John « DJ » Desjardin began his career in visual effects 20 years ago. He worked on projects like TIMECOP, BROKEN ARROW, RELIC or MISSION: IMPOSSIBLE II. He becomes Production VFX Supervisor on the two MATRIX sequels. Then he took care of the VFX for films such as like FANTASTIC FOUR, X-MEN: THE LAST STAND or THE KINGDOM. MAN OF STEEL is his third collaboration with director Zack Snyder.
What is your background?
Wow! How far back do you want to go? Let’s see… I’ll say I had my first job in the VFX field way back in summer of 1984. I was an apprentice to a small group of VFX geniuses who worked on BLADE RUNNER, STAR TREK: THE MOTION PICTURE and CLOSE ENCOUNTERS OF THE THIRD KIND. I learned from them, and rode the wave that took us as artists from analog to digital. After a ton of hands-on experience, I eventually became a VFX Supervisor. My first production-side VFX Supervisor job was on the MATRIX sequels, working directly for the Wachowskis which was a very satisfying lesson in well-planned VFX techniques and shot design. Years later I was very fortunate to meet Zack Snyder for WATCHMEN, and I’ve been very lucky to work for him and his very talented film family ever since.
This movie is your third collaboration with director Zack Snyder. How was the collaboration on this new movie?
As always, collaborating with Zack and his key artists is (more than anything) FUN. Zack is extremely inspiring, and as a result we all end up inspiring each other. After all of this time the conversations tend to be short yet meaningful while the pop culture references keep flowing. There is a point to all of that, too, but it translates into a fun experience for me.
What was his approach about the visual effects?
Zack was clear from the start that while our previous movies together tended to be more stylized when it came to the visuals, this one needed to be as realistic as possible. Even though it was a very sci-fi take on superheroics, he wanted it to be shot hand-held and always at 24 frames per second (« the speed of life », as Zack says). No slo-mo, no complex camera rigs (no technocranes or fluid dolly moves).
Did he gave you specific indications for the visual effects sequences?
Yes. Zack is VERY specific with what’s in his head. He draws his own storyboards and is very specific as to what needs previs help and what doesn’t. He is also very forthcoming as far as any new ideas he has for sequences. He gets the information out to his key team members as soon as possible, which is great because it keeps us all current and on the same page every step of the way.
What was your feeling to work on a Superman movie?
I didn’t have an answer to that for the longest time. I was excited by the artwork and the script and Zack’s ideas, but I didn’t have an immediate reaction to the fact that I was working on a new take on an iconic character, one that is sometimes referred to as « The First Superhero ». But there was a moment on the final day of shooting. It was about 3pm and we were all nestled in a valley near Bishop, California. It was a sunny afternoon in early February, and we were shooting the scene where General Swanwyck confronts Superman who has just trashed a Predator Drone. I was looking at the scene being filmed and talking quietly with my very good friend and stunt supervisor, Damon Caro. And I think it hit us both at the same time as we looked at Henry Cavill in the suit, standing there as Superman in front of this crater containing the destroyed Drone. Damon and I looked at each other, and I said to him, « MAN! This is really IT. We’re making SUPERMAN! » That was the chill that went up my spine as the realization sank in. It took the entire seven months of shooting for that moment to hit me, but when it did, it really got me.
The movie has a fresh art direction with Krypton and the spaceships. Can you tell us more about this aspect?
Production Designer Alex McDowell and his team were charged with coming up with the basic concepts for Krypton and the spaceships. The script by David Goyer and Christopher Nolan contained some details about that world. Zack also gave notes to Alex about the history of Krypton. I inherited the concept paintings and Zack’s notes and worked with Dan Lemmon and his team at Weta to flesh out the world. We tried to show Krypton as a world in decay, as a planet and as a society. The city was mostly underground and consisted of strip-mining striations overhead and various structures below. There are three types of ships. Hammerheads are the big, Star Destroyer-type ships that slowly cruise around the city and countryside. These are stately sailing vessels. Scarabs are the Millenium Falcon-type ships which move fast and escort the Hammerheads. Then there are Zod’s Gunships which contain a blister of various guns and a troop-carrying cargo bay. All of these ships can be seen in the establishing shot of the battle as Jor-El runs out onto the Council Chamber platform. Hammerheads blast away at various targets while Scarabs and Gunships dogfight over the landscape.
The remaining ships to mention are Kal-El’s Starship (which carries the baby to Earth), the Black Zero, Drop Ships (very insect-like), the World Engine and the Scout Ship which we later refer to as the Fortress of Solitude. We tried to represent two eras of Krypton in the design of the ships. The most modern ship in the movie is the Black Zero. Jor-El helped build this ship to imprison Krypton’s criminals. It works in conjunction with the Phantom Zone Generator or Projector which hovers in orbit over Krypton waiting to receive the prison ship. The lines are very clean and while the curves are reminiscent of other Kryptonian ships it is devoid of ornamentation.
