Join us behind the scenes as VFX Supervisor Tara DeMarco pulls back the curtain going through the creative process step by step of the visual effects work (made by ILM, RISE, Rising Sun Pictures, Sony Pictures Imageworks, Trixter, Weta FX, Wylie Co.) for the Marvel Studios film, The Marvels:
WANT TO KNOW MORE?
Tara DeMarco: Here’s my interview of Production VFX Supervisor Tara DeMarco.
Disney+: You can watch now The Marvels on Disney+.
© Vincent Frei – The Art of VFX – 2024
Back in 2018, Alex Wuttke provided insight into ILM‘s visual effects work on Jurassic World: Fallen Kingdom. His journey continued with Six Minutes to Midnight and then leading him to become a part of the Mission Impossible universe.
How did you get involved on this film?
I met with Chris McQuarrie early in the lifecycle of the project, before pre-production had begun. He was starting to scout potential locations and we talked about his approach to crafting action sequences, about the ways that location informs action. We talked about the films I had worked on previously and the takeaways that could be brought to bear on the project. We started pre- production soon after.
What was your feeling to be back in this iconic universe?
Obviously very excited! I’m a huge fan of the Mission series, so the opportunity to be a part of the franchise was something I didn’t want to miss out on.
How was the collaboration with the Director Christopher McQuarrie?
Fantastic. Chris is an incredibly accomplished storyteller, and every frame within the film plays a part in telling the story. He would be very active in the blocking and framing aspects of our work, which was a fantastic education. In terms of practical shooting methodology, he was very much across the details, but ultimately happy to defer final approaches to the HODs.
What was his approach about the visual effects?
On these films, we like to shoot as much practically as possible, and this would extend to the way we approached the VFX. We would always endeavour to place a real element at the heart of the shot. This could mean SFX constructing a fully working, fully electric Fiat 500 for the Rome car chase, or sending a full size locomotive over a cliff and into a quarry to provide us with a central element for the train sequence. Having this sort of photographic truth at the centre of each shot gave us a tangible reality to base our work in.
How did you organise the work your VFX Producer?
Robin Saxen, the show’s VFX Producer, and I would first break down the film, with input from the other HODs in terms of shooting methodology. Once we had the breakdown, we could then break the work up into manageable chunks for bidding with our vendors.
How did you choose the various vendors?
We would cast our vendors based on their respective specialisms. This film had a very broad spread of different types of work, so finding the challenges that would play to their strengths was key.
Can you tell us how you split the work amongst the vendors?
ILM were our primary vendor, with work being completed between the London and Sydney studios across all the major sequences. Simone Coco headed up the Sydney studio, working on the train crash sequence, whilst the London studio, headed up by Jeff Sutherland, worked on pretty much everything else, from car chases to mask gags.
beloFX worked on the opening submarine sequence, whilst Rodeo FX and Alchemy 24 provided the pre- crash train interior work. BlueBolt produced work in the Airport and One Of Us helped out on the desert fight sequence. We also had Atomic Arts and Untold Studios on the show providing various shots across the movie. Blind LTD and Territory Studio provided our motion GFX, whilst Lola provided some complex face replacements in the airport.
Like every Mission Impossible movies, we are traveling a lot. Can you elaborates about it and your set extension work?
We travelled around the globe, with locations in Rome, Venice, Abu Dhabi and Norway. Whilst we would strive to shoot out as much of the action on location as possible, we always knew that we would need to take some set pieces back with us to our stages and backlot in Longcross. Whilst on location, we would capture as much information as possible. We captured each environment using a multitude of approaches.
On the day of the shoot, we would shoot tiles on A Camera for any shots that we felt might need a pickup, and would also sweep up after the main unit and shoot bracketed bubbles along the route of the action. For the Rome car chase, we put together an array vehicle, with an eight camera setup, giving us a 180 degree dome for use in any interior car work we might pick up on the backlot later in the shoot. This array footage also worked pretty well as a photogrammetry source if needed to patch up any holes in our traditional captures.
We also had a team from our friends at Clear Angle, who would sweep up locations once main unit had moved on, with Lidar and additional texture photography.
Which location was the most complicate to enhance?
It’s a tie between the Spanish Steps in Rome and the Train sequence.
The Spanish Steps are incredibly old and fragile, and its forbidden to even sit down on them! Therefore, sending heavy vehicles down them was not really on the agenda. Even shooting plates would be difficult; We considered drones, but that type of camera movement was just not in the language of our movie, so we had to consider other options. In the end, we decided to build our own Steps on the backlot in Longcross. These were divided into two parts, upper and lower sections, with reinforced steel sub structures. This gave us a practical base, to which we would add our digital extensions, constructed from extensive scans and detailed photography of the actual Steps.
The Train sequence was challenging in the sense that we shot it across multiple locations and times of year. We established the look of the sequence in Norway, where we shot beauty shots of the train running through the mountains. We also shot some of the rooftop fight between Ethan and Gabriel here. During this shoot, we were running an array of cameras to capture moving backgrounds to use in any interior stage based carriage work we might pick up later down the line. After a break in production, we came back to the sequence, but due to logistical constraints, couldn’t get back to Norway to shoot. Instead, we scouted UK locations and found a stretch of track in North Yorkshire that had a good topological match to the Norway terrain. Due to the differences in season, we had to augment the locations pretty heavily using data we had captured in Norway.
The Rome car chase is really cool. How did you enhance it with cars and crowd?
When we got to Rome, the city was still in semi-lockdown, so the streets were relatively empty. We would add parked cars and traffic, as well as digital crowds to get the sense of bustle back in to the city. We carefully orchestrated our cg vehicles, placing them in strategic areas to make sense of the swerving Fiat and Hummer. We paid special attention to the physics of each vehicle, rocking cars on their shocks under heavy braking and even animating brake lights. We wanted it to feel like Ethan was threading the eye of a needle with the Fiat and our CG cars.
How did you work with the SFX and stunt teams?
