SHERLOCK HOLMES – A GAME OF SHADOWS: Sirio Quintavalle – VFX Supervisor – Framestore

Sirio Quintavalle started as a trainee at Framestore almost 15 years ago. He has worked on films like LOST IN SPACE, THE GOLDEN COMPASS, THE DARK KNIGHT or CLASH OF THE TITANS. In the following interview, he talks about his work on SHERLOCK HOLMES: A GAME OF SHADOWS.

What is your background?
I studied Physics then entered the industry as a trainee at Framestore working in the longform television department on shows such as WALKING WITH DINOSAURS.

How did Framestore got involved on this show?
I’m not sure on the exact decision making process but our work on the first SHERLOCK HOLMES, the production being based in London and Framestore’s great reputation were probably all factors.

How was the collaboration with director Guy Ritchie and Production VFX Supervisor Chas Jarret?
Guy and Chas knew each other well from the first movie and they had a great understanding. Chas was great to work with and was always pushing the boundaries, whether with new camera rigs, robotic arms or even flying his own model helicopters to get the perfect angles for the effects shots. Guy is a larger than life personality both onset and during post review sessions which he always brought to life with his colourful turns of phrase.

What have you done on this movie?
Set extensions and stunt cleanups for the fight with the assassin; environments, green screens, steam train and lots of explosions for the train battle; environments for the alpine sequence, a lot of cg breath and a big, ultra slow motion fall of a waterfall for the finale.

How did you create the huge environment for the train sequence?
A mix of live action panocam footage take from a disused railway, helicopter plates and cg modelling using Maya and Onyx tree and our in house fur system for grass and additional foliage. The panocam footage was shot using overlapping cameras to allow us to stitch together a moving environment which covered all the angles for the green screen work inside the train and for some of the simpler exterior shots. Where possible we used real footage and even with some of the more complicated camera moves which didn’t allow a 2d approach we still kept filmed elements for the horizon and more distant environment. This approach ensured that the scene had a level of realism that is often lacking in fully cg shots.

Can you tell us more about the shots in which Holmes and Watson are outside the train?
We had a single carriage on a green screen so we were able to film Jude and Robert leaving the train and climbing on the outside but that was it. We knew that we couldn’t use motion control for the environment as we couldn’t put the rig on a moving train so instead we used a libra mount on a crane arm attached to a real train and filmed approximate camera moves. These were then reworked and manipulated to match the principle photography in post. We were then able to add in the additional carriages and steam train which we built in cg to match the set carriage. We also added reflections and lighting onto the practical carriage as well as steam whipping past to give a sense of motion.

Have you used previz for the shot in which we are inside the gun?
Guy felt that the whole plot point of the lipstick blocking the gun and causing it to backfire was being lost amongst all the action of the train sequence so he wanted a shot to show explicitly what was happening. He shot the gun tracking round the gun barrel then said now you push in and you’re in Chas (Jarret) land. We then set about prevising some ideas in house for how the mini sequence would work. There were a few key points that we needed to show; the firing pin hitting the bullet, the bullet igniting and the bullet hitting the lipstick case. We worked on the assumption that we were using a miniature camera that could work in the tiny confines of the gun barrel and created camera angles and moves accordingly. We presented a number of ideas to Chas and then Guy before we locked down the sequence.

Can you explain to us in detail the creation of this shot?
Guy is hugely knowledgeable about guns so it was important that we got little details right like the shape of the firing pin and the indentation that is left on the bullet casing. We were in macro world so we were able to play with some beautiful shallow depth of field and close up surface textures. In reality the interior of a gun would be dark and visually uninteresting so we had to work on making the lighting and textures somehow believable and engaging. To achieve this we used extensive ray tracing and the Frischluft defocus tools for Nuke.

Can you tell us more about the creation of the fortress and the Swiss Alps?
The fortress was originally going to be a live action helicopter plate which we would then track into a cg alpine environment with a waterfall. The shots developed a bit and the camera moves changed so we weren’t able to used the actual footage but photographic reference taken during the shoot proved invaluable when recreating the fortress in cg.

What references did you received from the director for the fortress?
The fortress was originally based on Hohenwerfen in Austria but it felt too military so we added turrets inspired by the Neuschwanstein Castle.

How did you create the huge water fall?
Chas’ intention throughout the filming was to do as much practically as possible and to that end he literally scoured the Earth for suitable waterfalls. The Reichenbach falls from the novels was not impressive enough so we filmed waterfalls in Norway, then Yosemite and finally Canada, we even shot a few extra elements in Iceland. They needed a certain amount of manipulation but covered all the waterfall shots, apart from one.

