Life On Mars
Issue: Volume 35 Issue 2: (Feb/Mar 2012)

Life On Mars

They aren’t little green men, but the large, strange-looking Martians in the science-fiction romance film John Carter are definitely green. And alien. The Martian barbarians, called Tharks, have four arms, no nose, and head flaps, and they’re violent. And, they aren’t the only inhabitants of the red planet.

Based on the Edgar Rice Burroughs story “A Princess of Mars,” published in 1912, the action-adventure film follows John Carter, an American Civil War veteran, who finds himself teleported to Mars, where he becomes extraordinary. There, a stranger in a strange land, he battles the barbarians, rescues a princess, fights evil, and finds himself a hero.

Credited as the first planetary romance, the 100-year-old novel and Burroughs’ subsequent “Barsoom/John Carter” series have inspired scientists, science-fiction writers, game developers, filmmakers, and one 10-year-old boy named Andrew Stanton. Stanton would grow up to write and direct two Oscar-winning animated features, Wall-E and Finding Nemo, receive Oscar nominations as one of the writers for Toy Story and Toy Story 3, and, most recently, direct his first live-action film, this year’s John Carter.

“About the same time I read these stories, I saw Star Wars with imagery I didn’t think was possible,” Stanton says. “So ever since 1976–1977, I wanted to see these stories on screen by somebody. I had never thought it would be me.”

At top, DNeg created crowds of green-skinned, double-armed Tharks as well as all the
creatures in the film. At bottom, facial-capture data from actor Willem Dafoe helped animators
create expressions for the hero Tars Tarkas, shown with actor Taylor Kitsch (John Carter),
but the animators keyframed the tall, lanky character’s performances.

The Walt Disney Pictures production stars Taylor Kitsch as John Carter; Willem Dafoe as Tars Tarkas, an unusual Thark whom Carter befriends; and Lynn Collins as Dejah Thoris, the Princess of Helium. Filmed in Utah and on soundstages in the UK, the production features CG characters and creatures acting alongside live-action actors. Carter is always Kitsch, the Heliumite princess Dejah Thoris is always Lynn Collins, but the Thark Tars Tarkas is a CG character voiced by Dafoe, who performed the character on set.

When Stanton took a detour from Pixar to direct the film for Disney, he brought along as a writer Mark
Andrews, head of story on Ratatouille the “John Carter” books, and then invited Michael Chabon (Spider-Man 2) to join the team. Wall-E producers Lindsey Collins and Jim Morris, Pixar’s general manager, became the producers along with Colin Wilson (Avatar). Filling out the group were production designer Nathan Crowley (The Dark Knight), cinematographer Daniel Mindel (Star Trek), and visual effects supervisor Peter Chiang (The Bourne Ultimatum) from Double Negative.

Three Soho studios created the visual effects for the movie: Double Negative (DNeg), Cinesite, and The Moving Picture Company (MPC). Legacy Effects provided physical maquettes and sent Pixologic ZBrush models to DNeg. Halon and NVisage did previs, with Halon handling the majority of shots.

As for the visual effects, “We started by splitting the work between DNeg and Cinesite,” Chiang says. “Cinesite did 860 shots with the red men, who are human actors tinted red with tattoos. They fly in spaceships designed by [art director] Ryan Church and Nathan Crowley, so Cinesite created that hardware and the cities. Double Negative took on everything with the Tharks, just over 1000 shots. But, at some point, there was an individual Warhoon sequence, in which John Carter meets these almost Neanderthal thugs, that we handed to MPC.” In addition to the Tharks, DNeg created various multi-legged creatures.


The Tharks’ most noticeable characteristic is the second set of arms under their main two limbs, which presented an interesting creative problem: What would the second set do? Would the Tharks use each arm and hand to do different tasks or similar tasks?

 “Early on, we did a test with a Thark eating from multiple bowls on a table,” says Eamonn Butler, one of two animation supervisors. “The animators did a great job, but Andrew [Stanton] said it looked like we were trying to prove how cool it is to have four arms. It was distracting. So we downplayed it and used the lower arms to support what the upper arms did.”

Adds Steve Aplin, the second animation super­visor, “Initially we tried to stay away from twinning, from having the arms mirror one another. But as soon as we separated them too much, we felt that our brains couldn’t handle it, so we went back to twinning. We’d separate the arms and try to find poses that worked symbiotically.”