The other ships mentioned that fly around the city (Hammerheads, Scarabs and Gunships) are from an age between the space faring years and modern Krypton. They have elements of the Scout Ship ornamentation, similar lines and materials but are more functional.
The Scout Ship is my favorite. It’s almost like a cathedral, a sculpted work of art crafted by a more elegant and enlightened society intent on exploring the universe. This design represents the older, bolder Krypton that Jor-El yearns for at the beginning of the film. The Star Craft that transports Kal-El is a nice bookend to that period, a throwback to that space faring age built by Jor-El in his modern, dystopian era.
All designs are based on underwater creatures or insects, right down to the materials used in the ships. The hulls, wings and engines look like corals or crab shells or beetle shells.
As I mentioned before, Krypton is a dying world. It is that way because of its short-sighted people. The Kryptonians seem to be thousands of years old. Their technology is very old — their robot servants, weapons and armor and some ships are thousands of years old. It is mentioned in the story that Kryptonians were a space-faring race but abandoned their outposts after some ill-defined time or event. We made sure to add ancient city ruins to the various landscapes seen in the movie. Unlike past renditions of Krypton’s story, Krypton’s sun, Rao, is NOT a Red Giant Star responsible for Krypton’s demise (as it was in the 1978 Richard Donner film). It is a Red Dwarf Star. I ended up reading a lot about such stars and discovered them to be the most plentiful type of star in our Universe. They’ve been around since shortly after the Big Bang and no one has seen evidence of them dying out. They burn cooler than most other stars — including our own — so they don’t use up as much of their fuel. It is thought that if there were life-sustaining planets around such stars, they would be closer to their Red Dwarf Suns. There is also a chance that the planets would not rotate on their own axes, so there might be a ring of a sweet spot where life would grow. It is also possible that a decent atmosphere could conduct the heat of the sun around the non-rotating planet. Still, there would be a constant day side and constant night side. But I made sure to place our city in the sweet spot as a way to render Rao as large and reddish near the horizon, like a constant magic-hour look to the skies. Visible in the constant twilight sky is a broken moon, which is a part of Superman comic book lore.
When it came to exploding Krypton, Dan Lemmon and I had long discussions about how it might look different from other planetary explosions. Fed by lines in the script about what the Kryptonians were doing to their planet’s core, we decided that the planet should basically give birth to a failed star. In an effort to ignite their core to provide a new energy source, the Kryptonians compacted the matter at the center of their planet. We thought maybe they were using one of their terraforming machines — a World Engine as seen in the Indian Ocean at the end of the movie — and the result of this process was the collapse of all of the layers of their planet. Beams of energy cut through the crust, shooting towards space as the star at the center radiated outward. Cutting to space to see the planet-wide results, you can see the layers collapse around the bright areas of the planet until it explodes but settles immediately into a debris field held in place by the neutron star at the center. That star is the glow at the center of all of the rocks and dust. That is what Zod and his team see when they awake in the Black Zero after being shunted out of the Phantom Zone by the broken generator.
What were your thoughts while reading the script and then had to put these huge sequences on screen?
My immediate thought was « WOW! That’s pretty cool! » I was happy to read a script that was such a fresh take on an all-too familiar icon. Once I’d settled down from the initial excitement of reading something new I got to work breaking it down into digestible pieces for budgeting and to figure out the technicalities of the work. In other words, I quickly lost sight of the hugeness of the project and took it little by little. It’s like climbing a mountain and concentrating on that small area of rock surface in front of you and not worrying too much about how high you’ve climbed or how far up you have yet to go. No matter what, you do finish eventually.
How did you work with Zack Snyder and the previs team?
Working with the previs team is a daily affair for me, part of my schedule during prep and during production. We even keep most of the team on during the first few months of post-production to do some postvis and generate a certain percentage of temps needed for Zack’s Director’s Cut. Generally during prep and production, all of the vis starts with either Zack’s drawings or Damon’s stuntvis. It’s my job to work with the previs team to work on fitting the action VFX pieces into the location or set photography. We also do some shot design, again primarily generated from Zack’s drawings or Damon’s stuntvis. One main goal Damon and I developed through SUCKER PUNCH and improved upon for MAN OF STEEL is the combination of stuntvis and VFX previs and editorial to create a visual shooting bible for the crew. Damon and I worked closely with our previs supervisor Kyle Robinson to make whole sequences that designated what should be shot and what was CG; what was green screen, what was location, etc. We made very well-marked quicktimes of scenes like the Smallville Battle, the World Engine, Metropolis Under Siege and Superman Vs. Zod. Bruce Moriarty (our 1st AD) would usually get these first as an early guide from which to schedule our shoot. The quicktimes would then get distributed to the keys in the crew so everyone would know what was going on. I would say that this tool has become very important to the way we collectively get through some very technical portions of the shoot.
