Jack the Giant Slayer

CGSociety :: Production Focus

7 March 2013, by Paul Hellard

Looking up at Digital Domain, the Lead VFX vendor on Jack the Giant Slayer.



Jack the Giant Slayer is a surprising VFX candy store of a production. A work of great scale, with an original story everyone remembers, while the screenplay brings a great range of comedy and drama to new heights. Literally.

Of course, there was an expectation of top quality in the finish, while there were also considerable technical and creative constraints. Digital Domain handled character development and animation on a large scale. There were eight hero giants and a hundred extra giants as well as a bunch of CG animals like pigs, horses, sheep and birds. Main giants included Fee, Fye, Foe, Fumm, Cooke, Sentry and General Fallon (two characters in one). DD also created secondary giant characters, many of which required facial controls to enable them to chant, perform limited dialogue and in the case of a giant called Bugler, blow a horn.

There was also fire, water, explosions, waterfalls, mist, clouds and rain to be generated, digital doubles on all the lead characters, for DD’s 250 artists, there was 461 shots.

The story gets moving when a brave team of humans climb the beanstalk, gets to the top and splits into two, determined to explore. They end up in a set environment called the ‘watering hole’. This is the first time we see one of the giants. “While it is a single giant, it really defines where we’re going to go for the rest of the movie,” says Steven Rosenbaum, Digital Domain's VFX Supervisor. “We were not holding back by any means for the creative and the technical. We have close up, personal contact with these giants, with dialogue, and even an extreme close up to the giant’s eyeball. Handling that kind of a shot in stereo is a hurdle. We had to dial the interocular to almost nothing, to avoid offending the eyes of the viewer.”

personal personal

One of the reasons Rosenbaum enjoys this scene with the waterhole is because it was no-holds-barred in terms of the technical hurdles DD had to get over. There are four giants in the initial group, (Fee, Fye, Foe and Fumm, of course). The giant Fee has laid a trap for sheep and some of Jack’s friends have been caught in a net. Jack is pulling them out of the net when Fee shows up. Jack has no place to run so he ducks under water.

One shot begins above water as Jack tries to hide. Just as Fee reaches down to grab the sheep, he dips below the water surface and the camera tracks down under the surface as well. “We see Jack holding his breath, still looking up,” Rosenbaum explains. “Fee’s hand plunges down and grabs the sheep. With all the water splashes and particulates, this is a scene never shown in 3D before. Going from above water, to below water, looking up through the water with foreground action as well, it was very tricky.

personal personal personal personal personal personal

“There was a lot of debate about building a real set and submerging the camera in real water for that initial Fee giant sequence,” VFX Plate Supervisor Swen Gillberg explains further. “But seeing the Plexiglass tank, the water and the size of the 3Ality rig for the stereo camera, there practicalities of this moved them to go straight for the CG option for water and the hair.”

The first stage was the performance capture for the chase out into the waterhole. The production design crew had designed a physical set for the humans’ component. Then an identical set a quarter of the size was prepared for the second stage, with the giant arriving in the same arena. This whole performance was captured in real time at Giant Studios' performance capture studio,” Rosenbaum says. “We were doing this in collaboration with Digital Domain’s Virtual Production Studio which handled facial capture.”

The next stage was to go over to the live action, full-scale studio and shoot a background plate. “We shot this native stereo using the Red EPIC cameras. We had Jack, played by Nick Hoult, lying down in a dry pond bed, with his hair slicked back, tight. The camera would come down behind Jack’s head and tilt up. We would replace the background, the water, insert distortion, even Jack’s hair is CG,” says Rosenbaum. The base shot was eventually captured with a Red EPIC camera on a Super Technocrane and a crew of five. In this scene, all animation was done in Maya, and the water simulation was in NIAID. Cloth and hair solving was done in Maya. KATANA was used for all the lighting and Arnold was the renderer. “This was the only time I’d ever strayed from using RenderMan,” Rosenbaum explains. “The idea of using a true global illumination solution, to have a really nice ray tracer at our disposal and really nice HDRs made it really straight-forward for us to light and set up the photo-real result for this shot.” There was no real water involved.

personal personal personal personal


VFX Plate Supervisor Swen Gillberg was working the location VFX coverage out on an old estate in England. The castle was recreated in CG by SohoVFX. Here, there was knights riding, giants running and a large-scale chase up to the castle. While a previsualization of the chase was done a year earlier, the edit was changed the night before the shoot. It was Gillberg’s job as Onset Supervisor to sit with VFX Supervisor Hoyt Yeatman and the second unit director to make sure there was enough plate material to create a working edit, and appropriate HDRIs to gather lighting details on the shots. “We had stuntmen running around with giant-height eyeline poles to help frame giant’s heights for the background plates,” Gillberg remembers. “P&C did the meter point LiDAR [Light Detection And Ranging] on the entire area, and this was a great help.” This magnificently appointed property was taken over by huge camera cranes, low-loading camera support trucks and the place was strewn with a bunch of bright orange volleyballs. It must have made quite a sight.

”The stereographer on Jack the Giant Slayer Chris Parks, and Bryan Singer are both firm believers in wide lenses, which made the job a little easier in that most shots could then hold a six foot man and a 24 foot giant,” Gillberg adds. “A wide lens allows more light, so you can crank down the aperture, bringing a long depth of field, so everything is very sharp." Parks came up with something called ‘Giant Vision’ - extremely wide IO, set up at 18cm instead of the normal two and a half, which makes everything look extremely small.