Sometimes the arms worked together; other times they complemented one another. For example, in a scene where Tars Tarkas tries to persuade John Carter to jump to great heights, which is something Carter can easily do in the lighter gravity of Mars, he stands with his big arms on his hips and small arms across his chest. “We used the arms to reinforce his tone and attitude,” Butler says. “That was the challenge—we can’t have the arms all doing the same thing, but they needed to have the same attitude. If they are loading weapons, we’d have them hold the weapon with their upper arms, and have the lower arms load and unload. One of the great things about this film is that we were working on characters that were carrying shots.”

Riggers at Double Negative use an in-house system called Pinocchio, with which they could quickly assemble a bipedal rig with the extra set of arms. The challenge, however, was in moving the shoulders. “The rigs for the upper and lower arms were exactly the same, but we had a bone that tried to act as a scapula and collarbone,” Aplin says. “We considered doing a second scapula and collarbone, but it looked like the parts were swimming because we’re so used to seeing a solid area.”

Ultimately, the riggers gave the lower arm a smaller range of movement that affected the skin area less, and then skinned the creatures so that secondary movement was visible, especially in the back. That secondary movement was procedural, built into the rig.

All the hero Tharks are CG characters voiced by actors who performed the Thark roles during filming. To provide the second set of arms on set, someone in a green suit stood behind each actor so he or she could, for example, hand something from one set of arms to another.
“Andrew was keen to have a representation of the principal actors so all the eye lines were correct,” Chiang says. “So, the hero Tharks were always represented on set.”

The Tharks ranged in size from 7 feet 6 inches tall to 8 feet 6 inches, so to have correct eye lines, the actors playing the characters sometimes stood on boxes sized according to the eye height of each character, and the actors learned how to walk on stilts. “If they needed to do a 20-foot run, we would use Lego-like sets of decking,” Chiang says.

Although the crew considered using an optical motion-capture system, after analyzing the difference between the actors and the characters, they realized all the movement would have to be re-animated. So, they decided against motion capture for the physical performance. Witness cameras on set made it possible to later triangulate a 3D track from the footage for rotoscoping the actors on stilts or boxes, which helped animators, but gait differences often meant animators needed to create the performances from scratch.

 “The Tharks have extremely long legs, so they have a wide gait, which means they get to objects quicker than an actor,” Chiang says. A human would have to jog to keep up with a Thark who was walking. However, rather than have Dafoe jog to hit his mark, they decided to film him walking and fix his gait in animation.

“Eye height in relation to principal actors, and facial capture, were more important,” Chiang says. “Jogging affected facial performance, and it was important to protect the facial capture to have the right emotion.”

Facing FACS

Each actor playing hero Tharks went through a MOVA capture session to determine FACS shapes, and on set, they wore head rigs mounted with two NTSC cameras. A procedural system tracked dots applied to the actors’ faces and translated that motion into a curve-based system that animators could control.
“We needed to redo our facial-capture system for this film,” Chiang says. “We rewrote code and redid the translation and the procedural system that read the tracking of the dots to make it editable for animators. Animators like to work with particular controls, and we needed to conform to that.”

One problem was that the Thark faces didn’t conform to human faces. Thark ears were roughly in the same place as a human’s ears, but they were only holes. The Tharks have flaps on the top of their heads that the animators could control when Stanton asked for more expression. Lastly, and most importantly, Tharks don’t have noses.

“We lost all the emotion captured in the nasal area, so we went through a few iterations trying to convey that emotion in the mouth, eyes, and eyebrows,” Chiang says. “We put in the whole muscle structure based on the FACS shapes, but there weren’t even holes to suggest nostrils, so we had to dull down all those nodes. We got there in a procedural way, in the retargeting. Once we cracked it and retargeted the motion, we would keyframe on top. Then, we baked that in and that would drive the procedural muscles, around that area so we had consistency in how the muscles behaved.”

Even so, creating emotional, photoreal performances for characters with such unusual faces was an animation challenge. The facial capture gave them a blocked version of the expressions quickly. Then, they needed to refine the emotion. “A nose gives you certain tells,” Aplin says. “We had to find ways around that.”

That was especially important in scenes without dialog, when the character’s face tells the story, as in a scene with Sola. Sola is the daughter of Tars Tarkas, and one of the few Tharks to feel empathy. When the Tharks captured John Carter after his arrival, Tarkas assigned Sola to guard and teach him.