What was a typical day for you on-set and then during the post?
A typical day on-set starts very early in the morning either on location or on a sound stage. I would participate in a walk-through of the first setup with Zack and the crew, then I would meet with my VFX crew and download what I thought was necessary for the day, including how to split up our presence because most times we had a main unit and splinter unit going at the same time. Often I would work directly with Damon Caro on the splinter unit which covered a lot of the Superman fight and flight action while Zack was on another stage working through dramatic scenes with other main actors (like Russell Crowe). Many times the on-set work required attending to previs for upcoming scenes or walk-throughs of other sets or locations. A lot of my work day can be spent making sure things are in order for upcoming shooting. I try to stay a couple of weeks ahead, more if there’s heavy action involving VFX work. My VFX team is there to handle the setup-by-setup data gathering, which is the most fundamentally important work on the shoot. And I don’t even do that! I just approve of what we’re trying to get on any given scene.
A typical day during post consists of REVIEWS REVIEWS REVIEWS! My morning starts with internal VFX dailies. That’s me looking at everything that came in overnight from our teams working around the world. Cinesyncs — basically, teleconferences where I and a facility team can talk to each other, go over notes and quicktime dailies which we can both draw on and detail — occur throughout the day. I would talk to Ged Wright at Double Negative (London) first; then Chad Weibe at Scanline (Vancouver & LA); then Guillaume Rocheron at MPC (Vancouver) usually in the afternoon; and at the end of our day, Dan Lemmon and Keith Miller at Weta (New Zealand). Sandwiched between all of that we would also have reviews with Zack and David Brenner, our editor. That is the basic day-to-day that ran from mid-February 2012 until mid-February 2013.
Can you tell us more about the digital doubles creation?
Digital Doubles were created for all of the major fighting characters in the film. The process is probably similar to many other films. We cyberscan each character and do FACS sessions for each of them.
One difference is how we handle capturing data for the doubles on-set rather than relying on other performance sessions outside of the regular shooting process.
The theory we’ve followed since realizing Dr. Manhattan as an all-CG character in WATCHMEN is we try to get as much lighting and performance detail as possible while shooting the scene.
In the case of Superman fighting in Smallville, we will shoot a sequence of the fight normally, with everyone in the moment performing Damon’s stunt choreography. If there’s a section that we feel will handoff to CG, once Zack has an overall take he’s happy with, we’ll bring in a piece of equipment called the Shandycam (named after Shandy Lashley, our VFX coordinator/ data wrangler whom I tasked with building the two versions we’ve used).
The Shandycam is a 6-camera array on a lightweight pipe rig which is placed about 6 feet away from the subject. The rig has 6 Canon 5D Mk II cameras — 3 on top, 3 on bottom — with the left and right cameras panned inward to provide coverage slightly side-on to the character. These cameras are synced and remote-activated. We would use this rig as the final bit of photography of a setup. In this case, we would run Henry through some of his performance at a reduced speed as we snap away synchronized stills with the Shandycam. The advantage is we get Henry Cavill’s performance — body and face — at an extremely high resolution, one that exceeds what we get with the motion picture camera. The lighting and emotional moment of the scene are maintained because we’re capturing while remaining in that moment, as opposed to trying to re-create it in a vacuum down the line. This image data is fantastic for projecting onto our match-moved CG character to ensure we’re getting the most photographic, on-set reality out of it.
How did you approach the flying sequences for Superman?
Our approach to Superman flying was dictated by Zack’s direction for the photographic style of the movie: documentary, cinema verite, hand-held 24 fps. In other words, we always tried to have a reason for where the camera was and how it was operated when it came to showing Superman in flight. Zack first showed us what this might be like when we shot our test for the Kryptonian fight scenes (see « biggest challenge » question below). We put one of Damon’s guys — Ryan Watson– in a belly pan rig. Ryan laid down and put his arms out Superman-style, then Zack grabbed the camera and got really close to Ryan, just to his hands outstretched. Zack shook the camera while he moved and twisted around Ryan, always close to him. It provided the feeling of a ton of high-speed energy.
That simple video became our model for how we approached the flying shots.