All of the departments on this show worked incredibly closely to achieve these shots. Since Tom insists on doing his own stunts, we had the opportunity to build some viscerally real sequences around him, taking the audience along for the ride. Wade Eastwood, our stunts supervisor, would design action with Tom and Chris McQuarrie, and between myself and Neil Corbould, our SFX Supervisor, we would figure out methodologies for achieving the action. This would come down to Neil building a series of incredible rigs, giving Tom dynamic environments to work with. We would then cleanup and extend where necessary. We had a goal of keeping our work grounded in the real, so to this end would go to great lengths to shoot practical elements. Neil and his team were key in this respect, giving us incredible elements to work with.
Can you tell us more about the impressive cliff jump sequence?
This was the very first thing we shot on the movie. The construction of the ramp alone was an incredible feat of engineering. Every part of that ramp had to be taken up the mountain by helicopter. Neil and his team designed the ramp using simulations of Tom’s trajectory, figuring out exit speeds, ramp inclines etc. We then constructed a twin ramp in the UK for Tom to rehearse on, with wires for safety. For the actual jump, we made sure we had extensive coverage of the event, shooting close coverage on a drone, a heli shooting long lens, and several ground based positions including a technocrane that had to be lifted by helicopter up the mountain. In the end, the shots were almost entirely practical barring the ramp paint out.
The train sequence is really intense. Can you tell us more about its shooting and especially the train crash?
We shot this sequence across a number of years and locations. We started shooting in Norway on a section of track secured by production. Art department dressed some rolling stock to look like the Orient Express and Neil and his team built a locomotive to put out front. This procession formed the basis for our beauty shots. For shots of the cast on the exterior of the train, we removed some of the art dept cars and switched in a low loader equipped with a camera crane and the vfx camera array, capturing backgrounds for our interior stage based carriage work. At the same time, we had a pursuit vehicle chasing the train in contact with us over radio, picking up HDRIs along the way. We also captured extensive scans and panos for use later. We ended up shooting train pickups on a stretch of track in North Yorkshire, since the terrain was a reasonable match for our established Norway look. Due to time of year issues, we ended up doing fairly extensive digital augmentation to pull the looks in line.
For the loco going over the cliff edge, in line with our desire to put reality in the heart of each image, we sent a full size SFX loco over some track into a quarry, shot from various fixed positions and a heli. We also had fixed positions on the loco itself, with cameras in crash boxes fitted with explosive bolts, which would be jettisoned from the loco prior to impact. The surrounding environment was a CG build, based on an amalgamation of features captured in Norway during our time there.
The slow motion train wreck, with Ethan and Grace navigating a series of jack knifing carriages, was shot on the backlot at LongCross on a variety of gimbals. The main gimbal held various full size carriages, capable of more than 30 degrees of tilt. This allowed us to put our actors in positions of real jeopardy, allowing them to navigate tumbling furniture and obstacles for real.
The final piano bar set was mounted on a gimbal capable of going level to vertical. For the dangling piano, we shot various elements of an actual piano suspended on a release rig, and shot elements of furniture crashing through the back of the carriage for real. The idea was to capture as much of the chaos of these real events as possible and lean in to the imperfections that would occur naturally in the process. The same was true of Tom and Haley’s performances. Most of the time, the tension you see in their faces is the real deal!
What was the main challenges with the train sequence?
It was an incredibly complex sequence with lots of moving parts and required absolute cohesion between all departments. We worked incredibly closely to have it all work out, and it required a huge amount of planning.
Can you explain the effects for the different face masks?
Due to the language we’d established for the film in terms of how we would move the camera, and due to the fluid way we worked, we didn’t want to be bogged down with motion control and the restrictions that process imposes. Thankfully our A Camera operator, Chunky Richmond, happens to be a human motion control rig, so could quite faithfully replicate moves on his steadicam. We would shoot a series of passes; The A side actor, then the B side actor followed by a pass of the A side actor pulling off a prosthetic rubber mask. We would stabilise and align the takes in the comp, occasionally re-projecting to correct variations in the lineups, then use the leading edge of the prosthetic mask as our wipe between the elements.
The movie is full of motion graphics. Can you elaborates about their design and animations?
We had two teams running out our motion graphics for this movie. Territory handled the Entity and Venice satellite imagery, whilst Blind handled almost everything else. For Eddie Hamilton (the editor) and McQuarrie, graphics do a lot of heavy lifting in terms of story telling, so it was critical that the graphics could convey a lot of information in a very short timespan. Clean and uncluttered design therefore is crucial. For the Entity, we looked at a lot of visualisations of Big Data, and ended up with this concept of an all seeing eye constructed of data, that appears to be looking around the room. This was important in lending the villain of the piece a sense of character and presence in the movie.
Which sequence or shot was the most challenging?
The opening submarine sequence provided challenges in the sense that it was the one sequence we couldn’t capture anything in camera to hinge the work off of, for obvious reasons! There’s also very little reference, in terms of what we could realistically ask an audience to believe in outside of certain hallmark movies such as The Hunt For Red October and Crimson Tide, so we made the decision to lean into these movies as a visual touchstone. In the spirit of approaching things in an authentic way, we explored shooting in the same style as these films, ie: flooding a sound stage with mineral smoke and pulling big miniatures through the volume. Due to logistical constraints and knowing what our miniature would need to do, we moved to a digital approach, but tried to hold on to the same aesthetic. This extended to the ways we approached the execution, with Joel Green and his team at Belo coming on board to execute the work. We filled our digital world with a 3D density field, and pulled our sub through this, giving us a nice falloff and diffusion, mimicking the look of these early movies. The physics of the sub explosion also took cues from Crimson Tide. It was a challenging sequence to get right creatively, but we’re all very happy with the results!
Is there something specific that gives you some really short nights?
We’ve been working at pace on this franchise for so long now that the panic has become normalised! It’s not called Mission Impossible for no reason.
What is your favourite shot or sequence?
So many amazing moments, its impossible to choose.
What is your best memory on this show?
Again, so many great memories. Sunsets in the desert, dawn commutes along the Grand Canal in Venice after night shoots, hurtling through the Norwegian mountains on a low loader carriage, but the biggest thrill was finally seeing the completed film on the opening night in Rome and seeing everyone’s hard work on the big screen.