“In one part of the film, Sola is punished for an indiscretion, and we had to tell what was going on through facial acting,” Butler says. “It’s amazing how much you miss when you don’t have the lines that flow from the corners of the nose to the mouth. We pushed other parts of the face harder, and we started introducing the neck, chest, and abdomen. We had people [at Double Negative] say lines from the film so we could look at their necks. The neck is so alive; a lot of noise is happening in the neck, and it’s different from person to person. Tendons flare to the side. We saw the voice coming from farther down in the throat. So the creature-effects team hand-animated neck muscles, tendons, veins, and muscle bulges driven by the main performance.”

 Mars Attack

During a stand-alone sequence in the film John Carter, the hero John Carter, Princess Dejah, and Sola, a Thark, are traveling across the red desert when a savage tribe of desert nomads, led by the antagonist Matai Shang, attack them. The Martian nomads, known as Warhoons, look like larger, thicker versions of the tall, double-armed, noseless green Tharks. Adam Valdez, visual effects super­visor at The Moving Picture Company (MPC), which created the characters, describes the sequence.

“We follow this pack of several thousand Warhoons chasing the heroes across the desert, and then John Carter realizes they can’t outrun them,” Valdez says. “Because he has enhanced strength in the reduced gravity, though, he leaps over them, attacks, and then leaps somewhere else before they can react. He causes chaos and huge pileups of Warhoons.”

On set, actor Taylor Kitsch, who played Carter, fought thin air. It was up to the animators at MPC to line up the CG action and animate the Warhoons in a way that made the battle believable. Making that animation challenge more interesting, director Andrew Stanton intercut this sequence with flashbacks of John Carter’s previous life on Earth to tell, for the first time in the film, his backstory. Carefully designed choreography made it possible for the audience to follow the battle action despite multiple cuts in the sequence. For example, if the focus in a flashback led their eyes to frame right, and the next frame was of the battle, Carter would be in that same frame-right spot.  

“Andrew produced a set of drawings, which were on top of the plates, so we understood the intention in the rough staging,” Valdez says. “We could see where he wanted characters to enter and leave the frame. With his animation background, he could produce rough drawings that gave us the rhythm on screen that he liked. The sequence was cut to a piece of music; it was all designed that way. So, we carefully monitored the progress of the scene to not water down the primary indication of the drawings, the editorial rhythm. We had to hold onto the design, while adding gobs of complexity and making the shot clear. It was a quintessential scene to work on. You don’t always get a chance to work on scenes where there is that much intention, where the scene is that designed. It was interesting to fulfill that design, satisfying to work in the medium and not just do photoreal effects work.” –Barbara Robertson

Green Man Group

For the crowds, which totaled 1500 Tharks at one point, the animation team adopted a three-pronged approach: keyframe animation for characters close to camera, motion-captured cycles for the rest in two ways. Animators individually placed the characters in the first few rows behind the hand-animated Tharks. Beyond that, Double Negative’s crowd-animation system handled the rest. “It’s a bespoke system in which the objects are aware of others around them—they don’t bump into each other, and they respect the terrain. But, they aren’t choosing from a database of moves. The demands of the film were that the crowd needed to roar, run away, move across the terrain.”

At left, animators at Double Negative started with facial-capture data for Tharks, and then
found ways to create expressions for creatures that have no nose. At right, Woola also
presented animators with a unique challenge: They needed to move the heavy, 10-legged
creature, a cross between a bulldog and toad, easily across a desert. 

Variation in the Tharks’ skin color, which ranges from blue-green through yellow-green to a reddish-green helped make the crowd believable. And, textures achieved through bump and displacement maps that mimic human skin added another layer of believability.

“When we first got the illustrations, the Tharks had black sclera and red eyes,” Chiang says. “They looked like they wore contact lenses. But, once Andrew [Stanton] saw the performance from Willem Dafoe, he wanted to retain it, so we reverted to white scleras, irises in human colors, and black pupils. And, we started to go down other routes to make the Tharks more relatable—the subsurface qualities, the cellular breakup on the skin—we even made the palms of their hands lighter than the backs of their hands. It was an interesting transition from totally Martian to something you could relate to, but that retained a Martian quality.”

Even though the Tharks had blue blood, the interior of their jaw and gums were red, and their tongues have an underlying pink fleshiness. “They had a purplish tint that was red-purple, not blue,” Chiang says. “But they had blue blood coursing through their veins.”

For costumes, the team used wire simulations in Side Effects’ Houdini and in nCloth working with Autodesk’s Maya. “Loincloths on the hero characters were nCloth,” says CG supervisor Christoph Amman. “And we used Houdini wire sims for necklaces, belts, and baldrics that interact.”