Still, there were times when we attempted to emulate the original Richard Donner movie’s flight. There were a few boards that Zack drew which we prevized (and ultimately shot elements for) that showed Superman flying as we tracked with him, even dollying in on his face. This was exactly the style of the 1978 Superman film which relied on the Zoptic technique for Supes’s flight. (Zoptic is a rear-screen process in front of which Superman was hung on an articulated wire rig, finished flight imagery playing on the screen which is synced to the film camera on a boom or a crane arm. This rig implied we were always tied to Superman or at least operating within a few feet of him, magically booming and dollying as we flew with him in front of the rear screen projection.)
Once we saw these prevized and postvized shots in the edit, we realized they just didn’t work with the visual language of the film. We had to re-imagine them in the style Zack demonstrated to us in that original video test.
When Superman flies, the camera is either really close to him (Henry in the belly-pan against green screen) as if it’s attached to him somehow; or we’re far away from him on a plane or a helicopter or even a satellite, trying to zoom and focus to get a glimpse of him. Because of this, in most cases, he is CG as he flies through the often-CG environment.
The movie features a lot of different environment especially on Krypton. What was the real size of these sets?
For Krypton, the largest sets (50+ feet in diameter) were the Council Chamber where Zod explodes into the scene and executes one of the Council Members; the Black Zero Bridge; Jor-El’s Lab; and the connected interiors of the Scout Ship. All of these required some kind of CG set extension and/or additional CG set dressing.
How did you help the crew and the actors to visualize the effects?
The same production-guide quicktimes we generated for the crew (as illustrated above in the previs team question) were good visual tools to show all of the actors what the scene was all about. Zack was also very descriptive as to what was going on. I would sometimes have concept art on-hand or any animation or look development tests on-hand to flesh out conceptual details. For example, Double Negative sent me a very good test of CG Zod in armor walking around. I showed that to Michael Shannon when we were shooting his scenes on the Black Zero to give him a good idea of the volume and look of the armor he would be wearing in the movie instead of the performance capture suit he wore on set. Those were the kinds of tests we tried to get in front of the actors whenever Zack thought it was appropriate to give them a better self-concept of what was going on in a scene where they had to visualize quite a lot when very little set or hero costume was before them.
Can you tell us how you proceed to get all the necessary informations for the VFX studios?
That’s a combination of my production-side VFX team and our cinesync dailies sessions. The studios send us their work throughout the day. Our I/O supervisor — Arnand Kularajah — grabs it from them and notifies our VFX editor — Craig Smith — and our coordinators and production manager — William Marlett, Ana Marie Cruz, Robin Williams and Matt Magnolia — who work together to organize it for my producer (Josh Jaggars) and me to examine during our internal dailies sessions. Josh and I give comments that our team takes notes on. We generate more notes in our cinesync sessions with the VFX studios. And then we generate MORE notes during our almost daily reviews with Zack and David (our editor). All of these copious notes are entered into our show database and distributed to the various VFX studio teams. That way, we always have a history of every shot on the show which we can refer back to if there’s ever a question about a change or a comment as to how to proceed.
The final sequence involved a huge amount of CG and destruction. How did you approach this sequence?
The first thing I did was have a conversation with Ged Wright (supervisor at Double Negative) about creating a procedural city generator, on the order of what Sony Imageworks had created and honed over the years with their work on the SPIDER-MAN movies. Ged’s primary task was to build Metropolis so that we could fly around it with our superhuman characters and also destroy it and show its layers.
Metropolis, as described by production designer Alex McDowell, was comprised of three giant city centers, with characteristics of LA, Chicago and NYC. We were scheduled to shoot in Chicago for our primary city location work. The VFX Team executed a massive data capture session over a couple of weeks. We had ground-based LIDAR moving through the canyons of the city. We had a Spacecam-mounted LIDAR rig on a helicopter flying over streets. We also had teams going up into buildings to get LIDAR and photography throughout the city. For the Indian Ocean portion of the destruction, we sent a splinter unit to Bora Bora to get some hero and survey views of the island and the ocean.
All of this added up to fantastic reference for Ged and me, because we knew that when the World Engine made its crash landing on Bora Bora, and when the Black Zero started pounding away at Metropolis, huge portions of the motion picture photography would have to be replaced with massively simmed and rendered CG destruction. While the camera work was meant to be very realistic and hand-held or aircraft-mounted, what was happening within the frame to the city or the Indian Ocean was on a massive, fantastic scale. So we knew at the beginning of the shoot that these sequences were destined to be heavy CG work.
We still tried to shoot as much as we could. We flew camera helicopters above Chicago and attempted to replicate the previs camera moves we’d created back in our production offices. But we always knew these images would be reference for the look and shape of our CG camera moves over our CG city.