How long have you worked on this show?
I started in 2019, and we finished our last shot exactly 4 years later.
What’s the VFX shots count?
In total, around 3700.
What is your next project?
We’re well under way with the next installment!
A big thanks for your time.
WANT TO KNOW MORE?
ILM: Dedicated page about Mission: Impossible – Dead Reckoning Part One on ILM website.
Atomic Arts: Dedicated page about Mission: Impossible – Dead Reckoning Part One on Atomic Arts website.
beloFX: Dedicated page about Mission: Impossible – Dead Reckoning Part One on beloFX website.
BlueBolt: Dedicated page about Mission: Impossible – Dead Reckoning Part One on BlueBolt website.
One of Us: Dedicated page about Mission: Impossible – Dead Reckoning Part One on One of Us website.
Rodeo FX: Dedicated page about Mission: Impossible – Dead Reckoning Part One on Rodeo FX website.
Territory Studio: Dedicated page about Mission: Impossible – Dead Reckoning Part One on Territory Studio website.
Untold Studios: Dedicated page about Mission: Impossible – Dead Reckoning Part One on Untold Studios website.
© Vincent Frei – The Art of VFX – 2024
A couple of months back, Erik Henry walked us through the visual effects craftsmanship of the second season of Invasion. Today, he delves into his experiences working on Percy Jackson and the Olympians, discussing the many challenges he encountered.
How did you get involved on this series?
I got involved with Percy Jackson because I had been working for more than a decade with the show’s creators. We had done Black Sails together and were in the middle of shooting The Old Man for FX when they asked me to join the team for Percy.
How was the collaboration with the showrunner and the directors?
My collaboration with the showrunners and directors was super smooth. We all bought into the idea that to be successful the show needed not to be too scary for the little kids while also being real enough that adults would be intrigued. The VFX team adopted that principle when making the monsters throughout the season. The Minotaur has just a hint of Teddy Bear to it.
How did you organize the work with your VFX Producer?
I’m lucky enough to have a stellar VFX Producer in Matt Robken. He and Amanda Newby Fitzgerald would take care of literally everything except the collaboration with the show’s creatives. This gives me the time to stay close to the art department, DP and all the creatives without having to worry about tax credits and the like.
How did you choose the various vendors?
One of the things the three of us did was vetting various companies for their creature capabilities, and for who was still at the particular house that worked on the shots we liked. Then the next step was to play to strengths at each house. MPC in London got the Chimera in episode 104 while ILM did the Minotaur, Cerberus and Medusa. Both companies worked on the winged Furies. When it came to the Nereid, an underwater creature, we had seen a demo of an underwater sequence that showed great promise from the VFX house MARZ in Toronto. When we found that shooting an actress underwater was going to be impossible we hired MARZ to create the Nereid completely in 3-D and used an actress to voice her. Environment work was given mostly to our longtime collaborators, Raynault VFX in Montreal and Storm Studios in Norway. Their eye for detail is essential for our grounded and realistic ideals.
Can you take us behind the scenes to create the visual effects for the mythological creatures, such as Medusa and the Minotaur?
A little behind the scenes of making Medusa – The show’s creators had a very important directive for Medusa, no silly snakes looking at camera! In fact they believed, rightly, that snakes generally move very slowly and therefore the snake ‘hair’ should exhibit very small slithering movements at most. We looked at all the scenes and felt we could possibly come up with a cost effective 2-D approach to the majority of shots and let ILM tackle a small subset of very hero moments where the snakes needed to perform. Storm took on the 2-D approach. Using warps and the highlight augmentation, we were super happy with what looks like undulating snakes under a hat the actress wore. For ILM’s work the actress wore a trackable tight fitting cap that kept her hair pulled up and out of the way. The snakes were modeled with a bit of variation in girth with the larger ones animated to move more slowly than the smaller finger thick snakes. variations in color were added to help differentiate the otherwise black mass. One shot had two snakes hiss and snap at each other, for that the animation team watched hours of real snakes, an iconic scene from « Indiana Jones », and even pulled from the expertise of those who worked on « Snakes on a Plane! » When we felt we were close the studio head asked if we could add some eye fx to subtly add that final x-factor. This was accomplished in the color suite where our color timer tracked the eyes in each shot and subtly added a slight colored glow (very slight) that pushed Medusa over that final edge from woman to creature!
What sources of inspiration did you use to bring such iconic mythological creatures to life on screen?
With the Minotaur we saw an opportunity to do something different in that we wanted it to look more bull-like than man. We also like the white and grey bulls we saw in our research but wanted to give it some back story. We added earrings and braided hair in order to give it a sense of individuality that we had never seen in other iterations throughout its history. We had the extra challenge of needing to have the actor climb on and interact with the monster. Our SPFX team helped by building a motion based fullscale blue Minotaur gimbal complete with fur, and with the help of the costume department, a custom pair of tighty-whitey’s! The fur helped our actor hold on in the torrential rain as we had the motion based buck and whirl and the underpants provided a key moment of comedy mixed into the danger. The scene was the big climax for the pilot episode so it was a make it or break it moment and we knew this monster had to be special. Early on we decided that bulls are most dangerous when they run at you on all fours. We decided our bull-man would stay on all fours as well, until a key moment when he stands to his full height and scares the hell out of Percy. We also loved the idea of having hooves break open into fingers as the monster rose up, super creepy! At the same time we struggled as we approached final on several close ups because the showrunners thought the face was getting too Zombie-like and therefore too scary for a segment of our audience. ILM smoothed some shapes and adjusted how we saw the teeth (yes we decided this bull-man needed to have a full set of teeth, bulls do not), and that seemed to work! The climax is that Percy snaps one of the Minotaur’s horns off and plants it into the beast’s head. As the Minotaur dissolves into dust Percy falls to the ground. For this we suspended the actor from the volume floor and SPFX used a descender rig to drop him for real. In post we found the animation of the Minotaur and its dust sim, requiring us to transition to a digital double of Percy to have him interact more appropriately with that simulation and displace it correctly and seamlessly.