Can you tell us in detail the creation of the impressive shot in which Holmes and Moriarty are falling in variating speeds?
This was always the biggest shot of the film. Guy wanted to see Holmes and Moriarty tumbling towards us and then enter a moment of almost frozen time as we fall with our protagonists and see their different facial expressions, Moriarty screaming in terror with Holmes in a zen time like calm. Knowing how close we would have to get to the actors digi-doubles were ruled out early on but to get the dynamism of the fall we couldn’t use traditional wire work. Chas had been interested in the precision and versatility of industrial robotic arms normally used for assembly line construction and wanted to use them for filming. So after some extensive in house previs work to make sure we hit the time ramp in and out of the slow down and the general camera move just right we got two « Kuka » arms and attached specially moulded fiberglass harnesses to them. We transferred our previs camera to a motion control rig and programmed the arms to rotate about each other using stunt doubles to match out animation. We wanted time to slow down, not quite to a stand still but slow enough that we could really focus on the faces and see droplets of water in sharp focus. It was also important that we used a multiple of 24 fps so we wouldn’t have interpolation problems when returning to real time at the head and tail of the shot so we shot at 432 frames per second on the Phantom Flex camera.

We pushed really hard to squeeze as much live action water into the shot as possible, the final section as they fall away from us was the most straight forward as the camera had almost come to rest so we were able to film a matched camera from Hunlen Falls in Canada. The central section looking straight on was slightly trickier but Chas shot some great elements with a specially prepared rig to dangle a Phantom camera on a wire from the base of a helicopter to allow it to get safely close enough to the falls. These allowed us to layer up and tile elements for this section. Even with all the filmed elements there were still two big sections which we needed to generate in cg. The top part where we start very close to the lip of the falls and a joining section as we come out of slow motion into the final framing. We used Naiad for the first and mixed it with Houdini mist elements. For the other section we had to match the look of the live action plates before and after seamlessly and one of the things that we found was that the waterfall very quickly became aerated and developed beautiful v shaped plumes with trailing mist. This was going to be very difficult to achieve using a sim based approach especially given that the water had already dropped 500 meters to reach that stage so we went for a more hand crafted approach in Maya and placed leading edges to match the footage with particle emitters driving the mist. The droplets in the frozen time section were made using Nuke’s new particle system which gave us enormous creative control and quick turn around.

How did you manage so many elements and the huge amount of data storage?
We used our in house asset tracking tools,had to be very careful about deleting outdated versions and used some big disks.

Have you develop specific tools for the particles?
A couple of tools fCloud which allowed us to reuse cached particle simulations in both Maya & Houdini and fMote, an elegant node based particle processing system. We also developed some tools to deal with the time ramping.

Can you tell us more about the breathes for the final sequence?
We tried to do this in camera at first so the whole set was kept very cold, the crew were all freezing but as soon as the lights went on and the action started in warmed up so much that no cold breaths were visible. At first I was nervous about adding the breaths in post as it was a very dialogue heavy scene and has been done with varying levels of success in recent films. We had a lot of shots to do so we needed a versatile system but also wanted to use real elements as much as possible. We used a variety of individual syllable breaths filmed against black and emitted them from a tracked source using the audio track to drive the timing. In this way we were quickly able to have a pretty good first pass of the whole sequence. Then we would select the appropriate element for each individual breath after which we made minor tweaks to get the look just right. This was packaged up to run in Nuke in a very controllable way so multiple artists could pick it up and get through the volume of shots needed.

What was the biggest challenge on this project and how did you achieve it?
The waterfall shot was definitely the biggest challenge and some of the unsung heroes were from the paint and roto department. The clean up was hugely difficult not least because the preferred performances for the actors came from different takes and had to be spliced together. The actors’ harnesses distorted their costumes and large parts of their bodies had to be replaced and although we were shooting on a green screen stage there was no way of preventing the actors from passing in front on wind machines or robot arms at some point in the move. Of course most of this had to be done at super slow motion as well.

Was there a shot or a sequence that prevented you from sleep?
I can be a bit grumpy if I don’t have enough sleep so I try not to let it get to me.

What do you keep from this experience?
Being dangled out of a helicopter to take stills photography for the recreation of the Swiss Alps was pretty memorable.

How long have you worked on this film?
About a year in total.