Once we got into the Superman vs. Zod fight, it quickly became evident that the backgrounds were in such destructive flux and our Kryptonian adversaries were moving so fast and creating huge shockwaves of force that we needed to shoot our main characters in a green screen environment. And of course once they started flying and fighting, then we moved into all-CG territory. Still, we always tried to ground some part of the image with real photography, even if we were just projecting Superman’s and Zod’s faces or body details for a few frames during the fights. Those grounding elements always help maintain realism during over-the-top action.
How did you split the work amongst the VFX studios?
We divided the work in the following manner:
Weta – Krypton, Black Zero in space, Drop Ships in space, Outpost Planet, Kryptonian ship interiors.
Scanline – Oil Rig Rescue, Tornado Scene, X-Ray Vision FX.
MPC – (Initial R&D for flying and fighting, cape R&D) Arctic Sequences (including the Scout Ship reveal and Zod’s visit), Superman learning to fly, Superman Surrendering (including Edwards AFB first contact), Escape Pod Rescue & Smallville Battle.
Double Negative – Black Zero over Metropolis / World Engine over Indian Ocean, Superman vs. Zod.
LookFX – Bus Rescue Scene, Video Screen composites.
Teamworks – All kinds of fix-it stuff.
What was the biggest challenge on this project and how did you achieve it?
The biggest challenge in my mind from the beginning was how to visualize the superhuman fights. It was the main thing we actually tested during pre-production in LA before moving to Illinois in Spring of 2011. I sat down with Damon Caro, our stunt supervisor and splinter unit director, and dissected his stunt viz of the Smallville Battle with him. We picked a fairly complex section of the street fight (Superman vs. Namek & Faora) and attempted to photograph it how we thought we might on the streets of Plano, Illinois (which was the actual location of Smallville in the movie). Zack, Damon and I agreed that the Kryptonians needed to move and fight fast — with a lot of destructive force — so that a human cameraman might have difficulty keeping all of the action within his frame.
We already had a decent, lo-tech way of getting what we needed on-set for our Real-to-CG doubles (as illustrated in the above Digital Doubles question). The freedom to switch out Real-to-CG characters was not the problem. What became evident was we needed a camera that could preserve the look of the location at the camera operator’s (John Clothier’s) position for each setup. Looking online for such a device, Guillaume Rocheron and I decided to try the Roundshot camera rig. It’s simply a Canon 5D Mk II with a 50mm lens mounted to a motorized head. It takes 74 stitchable tiles, each with extremely high resolution. The resultant environment sphere is so hi-res we found we could pan, tilt, dolly a bit and zoom within the captured space, which was inherent to the photographic style of the movie.
The « Enviro-Cam » has become a regular part of our digital character & set data capture process. That piece of the location work puzzle allows us to roto out any real characters from the photography that need to be taken over and override the film camera when we need to take a portion of the shot into completely digital territory. Since all of this capture occurs on-set at the time of the actors’ performances, all of the lighting and emotion is preserved, negating the need to re-create it in a vacuum at some later shoot date.
These simple techniques allowed us to further the energy Zack and Damon wanted to create in the MAN OF STEEL fight sequences.
Was there a shot or a sequence that prevented you from sleep?
Not really. I sleep pretty well these days. Some of my best ideas come just before I fall asleep or right when I wake up.
What do you keep from this experience?
I take away some very good working relationships and friendships from this project. It was once again very satisfying to be a part of Zack Snyder’s creative team. I had a fantastic experience with all of the VFX studios and would like very much to keep working with them on future films. It is always nice to continue to build upon the research and development we explore on a show. Furthering that growth and our techniques is what makes working on a movie FUN.
How long have you worked on this film?
About 2 years.
What was the size of your team?
Our team varied in size from about 10 during pre-production to 20 during production down to about 7 during post-production.
What is your next project?
Let’s just say I’m wrapping up work on Zack’s 300: RISE OF AN EMPIRE, surfing in Bali for a month, then returning to LA to commence work on The Untitled Warner Bros. Project.
What are the four movies that gave you the passion for cinema?
2001: A SPACE ODYSSEY, BLADE RUNNER, STAR WARS and EXCALIBUR.
A big thanks for your time.
// WANT TO KNOW MORE?
MPC: Guillaume Rocheron, VFX Supervisor
Scanline VFX: Chad Wiebe, VFX Supervisor
Look Effects: Max Ivins, VFX Supervisor
Pixel Liberation Front: Interview of Raffael Dickreuter, Virtual Camera Supervisor
© Vincent Frei – The Art of VFX – 2013