What were the main challenges in designing visual effects for mythological creatures, and how did you overcome them?
When designing the mythological creature for the show we always had Rick’s books for inspiration and Rick for guidance. He was a constant for us, always there to let us know what he intended. That was so valuable. It helped us understand how Chiron would dress when in his Centaur form and described the mashup that is the Chimera. The three-headed Cerberus was a special challenge because of its size. Thirty foot tall dogs have been done before and we knew we had to think of ways to make this better than in the past. To overcome the challenge we started by saying, what this needs is a little humor so the audience is with you from the start. The writers came up with the idea that Annabeth would use an old trick that works on pretty much any dog, scratching behind the ears to immobilize them in bliss. We set to work on a pre viz and when I brought it to the guys they laughed and said « are we really going to do this? » And next said they wouldn’t change a frame! so we shot it pretty much exactly like the pre viz with other silly parts including Grover being eaten by the dog and then climbing out of its stomach, and of course using a red squeaky ball to distract and send it away! The sequence is shrouded in fog not in any attempt to hide anything, more to reveal the dog at key moments and have you guessing where the threat is hidden throughout the scene.
How did you manage the balance between using digital visual effects and practical effects?
I’m always interested in doing things practically when possible, especially when it comes to fire. In episode 104 the Chimera tries killing Percy by burning with its fire-breathing capabilities. I asked Joel Wist, our SPFX supervisor if he could help and he was too happy to jump in with flame throwers! The idea was to paint the full scale set black after we finished shooting the first unit work. Next we added fire retardant and placed the cameras behind transparent barriers then Joel let it rip. We did end up combining a little CG fire in post but the overall result is a composite that feels so much more real than if we had settled for just a digital solution.
Which sequence or shot was the most challenging?
The challenges associated with extending sets. We had many set extensions for Camp Half-Blood that did require a deep think in post. We had one panning shot that really took in ALL of the world the kids train in each day. On the day we had a strawberry field and a fighting arena and the deal with the art department was VFX would add the rest. « The rest » became an interesting challenge after I showed the guys the first pass for the environment. It became clear it was not nearly enough. So we went into fan sites and again relied on Rick to help us. What we found was many ideas pulled from many books where fans and Rick had added tidbits that we thought could be visually arresting. I asked the team at Raynault to add a volcano, a waterfall, a lake and many things that, while not played out in this season, would become important in other subsequent seasons. We also had a big challenge in the Underworld where the sets were on a large scale but the finished comps would need to be on a monumental scale. The DP was super helpful in that he created a soft ‘sky’ to the studio ceiling that we were able to cover with our idea that the underworld was a huge god-created cavern with inverted mountains and an interesting alluvial flow of black on a field of white. It was a really good collaboration that paid off in that I’ve never seen anything like it. I hope the audience is also struck by the uniqueness.
Is there something specific that gives you some really short nights?
One of the most challenging scenes was the Dodds transformation into the Alecto fury. The challenge was that the actress, who was wearing a long leather coat, would have to turn that coat into a set of wings while we watched. In the end we had to work with two different companies to get to a place we were happy with the result. The simulation of the leather coat to wings was helped tremendously by adding live action extra crosses pulled from multiple takes and ROTO’d / color adjusted to fit into the FG. This worked because the scene was already filled with people and the addition of these crosses actually helped the sort of surrealistic moment that Percy finds himself in where no one around him sees what we see with him. The end effect was to give almost a sight of hand to the scene where you aren’t quite sure what’s changing from moment to moment, something the showrunners had hoped for from the start!
What is your favorite shot or sequence?
The Minotaur sequence was the sequence that gave me some sleepless night because, as is sometimes the case with the first sequence you work on, you spend way too much time on it! We worked on the Minotaur for the better part of a year to get it just right and it was right down to the last hours to get the final shots into the episode.
What is your best memory on this show?
I remember one night we were working late looking at final color on episode 107 when Dan Shotz, one of the executive producers got a Facetime call from our main actor Percy (Walker Scobell) and the actor who plays Grover (Aryan Simhadri). They were hanging out together and thought it would be fun to call Dan. He pointed the phone at the color suite screen and we played through the sequence where Percy and Grover come over a rise and see the huge Tartarus hole in the underworld. The actors had never seen anything other than the stage floor and some blue screens on the day so it was super gratifying when they both gasped involuntarily and said how incredible it looked. I knew we were in a really good place then.
How long have you worked on this show?
I worked on Percy for 3 years.
What’s the VFX shots count?
1800 shots.
A big thanks for your time.
WANT TO KNOW MORE?
ILM: Dedicated page about Percy Jackson and the Olympians on ILM website.
Disney+: You can watch now all the episodes of Percy Jackson and the Olympians on Disney+.
© Vincent Frei – The Art of VFX – 2024
In 2022, Kunal Ghosh Dastider had revealed to us the visual effects work made by DNEG on Sonic The Hedgehog 2. Then, he worked on the Netflix film Lou.
Back in 2019, Spencer Cook gave us a behind-the-scenes look at DNEG’s animation work for Godzilla: King of the Monsters. He then worked on Venom: Let There Be Carnage.
How did you get involved in this film?
Kunal – I had the opportunity to apply some new approaches for creature workflow for this particular show.
Spencer – I was taking time off after supervising the creature animation for Venom: Let There Be Carnage. During my time off I kept in touch with our DNEG studio manager, she would tell me about upcoming projects and as soon as she started talking about No One Will Save You, I was interested.
How was the collaboration with Director Brian Duffield?
Kunal – Brian had many cool ideas and wanted to try out different ways to creatively approach VFX. There was an emphasis on the character-centric storytelling and keeping the approach bizarre and weird. He would love to act out the motion of the aliens and send us the videos. We then discussed it in our calls to help guide through our process.
Spencer – It was great. Brian was very open to collaborating and experimenting. Brian’s script gave a good idea of how weird and unsettling he wanted the aliens to be. It was important for the story to find ways to hint at a culture for the aliens, a style of movement for each that indicates their place in the alien society.