How many shots have you done?
Just over 500.

What was the size of your team?
100.

What is your next project?
Carl Rinsch’s 47 RONIN with Keanu Reeves

What are the four movies that gave you the passion for cinema?
THE GODFATHER, IL CONFORMISTA, BIG WEDNESDAY, WITHNAIL AND I.

A big thanks for your time.

// WANT TO KNOW MORE?

Framestore: Dedicated page about SHERLOCK HOLMES: A GAME OF SHADOWS on Framestore website.

© Vincent Frei – The Art of VFX – 2012

THE DARKEST HOUR: Dmitry Tokoyakov – Senior VFX Supervisor

Dmitry Tokoyakov began his career as a compositor and rose through the ranks to become VFX Supervisor. He worked many years at Bazelevs and oversaw the effects of movies like WANTED or APOLLO 18. In the following interview, Dmitry talks about his work on THE DARKEST HOUR.

What is your background?
I started out as a compositing artist at one of Russia’s biggest VFX houses of the time, then gradually moved to a supervising position. Many of our early projects were commercials, which gave me a good survival experience, when the work had to be delivered no matter what the obstacles were. Eventually, I was invited to work on WANTED, which was the biggest challenge that Russian VFX industry has faced. I’ve met Stefen Fangmeier, movie’s overall Supervisor.
Working with him, I’ve learned a lot and that led to a next step. After the show was wrapped, Timur Bekmambetov has offered me to lead his VFX house and put together a new team.
With Stefen’s advice and his Russian working experience, we set up a facility based on ILM’s management principles. That allowed us to complete over 700 shots for the next Universal’s feature, BLACK LIGHTNING, in only 4 months.

How did you got involved on this show?
After we put the oldtimer car in the air on BLACK LIGHTNING, I’ve met Chris Gorak, who came to Moscow to see the future locations for THE DARKEST HOUR. We started discussing the script and ideas came sparking in the air right away.
Our team has started developing all of the visuals for the project and shaping creative ideas into images and live action shots.

How was the collaboration with director Chris Gorak?
Chris has a strong artistic background, which made our work incredibly effective. Some of our concept artists didn’t speak English very well, but when Chris would sit down for a minute and put a few sketches over our concept art, there was no need for words any more. He was very open to ideas and has adopted many of them in the movie.

How was the collaboration with Production VFX Supervisor Stefen Fangmeier?
I was very happy to work with my mentor again. He joined the project at a later stage, when principal photography in Moscow was about to begin. Stefen’s experience is just overwhelming. He reminds me of a big sailboat captain, who can oversee the entire ship and not miss even a small detail. His jokes have kept the crew smiling even when we shot long nights.

What have you done on this show?
After completing the RnD and look development work, I worked on set as a second VFX Supervisor along with Stefen. We usually split between the two units. After the shoot was over, I was responsible for all of the work done in Russia.
Our team did all of the « shred » effects, when the aliens destroy humans, plus a number of « alien presence » effects, which is a subtle revelation of an alien force field.

Can you tell us more about this original alien design?
We have made about 50 different designs for the alien, using lots of approaches to the creature. Jelly fish, demons, spiders, all kinds of fantasy monsters, even microorganisms were used in designing an alien that had to be unique and would work with the movie’s energy-based idea. This was the « visible » design, when the alien is stripped of his cloaking shield. The actual movie features a different alien, designed by BUF in France.

What was the main idea for the Aliens?
The script had them described as energy-based beings. Our early discussions were mostly around degree of their physical visibility. Then we agreed that the aliens will be cloaked by some energy shield and come to light when that shield is damaged by the microwave gun. The main challenge was to make them feel physically possible, not magic. And, they had to be able to shred people too.

How did you create the impressive shred effects?
We started by putting different streams of particles together, playing with color and density. At some point, it formed into an interesting burst of ash that everybody liked. Then we worked a way to connect the shred to a character and designed the way a character gets snapped into the alien force field. Technically, it was a combination of Houdini particles and our in-house tools, which then was put together in Nuke.

How did you manage the particles aspect for the Aliens?
Our RnD team have written a C++ code that created interesting elements that consisted of millions of splines, overlapping into multi-dimensional gradients. Those elements were animated and combined with Houdini particles to produce passes for our compositing team. Also, we have enhanced the ash that the aliens have turned all of the people into. In some shots, we had to add or remove ash to keep continuity.