All told, the digital costume department created 44 different costumes with individual digital setups, which added to variations in the crowds, as well. “Although we had some variation in the green skin, most of the crowd variations came down to costumes and face paint,” says Justin Martin, CG supervisor.

For hair, the team again used nCloth strips and Houdini wire sims. “The Houdini wire solver is sort of set up as a hair simulator, but it’s useful for quite a lot of other effects,” Martin says. “It’s easy to simulate the interaction between different materials. We built proprietary nodes around the main solving engine.”

Thousands of Warhoons

The Moving Picture Company (MPC) artists received models and textures from Double Negative as a starting point to create the super-sized versions of the Tharks that DNeg had built. They modified the models, changed the textures to vary the look and make the characters readable in the distance, and installed new rigs. Only a few Warhoons came close to camera; the artists built the rest for mid to wide shots using a combination of Pixologic’s ZBrush, Adobe’s Photoshop, Autodesk’s Mudbox, and The Foundry’s Mari on an Autodesk Maya-based pipeline. For the widest shots, MPC’s Alice crowd-simulation software generated the action using a low-resolution model.

“We’ve gotten used to tracking variations of characters and weapons through our pipeline,” says MPC visual effects supervisor Adam Valdez. “We try to create some fun in the animation and hang onto a physical reality. We don’t want a cartoon, but at the same time, there’s an animator’s aesthetic. This is not a film with invisible effects. It features effects proudly. So you want big ideas and a broad execution of those ideas with lots of detail to help make it believable. It’s a crazy fight, a big, outlandish piece of action. But we never wanted people to look at it as animation. It’s a desperate battle with edge. We placed lots of blood deliberately to be readable, but not push ratings.” As with the Tharks, the blood is blue.

To create the environments, the matte painters started working in Maya when they needed complex 3D geometry, then moved into The Foundry’s Nuke. Pixar’s RenderMan handled the rendering, although the team sometimes called on Mental Images’ Mental Ray for volumetric effects.

The layout department worked within Maya to line up shots in virtual space, and used Alice to block in the characters. “The layout artists know how to deal with different live-action plates meant to look in the same direction but shot differently,” Valdez says.

Once staged, the team divided the shots according to whether Alice would handle the crowds or animators would keyframe the action. “We thought the shots would be crowd-driven based on the story,” Valdez says, “but because every shot is so specifically choreographed within the frame, we used as much, if not more, keyframe animation. Sometimes we might start in the crowd software and then animators would add detail and unique performances. Other shots would start in animation, the crowd software would help with ragdoll simulation, and then the shots would go back to animation for cleanup. It was tricky making it feel like that large volume of characters were in the space making contact with John Carter’s sword.” –Barbara Robertson

Quick Loads

To match the lighting on location, a crew took 360-degree HDRI images and then extracted light sources from those images. “We had every lighting setup, so we could create generic light rigs per sequence and push that out to the lighters, who could refine it from there,” says Katherine Roberts, CG supervisor. “A lot of times we have a sparse set of HDRI images, but because the production was so good about letting our team on set, we got spheres for every set. That meant our lighting leads could generate rigs quickly.”

Even though the team derived the lights from the images taken on set, lighters could change the color and values using a classic CGI approach. To give the images a global illumination look, the team subtracted the extracted lights from the original HDRI images. “What was left when you take out the key lights is the ambient light,” says Amman. “So that gave us the combination of tweakable hero lighting and an overall, consistent relative brightness. The ambient could not be tweaked, but it didn’t need to be.”

To help the lighters work quickly on shots with many characters, the team used a proprietary system called Mob, originally developed as a crowd system, which reduces the need for geometry caches: Rather than importing geometry into a lighting scene, they can use Mob shapes. “It’s very light because there’s no geometry,” says Roberts. “But at render time, we can associate and swap the cache in.” For rendering, Double Negative uses Pixar’s RenderMan.

The same benefits—fast scene loading—applied to animators, as well. “The rigs were reasonably heavy,” says Martin, “so when there were too many characters in Maya, the scenes became unworkable. Mob gave animators a way to see lightweight characters, and the animators could swap between different levels of details. For example, they might have one character with a full rig, three with a lightweight rig, and two others as Mob. The amount of information in Mob characters is quite small. We just save out the joint information.”

Thern Blue

In the Martian world revealed in the film John Carter, a peculiar kind of nanotechnology plays a starring role. Called “thern,” which is also the name of a tribe, the always CG nanotech assumes many guises.  