In order to find the body language for each alien I suggested we set up a mocap session where Brian and I could try out ideas on the fly and quickly iterate. We could’ve started with animation tests but I felt we would find the body language of each alien quicker if we were directing a performer. I made a list of actions and performances based on Brian’s script. We hired professional mocap/creature performer Richard Dorton (who I’ve worked with before), and spent two days on a mocap stage in LA playing around with different ways to convey the actions in the script. We never intended to use the mocap data in the final shots, Brian and I both felt the aliens should move in inhuman ways, but it gave us a good foundation and clear direction for the animation tests back at DNEG.
What was his approach and expectations about the visual effects?
Kunal – Everything needed to have a bizarre touch of science fiction but be somewhat grounded in reality. Feeling realistic yet set in a dark scary environment. Apart from the aliens and the parasite, the flying saucer and surrounding “alien” clouds needed to also feel like a character. Even the “alien” light beams needed to have its own character and personality. We digested all these challenges with Production VFX Supervisor Stephane Paris, then we focused on how we translate that best into our world with our team.
Spencer – Brian wanted the aliens to be weird and inscrutable to humans, to move in unsettling ways.
An aspect that we focused on a lot in animation was the multi-layered language of the aliens. In addition to the sounds they made, their physicality is filled with nuance and complexity beyond human understanding. To convey these ideas visually, we started with mouth movement, then added ripples in the skin around the neck and temples, we also added small head twitches/clicks and hand gestures to suggest the aliens use all parts of their body to communicate with each other. As Brian put it, these extra gestures are the equivalent of humans using italics or bold type in an email. A lot was made in the press that there was no dialog in the movie, while that’s true for the humans, the aliens speak quite a bit throughout the movie, we just can’t understand them.
Can you elaborate about the design and creation of the aliens?
Kunal – We received concepts from production design and over time it evolved at DNEG into a lot of fine details and nuances with the aliens. Since Brian Duffield avoided having dialog it was even more crucial to have very sophisticated facial expressions to translate the emotions to the viewers, even if we kept facial animation at a minimum.
The eyes were especially a challenge because they were hidden behind a membrane that acted as a kind of shield and we would want to reveal or see the eyes only in certain moments for story telling reasons. Our Compositing Supervisor Mohsin Kazi came up with a slick trick of rebuilding the shader in compositing and adding some 2.5 magic to it, so we could dial the look in without needing to re-render in lighting every time we had creative notes per shot.
Spencer – Brian wanted us to visually depict a culture through body language that humans don’t understand. We developed unique physical characteristics for each alien, We wanted to give them distinct personalities and a sense of who they are in the alien culture;
-Main Gray – Leader of the group. Calm, intellectual, elegant, aloof like a cat
-Little Gray – Worker class. Short, chimp-like. Hyper aggressive like an attack dog.
-Daddy Long Legs – The high priest of the group, spiritual leader. Moves with weird ritualistic gestures.
Facial animation was kept to a minimum for the Main Gray. Brian wanted him to be mysterious and creepy, his black eyes giving little indication of what he might do next. The Little Gray was the most expressive. He would easily fly into a rage or whine when injured.
While each alien presented its own creative and technical challenges none were more challenging than the Daddy Long Legs. The unusual alien anatomy with long legs and extra joints, some bending backwards, required our team to customize the rig well beyond our usual template.
How was their presence on set? How did you handle the challenges of weight?
Spencer – On set we had a Main Gray head, mostly used for lighting reference and a full body of the dead Main Gray. My concern on set and location was mostly that the camera moves weren’t too fast for a big creature like the Daddy Long Legs. Another consideration was framing for the aliens. The stunt guy stood in for the Main Gray and Little Gray so that was fairly straightforward. For composing Daddy Long Legs shots, since he’s much taller than the others, we used a “monster stick”, which is just a long pole with a blue ball at the top where his head would be. I usually had the job of holding the monster stick so the crew could frame the shots. I stood in a field in the night, during the Louisiana summer, moving the monster stick around for Kaitlyn to react to. Not a glamorous job but fun nonetheless!
How did you work with the SFX and stunt teams for the aliens’ interactions?
Spencer – We had a stunt person in a green suit stand in for the aliens. Production VFX Supervisor Stephane Paris and I helped direct him based on things we learned in the mocap sessions and animation tests. This also gave lead actress Kaitlyn Dever something to react to, especially in fight scenes.
Can you tell us more about their rigging and animations?
Kunal – The Creature team was led by Creature Supervisor Anaïs Ringenbach and Build Supervisor Andreas Loose. They worked hard to come up with solutions to the challenges of the bizarre motion and behavior that the aliens had to deliver. Since none of the aliens should move too humanoid-like, especially Daddy Long Legs who would also move in a reversed quadruped manner, we had some challenges in terms of rigging and building the creature. With some head banging and combined brain power, the team brought a solution to the table that was achieving all our goals and giving our animation team enough control to bring the creatures to life.
Spencer – The Main Gray and the Little Gray were fairly straightforward in terms of rigging because they were basically bipedal designs. However, one challenge with the Main Gray was when he walks on his toes. He mostly walks flat footed but in one sequence he stands up on his toes with each foot crawling like a spider.
The Daddy Long Legs was the most difficult to rig. The multiple limbs, some joints facing backwards, took some time to work out. It’s generally a back and forth process between animation and rigging. We initially suggested rigging what we thought we needed to animate a creature. Rigging did a first pass and gave it back to us. We did animation tests and learned new things then went back to rigging with updated notes, and so on. As our animation tests progressed and we dialed in how Daddy Long Legs should move and behave, we gave rigging more specific notes on exactly what we needed and what we didn’t need.
DNEG Lead Animator Sebastian Weber helped design each alien rig and animated many of the key shots. He was especially instrumental in making the complex rigs work for the Main Gray toe walking and the unusual anatomy of the Daddy long Legs.
What kind of references and influences did you receive from the director for the aliens animations?