Did you create digi-doubles for this effect?
We shot traditional T-pose textures right on set, both for actors and their stunt doubles. Then we also scanned them in minutes using an Artec 3D scanner, which was actually developed in Russia.

How was filmed the stuntmen for the shred effect?
They were pulled using wire rigs in most cases, but for the multi-shred shot when the aliens rush though a crowd outside the club the stuntmen (including two girls!) actually jumped and fell on asphalt without any mats. We placed witness cameras to have enough data for rotomotion and tracking.

Did you used some interactive lights on set to simulate the aliens presence?
The aliens affect all kinds of electric devices, causing them to turn on and light up. In order to show the audience there was an alien approaching, every light bulb on set was rigged to a controller that would flash the lights around the virtual alien.
In some of the shots, we were asked to re-light the scene, so that the alien could move in a way different from what was intended on set. Having HDRI photography for every setup, we were able to achieve that.

How did you design the Aliens POV?
The aliens perceive everything as energy. So, they would clearly see a network of electrical wires or human nerve system. But, we had to make sure the audience would be able to see the location elements in the shot as well.
We’ve developed around 20 designs, combining particles, x-ray and proxy geometry with practical shots. Later, the team at Tata Elxsi designed another POV, which was used in the movie.

Can you tell us more about the Alien Towers?
Alien Towers are basically oil rigs that pump minerals and precious elements out of the Earth. Their designs were ranging from a concrete reinforced structures to a swarm of aliens. The towers were designed as a strong vertical element dominating over Moscow’s flat cityscape, showing scale of alien invasion.

Can you tell us more about the design for the Alien death?
When our heroes find a way to penetrate the alien shield and kill it, the resulting explosion bursts out a lot of debris, including a body piece that Sean later uses as a weapon.
We designed the piece, combining elements of bones, nerve system network and even fish scales. Based on the concept art, the piece was sculpted in Mudbox and physically printed using a 3D printer.

What was the biggest challenge on this project and how did you achieve it?
Having to create effects that had to be original, but line up with the movie’s idea of alien energy, we had to constantly think of the rules. With Chris and Stefen being over 6’000 miles away, we ran daily reviews and made sure our work matched their vision.
To make the shreds more violent, we sometimes had to switch to digi-doubles earlier in the shot that we expected, which was a lot of fun for everybody.

Was the movie initially intended to be 3D stereo?
Actually, it was planned to be a 2D release. But, with the AVATAR’s success, all studios naturally considered 3D a good option. But, since shooting 3D in far away Moscow meant a significant change in budget, the studio decided to shoot a test.
It was a short episode with two characters being chased and shred by an alien in an LA alley. We shot it with Sony F35 on a Pace rig. Then, the cut was handed over to our team in Moscow for VFX work. Shortly after presenting the completed piece, the studio has greenlit the show to be a 3D.

What was your experience with shooting 3D?
Most of times Scott Kevan (the DOP) shot with two rigs at the same time, which meant a lot of running for our data wrangling team. They had to make sure everything was documented, from the focal distance to the objects location and relative distances.
Being in the middle of a complex set, and making sure we don’t slow down the shooting process, our guys had to be ninja-accurate and fast. The data we provided was appreciated by colleagues from other VFX houses who relied on our records in their work.

How would you describe the VFX workflow on this movie from a 3D stereo standpoint?
This was a project that benefited a lot from the extra dimension. All of the particle effects worked were fun to create, especially when the shreds happened close to camera. Also, creating subtle alien presence effects in clear open of the Red Square was a challenge, since there was practically nothing to interact with the alien cloak. We have used CG ash to blow off when the alien passes by. Overall, the 3D stereo pipeline that we’ve built, have allowed us to quickly manage last-minute shots and comments and keep up the creative and technical quality to Hollywood standard.

Was there a shot or a sequence that prevented you from sleep?
Working with LA usually means staying pretty late for reviews, when the team on the other side of the world just starts their day. But still, we managed to run normal hours and even kept the weekends free of alien invasion as well.

What do you keep from this experience?
Most importantly, me and my team were the first ones in Russia to complete a large 3D stereo project. Being a part of an international team we stayed connected between Moscow, Los Angeles, Paris, Toronto, Prague and Mumbai.
It is always a pressure to run a major studio project, especially for a relatively new player on the international market, such as ourselves. But, we never surrender.

How long have you worked on this film?
My personal overall involvement was about 18 months and my team has completed all of our shots in 6 months.

How many shots have you done?
There was a total of 100+ shots that we worked on.

