Edge of Tomorrow, Doug Liman’s big screen adaption of All you Need is Kill is a film about trial and error – Major William Cage (Tom Cruise) is killed within minutes of his first day of combat, only to get trapped in a time loop that allows him and United Defence Operative Rita (Emily Blunt) to live, die and repeat their struggle against the formidable Mimics again and again.
Our Art Department and VFX team both started work on the project in late 2011, ready for some trial and error of our own when tasked with helping design not just the soldiers, but also the Mimics: superfast, almost indescribable creatures with a tangle of tentacle-covered retractable limbs.
Warning: spoilers ensue
We had to explore how they would move, how fast they were and how they would kill each other – all essential information that had to be determined before anything could be shot. Our visual effects, which account for more than 400 shots, include the film’s fiery finale in a flooded, fully CG Paris, requiring large-scale environment work and impressive simulations.
“With Edge of Tomorrow we took on everything” says VFX Supervisor Jonathan Fawkner. “Fluids interacting with themselves, close-up digi-doubles, extremely complicated character movement and massive fluid sims – we were asked to do every trick in the book and we rose to the challenge.”
We started with visual development and movement studies for the soldiers, but we soon had to consider the Mimic’s too. “We had a sort of arms race between the aliens and the soldiers,” explains Fawkner, “because if the aliens can move fast then the suits have to make the soldiers move fast too, otherwise why would they have built them?”
The basic Mimics, glowing yellow and with four limbs, each of those covered in a number of tentacles, are led by the larger but rarer Alpha Mimics – glowing blue, six limbed and again covered with extra tentacles. Both creatures’ limbs and tentacles can grow and retract as they move. The complexity was reflected in the rig, which had up to 2000 controls for the animators to play with.
“We had to find ways of making them not look like a traditional quadruped” says London Animation Lead Brad. “You never really see all of their limbs on screen at the same time because we cycled them so as one disappears another grows. All together it made it quite a complicated creature!” Aside from the Mimics the show’s main animation challenge was actually the sheer range of work – huge dropships, soldiers, helicopters, close-up digi-doubles and, in the fully CG shots, often the camera too.
The CG helicopter makes an appearance over the Alps, which we constructed using plates shot by production and digital matte paintings (DMPs), projected in Nuke. Integrated into the mountains is a dam – actually the roof of a sewage treatment plant in Essex. Inside, the Mimics get rare close-up hero treatment. “The film is a little bit like Alien, where you don’t see the creature too much, but what’s nice about this sequence and the one later on in Paris is that there are some big shots where you really see it” says Montreal VFX Supervisor Christian Kaestner.
Our major work comes late on in the film (so please don’t read on if you haven’t seen it yet) as Cage and Rita travel to Paris, now a dead, dark and flooded city. We animated the dropship that takes them there and topped up the set for the inside. Everything you see below as they drop is digital Paris, miles and miles of it, filled with water, smoke, fire and flying Mimic javelins – the fiery bolts shot by the aliens to bring the dropship down.
With Paris flooded he was unlikely to make it onto dry land, but Cage manages to drop straight into the Seine. There’s a digi-double take over here, flanked by two genuine takes of Tom Cruise holding his breath for some considerable time. He pulls himself out of the water and into our dark, digital Paris – the Place de la Concorde, the Tuileries, the Louvre and the surrounding area.
Lighting what was supposed to be a powerless city believably was a tricky balancing act. It had to look real, but then if it was really real you would barely see anything. We made it work with the clever placement of fog and fire. “We illuminated the fog as if lit from the sides, off camera, always trying to make it look like ‘movie night’ rather than real night, which is impossible to photograph” explains Fawkner. The water, fire and fog were all done in fLush, our in-house fluid simulator. It was a real bi-directional, sometimes tri-directional simulation. The lighting was done in Arnold using bespoke fMote fog and volume shading. Find out more about Edge of Tomorrow's technical challenges here.
Those water simulations needed to cover such a large stretch of Paris (it is over a kilometre from the crash site to Louvre) that they are biggest done so far at Framestore. Not only was there water, but also a restarted dropship, hundreds of pursuing mimics and, of course, plenty of explosions, all of which need to react with each other, creating detailed splashes, foam and spray.
In some shots the dropship has been filmed on greenscreen with the environment added later while others are fully CG, with several of those including close-up digi-doubles for Tom Cruise.
Surging along at top speed away from the mimics, the dropship has its wings clipped by the Arc de Triomphe (du Carousel, not the larger and more famous Arc de Triomphe a few miles down the road), creating a violent marriage of fire, water and destruction. It’s a spectacular shot that required us to make the fire and water work as one, interacting with each other and lit and simulated together.
Just in front of the arc lies the Louvre Pyramid, crystalline, glittering and built to an incredibly high standard. It is of course next on the dropship’s wish list of destruction. It’s a full simulation with lots of interactive lighting. The falling glass, debris, water and of course Tom Cruise as he flies out of the ship are all digital. “It was a digi-double pickup halfway through his flight, though the bit at the end, the bit that hurts, is all him” says Compositing Supervisor Matt.
From their Cage and Rita make their way into the Louvre’s underground carpark to take on their last challenge. Like the scenes above ground, this takes place in an often majority digital environment. “The partial set was used very wisely for what it was meant to be, which is this quite confusing, repetitive space, that’s dark and moody so you’ve don’t always know where everyone is, but for the wider shots we had to turn it into a big, moody carpark environment” explains Montreal VFX Supervisor Christian Kaestner. It was a big model build with lots of DMP work and more inventive lighting solutions. Some shots are almost all CG, with around a quarter of the environment being real and three quarters of it generated.
Find out more about Edge of Tomorrow's technical challenges here.