Spencer – Brian’s script was the starting point, then through conversations and reviews we would further clarify his ideas. It was my job to find visual references. In creature animation we always try to find real animal/people references to base the fantasy creatures on. The creatures will be more realistic if we can mix in real animal characteristics with the fantasy elements. For example, the Daddy Long Legs is generally based on video reference of spiders, crabs, praying mantis and giraffes, but also I collected references of various religious ceremonies looking for gestures that we could translate to the alien ceremony.
Tricky question, which alien is your favorite?
Kunal – For sure a trick question. I have to say no favoritism, I have the same love for all of them! It doesn’t matter tall, short, wide, thin or freaky, I love them all!
Spencer – I don’t have a favorite. All three were fun and interesting to create. Each alien was so different from the others that it was an enjoyable challenge to define the rules for how to clearly portray their personalities and intentions in each scene.
Can you elaborate about the creation of the flying saucers?
Kunal – As mentioned before, everything needed to have its own character, the flying saucer was designed to have multiple discs on top of each other with a Vantablack material finish. Due to the night sky making it hard to see, we added some stars and clouds so we could separate the discs from the sky and justified adding some lights that would come out when those discs were separating from the flying saucer.
Spencer – Brian wanted the ships to be mysterious. They were designed to be simple disc shapes with not a lot of surface detail. At one point the Daddy Long Legs communicates with the ship through a series of gestures, like sign language or semaphore. The ship responds with alternating discs moving in a cascading pattern. This made the ship feel like a character instead of just a vehicle.
What is your favorite shot or sequence? Which was the most challenging?
Kunal – One of my favorite sequences is when Main Gray gets stabbed in the head, and falls on Brynn (Kaytlin Dever) with all the little nuances in his face followed by feeling the loss of life in his body.
The most challenging part was the “alien” light beam. It was an emotional roller coaster, from crying to laughing to getting even more gray hair. The end result was pure VFX magic that needed lots of attention and love from our team to be able to get the characteristics and story of how the light beam needed to be.
Spencer – The Daddy Long Legs sequences were by far the most challenging but also my favorite because of the weirdness of that alien. The creepy spider-like quality and the ceremonial gestures of their alien religion were fun to design.
What is your best memory on this show?
Kunal – Apart from the team delivering the show with a big kaboom, it was also amazing to see the exceptional work delivered by the talent at DNEG on a daily basis. What made it unforgettable for me was the laughter and fun that we had while working on the show.
Spencer – Collaborating with Brian and his team was a great experience throughout. They relied on us to add ideas and help push Brian’s vision further, and of course, the DNEG team was fantastic, especially the animation team! I had great creature animators who would push beyond the notes and really make each shot a work of art.
How long have you worked on this show and what’s the VFX shots count?
Kunal – We worked on 271 shots, with 624 team members for just a bit over a year. All amazingly managed by DNEG Producer Patricia Martinez Arastey, and DNEG Show Production Manager Nelson Dsouza across the global teams at DNEG.
Spencer – I was on the project for a little over a year.
What is your next project?
Kunal – Stay tuned, the magic sauce is happening!
Spencer – Godzilla x Kong: The New Empire. We did some really cool sequences at DNEG.
A big thanks for your time.
WANT TO KNOW MORE?
DNEG: Dedicated page about No One Will Save You on DNEG website.
Disney+: You can watch No One Will Save You on Disney+ now.
Hulu: You can watch No One Will Save You on Hulu now.
© Vincent Frei – The Art of VFX – 2024
Behind the scenes of The Winter King, the teams of the Uk based studio Realtime faced various technical challenges in creating the visual effects. Let’s discover how every detail was studied to bring to life the epic universe of this adaptation of Bernard Cornwell’s bestseller:
WANT TO KNOW MORE?
Realtime: Dedicated page about The Winter King on Realtime website.
© Vincent Frei – The Art of VFX – 2024
Immerse yourself in the gripping story of the most epic manhunt of all time. The trailer for this new Apple TV+ series, Manhunt, takes you to the heart of John Wilkes Booth’s intense hunt after Lincoln’s assassination!
The VFX are made by:
Zoic Studios (VFX Supervisor: Kenton Rannie)
Refuge VFX (VFX Supervisor: Tarn Fox)
Directors: Carl Franklin, John Dahl, Eva Sørhaug
Release Date: March 15, 2024 (Apple TV+)
© Vincent Frei – The Art of VFX – 2024
Explore the subtle yet magic of the visual effects work on The Marvels, orchestrated by the London teams at RISE:
WANT TO KNOW MORE?
RISE: Dedicated page about The Marvels on RISE website.
Tara DeMarco: Here’s my interview of Production VFX Supervisor Tara DeMarco.
Disney+: You can watch now The Marvels on Disney+.
© Vincent Frei – The Art of VFX – 2024
Discover the behind-the-scenes of the visual effects in Lil Nas X’s music video, J Christ, brought to life by the teams of Mathematic Studio:
// Lil Nas X – J Christ
WANT TO KNOW MORE?
Mathematic Studio: Dedicated page about Lil Nas X – J Christ on Mathematic Studio website.
© Vincent Frei – The Art of VFX – 2024
In 2023, Graham Page presented the work of Framestore on The Matrix Resurrections. Today, he returns to discuss his role as Overall VFX Supervisor on Wonka.
With over 25 years’ experience in visual effects, Robert Allman has made his mark working for studios such as Mill Film, Prime Focus and Framestore. His impressive portfolio includes films such as Gladiator, Alien: Covenant, Dolittle and The Little Mermaid.
Carlos Monzon began his career in visual effects in 2001 when he joined the Buzz Image team. His career led him to work with several studios, including Meteor Studios, ILM and MELS, before joining Framestore in 2013. He has worked on shows such as Avatar, Paddington, The Martian and Don’t Look Up.
What was your feeling about this new collaboration with Director Paul King?
Graham Page (GP) // Working with Paul was great; he has lots of ideas and a fantastic eye for detail. He’s always trying to bring out as much humour as possible into every shot so there’s lots of fun involved, but he’ll always push very hard to get the best result possible as well.
Robert Allman (RA) // Paul really believes CG in visual effects is a pivotal part of the filmmaking process and as such he should be every VFX artist’s dream to work with! He has great respect for us and what we do and that is why the VFX in his films are always so great.