What was the size of your team?
We had about 30 people on the project, including management.

What is your next project?
Our team has moved to a new company to work on a number of local and international features. One of the projects is a story of the first Russian hydrogen bomb test, which will feature a very detailed look at the explosion and its effects. The other interesting projects feature a subway train crash movie and a children’s tale with lots of magic. We have also recently completed a movie about Vladimir Vysotsky, a famous Soviet singer and actor, who has passed away many years ago. This one had lots of face augmentation to make the audience believe he starred in the movie himself. It was a combination of complex make-up and digital face replacement.

What are the four movies that gave you the passion for cinema?
As a kid, I was greatly impressed by the STAR WARS and ALIEN. Probably those two have determined my career. WHAT DREAMS MAY COME has raised the imagination bar ever higher. However, my favorite movie is Jon Turteltaub’s PHENOMENON. I do believe that people tend to limit themselves and once you learn to step beyond the possible, you can make those dreams come true.

A big thanks for your time.

// WANT TO KNOW MORE?

Piastro VFX: Official website of Piastro VFX.

© Vincent Frei – The Art of VFX – 2012

Industrial Light & Magic The Art of Innovation – Pamela Glintenkamp – Abrams Books

Pamela Glintenkamp offers us with this book of 360 pages, a presentation of 35 years of groundbreaking work of the wizards from Industrial Light & Magic. The book begins on prefaces by Jon Favreau and Gore Verbinski.

The first part of the book focuses on the first ILM two decades (1975 to 1994). A short text explains the main changes in the visual effects created by ILM for each of the films from this period and reminds us of the various awards and nominations won by the studio. Those days are already fully covered by the books Industrial Light and Magic: The Art of Special Effects by Thomas G. Smith and Industrial Light & Magic: Into the Digital Realm by Mark Cotta Vaz.

The second part is divided into several chapters, one per year, covering a period from Casper to Rango. Each chapter introduces one to many movies. The author details the many challenges that ILM teams have to faced. The chapters are enriched with many explanations and testimony of ILM’s wizards. The iconography is truely rich with archival and behind the scenes pictures that allow us to admire the work done by the magicians of the studio. Throughout the chapters, the author also reviews the careers of many VFX supervisors of ILM and on the major advances achieved by the studio.

The book ends with a impressive list of awards as well as the long filmography of movie on which ILM have worked on.

In conclusion, its rich iconography and the many interventions of ILM wizards make this book a must-have object for all fans of ILM and visual effects.

Click on the image below to order this book:

© Vincent Frei – The Art of VFX – 2012

And that’s 2! The Art of VFX celebrates its new anniversary!

Hi everyone,

The Art of VFX has just reached its second year!

On this occasion, here’s an illustration done by Neil Maccormack (a big thanks to him!) and a slight change in the website layout and adding a new section that will be devoted to books.

I want to thank everyone who allowed me to make more than 140 interviews during the 2 years and my sponsor Autodesk. Not forgetting of course, the interviewees who gave their time to answer my questions!

A huge thank you all for your loyalty and your interest in my work. This is very important to me.
You are more and more to follow my interviews on Twitter, Facebook, Linkedin and Google+.

I’m looking forward about this new year and to offer you lots of new interviews.

Best regards.

Vincent

SHERLOCK HOLMES: A GAME OF SHADOWS – Seth Maury – VFX Supervisor – MPC

Prior to joining MPC Vancouver, Seth Maury has worked for studios like Warner Digital Studios, Cinesite and Sony Pictures Imageworks. His career counts films like HOLLOW MAN, BIG FISH, SPIDERMAN 2 or ALICE IN WONDERLAND.

What is your background?
I studied mechanical engineering in college and worked in product design for about 2 years before moving into visual effects. I first worked as a compositor at RGA Los Angeles, then Warner Digital Studios and Cinesite Hollywood. I then moved to Sony Pictures Imageworks, where I stayed for 11 years and worked as a compositor/lighter, CG Supervisor, and DFX Supervisor. Currently, I’m a visual effects supervisor at MPC Vancouver.

How did MPC got involved on this show?
MPC had worked together with Chas Jarrett, the show VFX supervisor, before on a few shows, so Chas was familiar with MPC’s abilities and strengths. For SHERLOCK 2, Chas was very interested in our in-house destruction tool Kali, that can produce very complex physical simulations, and which was a good fit for the work, such as the watchtower destruction.