Animation Director Jan Philip Cramer was at Weta Digital before Digital Domain and was asked to join the crew with Stephen Rosenbaum because he worked with him on Avatar and was excited to do another creature film together. The brief was ‘strong, fast and agile’ giants from Bryan Singer. “We tried all kinds of tests with stunt guys with weight-belts, and ankle weights and to give the actor a good feeling for that. They were doing a lot of experimenting with motion coach Peter Elliott.

It was critical for Rosenbaum’s team to have an accurate understanding of the 20-to-24-foot-tall CG giants’ performance, position and timing while shooting scenes where giants drove the primary action and interacted with live actors; as well as for the all-CG and mostly-CG sequences featuring giants. The Giants are mocapped and every mocapped movie has a lot of animation that goes on top of it. For each of the eight hero giants DD captured approximately 120 facial expressions using MOVA. When those were combined, each hero giant ended up having between 1,000 and 2.000 face shapes. Fallon, with two heads, had 4,000. Cramer’s main focus in animating giants was the 350 or more shots. General Fallon is the giant with two heads, has about 20 minutes of screen time and a lot of dialogue. There was a lot to General Fallon, beginning with the two actors that played him. John Kassir and Bill Nighy. The actors played the same character in the same shot, together, and then later, the animators putting the two performances together, matching two spines working in tandem.

personal personal

Bill Nighy is a delicate gentleman, unhurried. John Kassir is high-spirited, bouncing around with a joke every minute. Pairing these guys together worked a little like the comedy duo of a ‘comedian and a serious guy’. “Matching it with mocap done by the two and with virtual camera blocking, they could see their own performances on the giants. This allowed the two actors to play with what their giant would be,” says Cramer. “It was a fun experiment that allowed the fullness of the performance through seeing the final action on screen. John on one side as the small right head, and Bill as the big head on the left. They very quickly got used to working as a double virtual body with dual spines. It was a pleasure to sit with the actors, and let them take time to develop their giant.”

The approach here was to design the giants as characters, building the actor’s characteristics and trademarks into the characters. “We had multiple companies working on this,” explains Jan Philip Cramer. “We had Giant Studios doing the body mocap, retargeted with our actors. DD did the face capture, using an evolution of the face camera used on TRON: Legacy, but modified quite a bit. The DD unit collect all the face marker data, and we would solve them onto our character rig.” This face solver is something that DD has worked on for a long time. Back on Benjamin Button, then Tron: Legacy. This version allows interaction with the animator, where they can solve and animate at the same time. They can be working on the mouth, and still be able to solve for any other part of the face, without losing anything.

personal personal personal


In a motion-capture shoot prior to principal photography, sequences driven by giants were choreographed and directed by Bryan Singer; Giant Studios capturing all body performance and Digital Domain’s virtual production team capturing all facial performance (using four mounted face cameras) – simultaneously as a single, cohesive performance, in real-time in 3D.

Digital Domain’s virtual production team then projected video from its face cameras onto the giants’ characters, a process they called ‘Kabuki’. Bryan Singer and the editor Bob Ducsay could then see the actors’ performance quickly and cost- effectively within the realtime version of the captured performance. Digital Domain delivered the Kabuki asset to Giant Studios where it was integrated it into the real-time scene, which could then be viewed in real time in motionbuilder or played out as QuickTime files for editorial. To ensure all motions maintain the weight & scale of giants, most motions were slowed down by 15-20%.

Those cohesive giant performances were then used for on-set references during principal photography where they were viewed by Singer, DP Tom Sigel and VFX Supervisors through the Simul-Cam or viewed as Quicktime references, so they could see the giants’ performance, scaled correctly (at a factor of four) alongside live actors, within the CG environment. In addition to providing accurate on-set info, this process also created great time and cost efficiencies in the post-visualization stage as Digital Domain was able to respond quickly to directorial and editorial changes in giants’ position, timing, performance, eyelines etc.


With all these giants about, staring down at the humans, and the POV of the humans staring back, the eyes were a particular challenge. DD in Vancouver made sure the animators were completely across the physicality of eyes. They brought in biology professors to give them lectures about how the eye works. How the muscle behaviour made eyes dart around a certain way, for instance. Both Jan Philip Cramer and Steven Rosenbaum had spent the last three years working on exactly this problem, so the animators had a lot of support in their job to make the eyes all look as real as possible. “Eric Petey - Senior Anim Lead; Mark Della Rossa, and Clement Yip and Jee Young Park in their roles as Animation Leads really worked some magic into their work,” says Cramer.


The giant jumping on the beanstalk presented the two VFX studios MPC and Digital Domain with a challenge not attempted in such great detail. MPC animated the beanstalk and Digital Domain worked on the giants. This project required Digital Domain to have a very tight cooperation with MPC in Vancouver and Greg Butler from the MPC side worked very tightly with the DD crew. As the two animation rigs were proprietary, the two companyies' animation crews couldn’t share exactly the rigs they in turn were working on. “What we did was create intermediate rigs that were not proprietary that we could then pass back and forth to the MPC crew,” explains Cramer. “These were Maya based, so they could then be plugged into our proprietary rigs back inside our pipeline here at DD.”

Giants were climbing the beanstalk, being wrapped in the beanstalk, so there’s the tightest interaction in the two. “When we show our work, we’re showing the other studio’s work as well,” adds Cramer. “I’m really thankful to these guys. It was hard work but it was a lot of fun.”