Carlos Monzon (CM) // I had the pleasure of working with Paul on Paddington 1 & 2. Collaborating with him Wonka was great, he understands the value of the work we do and really helped us achieve the best results for his movie.
What was his approach and expectations about the visual effects?
GP // Having made the Paddington films, Paul has a fantastic understanding of VFX. He knows what is achievable and what the bigger challenges are. He was very keen to create realistic environments rather than anything too stylized or fantastical. Paul has a big fondness for animation so he loved working with our animation supervisors, Dale Newton and Meena Ibrahim, on creating the various creature performances in the films. He appreciated our approach which was always to try to ground the creatures in real world references.
RA // I think he always thought ahead that his shots could be perfected in CG, from everything from animated performance to smoothing out live action camera moves if the cranes etc. hadn’t quite given him what he wanted. The leap of faith he took in going for a fully CG Oompa Loompa, given the level of realism needed, was incredibly brave. Again, it really paid dividends and I hope we did that decision justice.
How did you organise the work with your VFX Producer?
RA // The decision was taken to put the Oompa Loompa in our London studio owing to the concentration of animators here and also that the follow up Hugh sessions would be done in London. The other work was more or less overflow from what Montreal was doing.
CM // Early on it was decided that the Montreal studio would take care of a large portion of the set extensions and creatures. Focusing mainly on the design of the Town Square, Zoo, Scrubbit & Bleachers, Galeries Gourmet and Abigail, the Giraffe.
Where were the various sequences of the movie filmed?
GP // Much of the film was shot at Warner Brothers studios in Leavesden UK but we also went to a number of locations around the UK. We shot the interior of the Cathedral in St Paul’s in London, it was incredible to see it all beautifully lit with 100’s of stunt monks running through and Rowan Atkinson being chased by a full scale giraffe Puppet.
Oompa Loompa land was filmed in Wales at Church Door Cove in Wales, just as we were about to start to shoot, Timmy Mallett arrived to go for a morning swim and we had to wait for him to swim off, out of shot. The background plate for Wonka walking up the beach to collect Cocoa beans was the first footage shot on the film with an element of Timothée Chalamet shot later at Leavesden.
The exterior of the zoo was shot in St Albans, north of London.
We also shot in Bath and Oxford and Wonka’s arrival to the dock of the Grand City was filmed in Lyme Regis. Large crowds turned out to watch the filming and our reference chrome ball drew attention with someone shouting loudly. ‘He’s got a Lolly Pop! He’s got a Lolly Pop!’
Can you elaborate about the design and creation of the various environments?
GP // Much of the design was started in pre-production by the art department who designed the sets and how they would be extended later, this design process carried into post with us refining and adding to these designs, both in terms of fleshing out the required details but also working the roof top designs to make sure they felt natural and also worked for shots where we are above them.
RA // I would echo what Graham has said. There was a lot of overhauling the design of the city in the wide shots, Paul really wanted a sense of reality to it, as if you felt you could take a holiday there. Planning and referencing real world scans of actual old towns was key.
What was the real size of the sets?
GP // The sets, brilliantly created by production designer Nathan Crowley were pretty huge. The Galleries set was connected to the town square allowing for shots crossing the threshold between the interior and exterior world. However, they generally stopped short at about 30 to 40 foot high and we needed to extend the tops of them. Painted backdrops were used in the townsquare down alleyways to provide naturalistic reflections for vehicles and reduce blue spill, these were then rotoscoped to allow them to be replaced with deeper 3d extensions.
Wonka’s chocolate shop used pretty much all of one of the biggest stages at Leavesden and involved a huge revolving tree and travelling chocolate river and candy boat. We enhanced this set with extra details such as the chocolate bunnies and owl and growing flowers. We also created the world above the canopy of the tree based on a theatrical backstage area with ropes and pulleys to control the floating clouds and rainbows.
What kind of references and influences did you receive for the environments?
GP // The film is set in a story book world with many influences taken from all over Europe. Before filming ended up in Leavesden the production had considered finding a real town square to film in such as the Grand-Place in Brussels but instead details were cherry picked from the best of these real world locations. The Galleries Gourmet interior is largely based on Galleria Vittorio Emanuele II in Milan. The powder gate tower in Prague features, along with various astronomical clocks and statues. The zoo draws influences from zoos in Berlin, Antwerp, New York and London amongst others.
RA // I would add Oxford to what Graham says about the Grand Old City. There was much talk about different districts and the university district was very much based on that.
Which location was the most complicated to create?
GP // The city and rooftop environment seen when Wonka and Noodle dance on the roof was hard to create as it needed to include lots of detail both near and far. With the Galleries roof in particular being difficult to design. The scene needed to look beautiful but also stay grounded in reality so we needed to balance the realities of street design while maintaining compositions which felt grand and effectively told the story of their travel from the zoo to the town square.
Can you explain in detail about the creation and animations of the animals and especially the giraffe?
CM // Asset Creation: Based on references provided by Framestore’s Capture Lab our asset team created 3D replicas of the Flamingos and Giraffe. Our models were modelled and rigged based on real life anatomy to define the bones and joints placements to have the animals correct range of motion.
During the asset creation the animation team under the Supervision of Meena Ibrahim started running tests to explore the physicality and different actions the animals could achieve. We made sure we could match exactly the interesting structure and behaviour of a real giraffe, in addition to the subtle facial performance to the finest detail tested the Giraffe and Flamingos.
The team also collected a large amount of video references that allowed us to match the behaviour of the real giraffe. The same references provided a target for the CFX team to match and build their rigs. The skin, muscle and fat simulations gave the giraffe a realistic look and feel. Once the tests were lit and comp onto a plate, it was almost impossible to tell the difference between the reference and our CG animated version .
Animation: The main challenge the animation team faced was to make the giraffe move in a believable manner, while still achieving the performance Paul wanted for the movie. In order to do this we had to gather even more references from all over the internet, as well as getting some acting references from Paul King during our client reviews. The team did keyframe animation for all the shots that involved a Giraffe, dog or a flamingo and based their animation on real references.