“It’s like a Lego set of building blocks that can form into different entities,” says Simon Stanley-Clamp, one of four visual effects supervisors who worked on the film at Cinesite. “It can build a gun, a cannon, a control panel, a sword, a transporter.” Or, thousands of these blue, fractal-like therns can grow an entire sanctuary, which Cinesite created for 80 back-to-back shots.

 “One thern looks like a fiber-optic strand, bigger than a needle, more like a pencil, but short,” Stanley-Clamp says. “The tips have a bright light that dies off as soon as they form, and multiple threads interweave in patterns.” The sanctuary began as a corridor that grew ahead of people walking, on a soundstage, on an opaque lightbox underlit with white light. Eventually the corridor opened into a massive chamber made of thern.

 “It’s like a wicker basket, a tight mesh of interlacing thern. When a thern is fresh, it’s bright blue. When it cools, it goes through a color change to a slightly deader blue, almost gray. So, we had residual lighting within the basket weave to give a nice ambient glow to the thing.”

One of the first times you see the thern, it is a gun. “One character motions to give it to another character, and rather than passing it across, it climbs off one hand like a Slinky, across to the other guy’s hand, and then reforms around his hand as a gun,” says Stanley-Clamp. “It’s like a bit of armor; his hand is inside the gun.”

The Cinesite team designed and built the thern using Autodesk’s Maya, Side Effects’ Houdini, and custom software. An advanced Houdini simulation combining particles and fluids and augmented with proprietary software sent the thern from hand to hand when it formed the gun, along the corridor, and up the walls of the room to create the sanctuary—and elsewhere. The thern grow along an animated path using a proprietary vegetation-generating system. Because the algorithm for growing the thern is separate from the algorithm that created the geometry, artists could change the animation without re-generating the model for each iteration.

“Initially, we generate a model of the shape and then generate a ‘scaffold’ pass. We grow the thern onto this matrix and then animate it procedurally. Once animated, the geometry goes to lighting for custom glow and internal light passes, then to rendering and compositing. Conceptually, this was a difficult thing, to nail how the technology would be realized, what color it is, how fast it moves. But, this is the kind of stuff I love to do.” –Barbara Robertson

Creature Features

In addition to the Tharks, the animators also worked on several creatures: the 10-legged Calot, which Butler describes as a cross between a bulldog and a toad; a Thoat, which is a kind of horse, but with four legs on each side; and the White Ape, a huge, ferocious, gorilla-like creature with two sets of arms.

To block out the character Woola, a Calot that becomes Carter’s companion, the team first tried moving all the legs. “It looked like a centipede,” Butler says. “A bit silly. So we took the position that when we blocked a scene, we’d hide the extra legs and work with four. That gave us a good, biological, quadrupedal motion. It was more efficient, but more important, it kept our animators sane. Then if we had to tweak the extra legs, we did.”

Making Woola’s performance even more difficult is that he’s a heavy animal that needed to move fast. “When he’s in the desert following John Carter, his best friend, we needed all this motion,” Butler says, “But, no creature with his weight could move as fast as he does, so we had to bend the rules with him. In some poses, you only feel the compression and extension. He’s blurred with the sand. When he’s walking around, he looks like a large, tired animal.”

The Thoats has similar challenges. They have large feet and the leg structure of an elephant, but they live on a planet covered with sand and dirt. “That’s a lot of weight to move across the sand,” Aplin says. “We motion-captured a horse, but when we applied the motion to the rig, it looked exactly like a horse. Then we motion-captured a camel and it looked fantastic. The uneven gait worked better for this sandy world. We ended up adding tiny things. We looked at elephants and hippos, mixed and matched. You reach a point where science and references get you only so far.”

Planet Hollywood

While Double Negative and The Moving Picture Company artists had their hands full of character development and animation, Cinesite moved ahead with the environments, eventually creating four different areas of Mars in 800 shots: the city of Zodanga, the city of Helium, the air battles, and the thern sanctuary Sue Rowe led the visual effects effort at Cinesite.

In Edgar Rice Burroughs’ version of Mars history, and in the film, Zodangans and Heliumites, two cultures of men whose red tattoos depict their station and rank, have been fighting each other for centuries, largely in solar-powered airships with wings that spread for hundreds of meters to capture light.
Zodangans live in a mile-long metal tanker that crawls over Mars, a moving refinery that drills for radium, a disappearing resource. Stanton describes Zodanga as a city that “picks a spot, hunkers down, takes what it wants, and moves on. A majority of poor citizens live wherever they can within the superstructure, just trying to make do, and then you’ve got the few elite who reside up in the palace.”

