while Larry and I were physically on stage with one technician. I would be in a kind of cubicle, and I would set up the scene for him. Then I would send it to our virtual production supervisor, who would set it up in Unreal and get it all going. Our virtual camera kit was basically a Vibe in a small Volume. We didn’t have a huge stage, but it was large enough to achieve what we needed. I would throw on a suit – I could either be Chip or Dale. Sometimes we brought in someone else to be the second character. If we ever needed to work with each other, we would scale it up so that Chip would be a six-foot person and the world would be ginormous. And then I would move around and Larry would film me like he was filming a six-foot person, but in a one-foot world.
AWN: So were you compositing this previs in an Unreal engine setup? Or did it always live within engine?
TS: We used the engine as a real-time solution to get a camera. Because everything was prepped in Maya, we had to stay within the Maya Viewport rendering system to keep the look consistent. We had to set up an Unreal test to see if it was going to work, but, due to schedule and time, we didn't have the luxury of rendering everything out through Unreal. So we used Unreal to capture real-time cameras, exported those back into Maya, and then we ran it through Maya with the character animation, and cleaned it up. Then Akiva, [editor] Brian Olds, and I would determine that we have the camera, we have the performance – now how do we reanimate to the camera? And we would basically massage each shot and each scene until it worked in the story. For the most part, until we got into postvis, we didn't get into compositing. Every once in a while, we would composite certain effects that we needed to create a kind of effects library. But otherwise, it was all done through Maya.
AWN: How long were you in postvis?
TS: Postvis started right around the holidays, and I worked on postvis all the way up until August, 2021. The movie's release date was originally slated to be in March, and so they had a drop dead of everything by October. My term ended in July, but they continued with postvis until about the end of September. Luis [Andrade] took over and finished up the last two or three months.
AWN: Part of the challenge as you described it is the fact that you've got cartoon characters in a real world – and they're small. Was that a challenge all the way through, or did you master it early on in a way that served you throughout the rest of the film?
TS: It was challenging the whole way through, which is why it was great having Larry involved from the beginning. As long as he felt comfortable that this world could be shot, then it made sense that this world could exist. And Akiva was very adamant about where and when he was going to push the interaction of real world with CG world. And I think the best thing that he did was that he made size-appropriate sub-worlds within this world.
There have been films that have tried to have it be all CG characters that live in our world, but they always have to be jumping up on a table or something. There's nothing of their size to play with. In the worlds Akiva created, they live with us like humans, and they're just smaller – or vertically challenged – or they're bigger. In an elementary school, you have adult-size tables, you have kid-size tables, you have kids’ chairs, adult chairs. There are different heights, but they’re intermixed. And I think that's how he was able to solve that world-building challenge, writing it so that the worlds lived together. He also made sure that there was never, like, a four-inch Smurf next to a an eight-foot basketball player. So once the rules were established, they were easy to follow.
AWN: Sometimes previs can be confusing for people who don't know exactly what they're looking at and get confused thinking that’s how something will look in the final film. How much detail did you provide in the visuals you produced?
TS: The previs was pretty polished. It wasn’t an Unreal polish, but we had lighting and shadows. When we were outside, especially, we found out what day Larry would be shooting, where the sun would be, and which way we would be facing. The look was as close as we could get. With the advancement of Unreal, it definitely can be even more polished, but it requires more time to prep and more money to get set up. And it's going to take more time to get an output from it. But our previs on this was super-polished. Our Chip looked 2D, so we made sure it was toon shaded with toon lines, and our Dale looked 3D. And all the characters that were in there were either pushed in, like we're in a cartoon world, or a stylized world. Everything was built to set spec or built to stage. Luckily, we received a lot from the art department that then carried straight into finals. We would get the CG sets, and we would put on rough textures, shaders, lighting and so on. If we had time, we would re-render the sequence out to make it even better.
Postvis, on the other hand, we did everything in our power to make everything look like it lived together. And so, outside of any fur or anything like that, we were using a lot of compositing. We made sure our tracks were spot on, and we had multiple passes, whether it be shadows, ambient occlusion, or depth. We really pushed the postvis side, so when you looked at it, it was a good first pass of a shot. And then they really amped it up for the final renderings. The only thing we never did, because Akiva didn’t want to spend the time on it, was lip syncing. So we never did facial stuff. We would do expressions, but we never did talk because he changed the audio and the dialogue so much.
Omar [Hytham] Morsy was the animation supervisor on this, and during previs, especially right before shooting, he was in there, giving suggestions on how to animate. And he was very involved in postvis, to ensure that the performance we were doing was in line with what he had planned for the final. All said and done, there was extensive collaboration amongst all the filmmakers – we met every single day to talk. And when we had reviews, everyone was involved. So it was a really collaborative project.
AWN: Any last thoughts?
TS: When my friend reached out to me to work on this, it just was a no-brainer. I was like, "It sounds like a Roger Rabbit-type film,” which is one of my favorite movies. And Akiva is someone whom I've always respected for the stuff that he's done. I just was happy to be involved. And then, having the opportunity to step up into the role that I was given, and to have the team that supported me, it was one of the great experiences of my life. I mean, it was a lot of fun.
Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.
Hammering Out the VFX in ‘Thor: Love and Thunder’