We see the giraffe in various close up shots. How did you handle this challenge?
CM // Based on early storyboards of the movie, we knew our asset would require a tremendous amount of detail to hold up really close to the camera. Framestore’s Capture Lab provided extensive footage and photographs of a real giraffe named Molly, from the London Zoo. This enabled the team to meticulously align our 3D assets to the references and create a 1 to 1 match. We had to create two variants of the giraffe, one that was used for most shots and another one that was built for all the closeups. This variant was a higher resolution model, used larger maps and had much more density for the giraffe’s groom.
How did you manage the interactions with the giraffe and the actors?
CM // On set puppets and eye line poles were used to interact with the actors. The live action footage required precise matchmove of the actors hands and fingers to create a flawless interaction between the CG and the actors. A handful of shots required us to fully replace the actors hands with CG, this allowed us to get a better interaction between the fingers and the fur.
Let’s talk about the impressive Oompa-Loompa. What was your approach for this character?
RA // After many different iterations at the concept stage, Paul arrived at the simplest explanation, that he was just a small Hugh Grant. From that moment, we were locked into achieving a high degree of detail in the facial performance but also resolving the classic look of the Oompa Loompa in that context. The costume designers made tiny costumes so we had a clear idea of the wardrobe of this stylishly turned out individual, which underlay the cool and capable side of his character.
Can you elaborate on his design and creation?
RA // We were half referencing the books, where the Oompa Loompas are tiny elf like creatures and half referencing the classic look from the Gene Wilder version of the film, green bobbed quiff, orange skin and white eyebrows. With this sort of comic appearance it became a huge challenge to keep it looking like Hugh. That could only be solved by scrutinising his every facial movement.
Can you tell us more about the animation challenges?
RA // The performance ended up being completely hand animated, mainly because we were dealing with exaggerated comic acting, something that is very difficult to get out of a more procedural, motion captured method. So much of a very expressive performance is subjective that it is really impossible to do it without an animator and sculptor resolving it subjectively.
How does his orange and green colour affect your work?
GP // The colours did offer their own challenge as Paul wanted them to be as vibrant as possible but particularly with the skin you’re limited with how much red can be pushed into the skin before the colour begins to look oversaturated and in some cases break out of the colour gamut. Similarly the green hair had to look like real hair not a plastic wig so a lot of effort was placed on getting variation within the colour and size of the strands.
RA // Yes, in many ways the worst possible scenario for a CG character who has to look real; bright orange and green! But that was a big part of the comedy, so the challenge was worth it. Lighting became very important as to make him look natural, you have to put more cool colour in the shadows than you would normally. The challenge was then that that would instantly break in a new sequence where the lighting was different. We were always starting again, the rebalancing was very tricky.
Can you explain in detail about the shooting and creation of the beautiful final sequence where we see the old castle turning into the chocolate factory?
GP // We found a great castle called Bodiam Castle in East Sussex which felt like a really good fit to create the factory within. The walls provided a great call back to the ruined stone walls we see in the 1970s film when we see Wonkas ‘world of Pure Imagination’ which we spent time analysing to try to recreate the look and feel.
The design was based around steam fairs and breweries for the mechanisms but also children’s playgrounds and swimming pools for elements of fun and vibrant colours. Jonathan Opgenhaffen designed the interior to have a logic and sense of flow in terms of how the chocolate powered the water wheel which in turned cogs and wheels around the factory.
Which sequence or shot was the most challenging?
GP // The first shot of Wonka climbing the mast was particularly difficult.
RA // Always the Oompa Loompa. The level of work needed to perfect the performance, the amount of animation, the lighting and look was as hard as it can be.
CM // During one of the musical numbers, there is one shot where the camera is flying over the plaza and hundreds of extras running towards Wonka. The shot is composed of over 11 plates that were conformed together to populate the town square with customers. This was an extremely complicated comp to achieve.
Is there something specific that gives you some really short nights?
RA // The chocolate factory was late in the schedule and when I worked out on the back of a beer coaster how many weeks we had left against the number of available machines to render everything, I immediately returned to the bar for another drink. And then quite a few more.
What is your favourite shot or sequence?
RA // The first close up look of the Oompa Loompa. We developed everything in that shot, so it must have taken a year and a half. Normally, the development shot is one of the worst, as you never revisit it and it doesn’t benefit from what you learn from the rest of your time doing the show. However, in this case, it was still as good as the rest.
CM // For me it was the sequence where Wonka and Noddle are dancing on the rooftop of the Galeries Gourmet. It was the first time we saw the city from this angle and the only thing that kept from the plate were the two actors. This meant replacing everything else with CG including the balloons.
What is your best memory on this show?
GP // First seeing the Oompa Loompa talking with all of the sculpt work and fine animation in place was an important moment as you could see all the detail in Hugh’s performance. The last thing we wanted was for people to be distracted by the CG and instead be able to enjoy the humour.
RA // The producers expressing their sadness that everyone would think we had shrunk Hugh Grant and filmed him. It was a great compliment and a hilarious thought!
CM // The reaction of Paul and the producers the first time they saw Wonka and Noodles dancing on the Galeries Gourmet rooftop. They really loved the result.
How long have you worked on this show?
GP // Around 2 years.
CM // I joined the show at the same time as Graham. Around 2 years
RA // Think it was a bit less than Graham.
What’s the VFX shots count?
GP // Across the whole film we had around 1163 vfx shots, Framestore delivered 922 shots.
A big thanks for your time.
// VES Nominee Spotlight – Framestore
// Trailers
WANT TO KNOW MORE?
Framestore: Dedicated page about Wonka on Framestore website.
© Vincent Frei – The Art of VFX – 2024
The Los Angeles based studio Filmograph skillfully designed an opening title sequence for The Beekpeeer that propels viewers into the heart of the film’s intensity through a clever fusion of macro cinematography, historical records, and ancient bee artwork:
WANT TO KNOW MORE?
Filmograph: Dedicated page about The Beekeeper on Filmograph website.
© Vincent Frei – The Art of VFX – 2024