Helium is the opposite. A solid city constructed of stone, grounded, with a Palace of Light in the center. “It’s white, pure, like a blue droplet of water in this harsh Martian environment,” Rowe says. To Heliumites, the well-being of its citizens is important. This is home to Princess Dejah Thoris, John Carter’s love interest in the film.

Rather than use matte paintings, Rowe had the artists at Cinesite build complete 3D environments. “I was keen about that after doing Golden Compass and Prince of Persia,” she says. “You need to build these things because of the scope you can get within a big 3D city. And, because there’s always a time when the director says, ‘Wouldn’t it be great if…,’ and we were able to accommodate that with Carter.”
When we first see Zodanga, it’s a power-of-ten shot through the clouds onto the surface. “Suddenly, you’re behind the city, looking at these holes in the ground where the city mined the resources and left it,” Rowe says. “When you go inside, it’s like Mumbai, a huge, bustling, overpopulated city. Then, when you go to the top, the quarters are all clean and beautiful.”

Cinesite modelers built Zodanga working from artwork by art director Ryan Church. To populate the giant vessel, the crew photographed people on set and used Massive software to control their movement.

“We shot on sets in the UK in winter, and every day it rained,” Rowe says. “The sets weren’t big enough. There were trees, motorways. I thought, How will I make this look like Mars? But we did a lot of preparation with greenscreen coverage. In a lot of the wide establishing shots, we removed the real set and replaced it with a digital set. That’s the advantage of building [complete 3D models of] these cities for real. We can swing the camera around and do what we want. One of the things I’m most proud of is the city of Zodanga.”

The production crew built the set for Helium in a warehouse in North London, which had enough interior ground space. “The problem was that it wasn’t very high, and the camera needed to look up 400 feet at a central building. So, working from Halon’s previs, I did a tech vis.” In the tech vis, given the camera and film format, Rowe calculated which lens would give the director what he wanted and where to place the camera.

Stanton shot much of the film in Utah, so Rowe and a few others came from London to shoot photo reference and watch the process. “It was the hottest place I’ve been in, but the environment was amazing,” she says. “My matte painters did a great job making sure the footage taken in London had everything we got from that red-hot environment. They used a lot of the HDRI photography and added their own touches, working in our Nuke environment, which can cope with these huge environments.”

Huge, in fact, describes the entire project. “It was the hardest and smartest thing I’ve worked on,” Rowe says. “We went from one city covered in dust, scarred, with metal gratings, to another made of glass, where we worried about getting lost in raytracing hell, and then we had a huge Master and Commander battle in the air. I’m glad I had done this for a long time before or it would have scared the pants off
me.” –Barbara Robertson

For the animators, working with a director who had come from the world of feature animation was a particular joy. “I learned a huge amount,” Butler says. “Andrew [Stanton] brought that Pixar legacy; it reminded me of my days at Disney. It was great to have a director who could talk our language. Andrew is skilled at getting the best out of people. He knows how to inspire without micromanaging.”

Aplin agrees. “He’d treat the animators like actors and would direct them like actors,” he says. “He’d try to get you emotionally involved in what the characters were doing. I knew we would learn an incredible amount working with Andrew Stanton, and we did.”

Cinesite created all the environments and airships for John Carter, including this fantastic,
solar-winged ship in the foreground for the Helium tribe, and the darker Zodangan ship behind.

Stanton is now one of many directors who have made the leap from live action to animation, and vice versa. But, these days, when it’s difficult to know what to call some films, it makes more sense than before: Is Tintin an animated film? Should Andy Serkis have been considered for a Best Actor award? Does it matter?

“There’s this assumption that I want to do everything animated, but I don’t go to the church of animation,” Stanton says. “The only reason you’ve never seen me do a movie like this is that I didn’t think I had the skills. But after 20 years at Pixar, I’ve become more knowledgeable about filmmaking in general.

“When I read the books, I would always picture a real person standing next to really tall creatures, and to do this movie, to realize that kind of imagery, you need to have CG characters with live action to make the whole idea work,” Stanton continues. “I think that’s what gave me the guts to [make the film]. When half the equation was something I was familiar with, I was willing to take the leap.” And, like John Carter himself, Stanton’s leap was extraordinary.