How was the collaboration with director Guy Ritchie and Production VFX Supervisor Chas Jarrett?
Chas was great to work with, he had some areas where he was very specific, and other areas where he gave us a lot of creative freedom to develop the environments and ideas for shots. Chas always has a strong focus on the intent of the shot, is it telling the point it needs to? is it giving the appropriate sense of time period and ambience? Guy has a lot of confidence in Chas in that same manner, he knows what the shot needs to convey, and for Guy, the shot either works or it doesn’t.

What have you done on this movie?
MPC worked on approximately 400 shots across 10 sequences, the main sequences being Baker Street, Paris Opera House, and the Factory Complex/Forest Escape.

Can you explain to us in details the recreation of Baker Street?
The precedent for the look and feel of Baker Street had been set in the first movie, so we had to be faithful to that same feel while working on views that hadn’t been seen in the first movie. The most complex Baker Street shot was the establishing wide shot that ran after the main titles, where we start looking across the rooftops of London and crane all the way down to street level to find Watson walking up to 221 Baker St. We started with a live action plate shot at Leavesden, with a partial set piece for the front of the building. We also shot some motion control elements of people and horse and carriages to dress along the second block. We also had an art dept layout of the intended look and feel, so we photographed many buildings around London that we could use for source texture. We modeled about 20 CG buildings that we used to extend the street, as well as some digital matte painting and projection work to take those models further. The opening wide vista was a mix of 3D buildings, 2.5D projection, and 2D paintings for the very distant buildings. The cloudy sky is a panorama that I shot from my balcony here in Vancouver. The clouds here in Vancouver have a different feel than those in London, so we worked on the pano in comp to give it a London feel, but also to retain a sense of drama. The art dept was very specific that the wide vista of London have little to no trees, and very few white window frames, all in keeping with the Sherlock feel. Then the entire shot was brought together in comp with a lot of effort and balancing and sorting of technical issues from bringing elements from so many different sources together.

What were the real size of the sets for the sequences in Paris, and how did you create those huge environments?
It depends on which angle you’re looking, but for the most part, the sets were small sets built for the actors to be on, and then we built a CG environment around them. For example, the shot where we follow Holmes and Watson out of the Opera House towards the Hotel was shot in Greenwich, and the only building we kept from that was the door and surrounding stonework from which Holmes and Watson run out from. The square was then made up of CG buildings built from reference and measurements of the practical buildings around the actual Opera House, but we then changed the geography and layout to make it feel smaller and more intimate by taking out some roads, and moving the buildings in tighter. The rooftop set was a practical set piece about 20 feet long, which we then extended in CG, based on photos we took on the actual rooftop of the Opera House. For example, the shot where we crane up over Holmes, Watson, and Sim was only a big greenscreen and nothing practical to anchor that wide vista. We populated the square with digital crowd and horse and carriages, as well as period-appropriate street furniture and lighting fixtures. We also photographed awnings, doors and restaurants all around Paris, trying to find the elements that still felt period, and then we built the lower floors around the square made up of these different practical pieces we found.

How did you take the references?
The references were mostly shot by our in-house MPC photographers. We shot as much as we could from as many angles as possible. For example, we got permission to shoot on a scissor lift in the area and roadways in front of the Opera House, so we covered it from different heights and different angles all around, and we were lucky to be out there on very overcast days, so we didn’t have to paint out hilites and shadows from the sun.

Can you tell us more about the explosion at the Hotel du Triomphe? Did you used pratical elements?
The hotel explosion was made up a mix of practical elements and a cg building. We shot 2 different explosion elements that we planned to use, but they ultimately didn’t work well enough to sell the explosion. We ended up using layered library-element fireballs and we modeled a CG version of the 2nd floor upwards of the Hotel and married that with a practical plate of the ground floor. Having a CG building allowed us to get a lot more accurate interactive lighting on the building. Initially we thought these shots would not be too difficult, but they ended up being some of the more challenging ones to sell in terms of impact and reality.

How did you create the Opera House and populated it?
We weren’t allowed to film inside the Opera House, so Chas designed the shots to be filmed on a set piece with environment set extensions. We were allowed to take photo stills inside the Opera House, so we shot panos from multiple locations on the stage looking out, and from the audience looking back. We built a 2.5D environment inside of Nuke with those high rez panos and some basic interior geometry. For the crowds, we filmed shot-specific crowd elements on a riser. The riser was fixed in place and we moved the cameras around the stage to get the different crowd plates we needed to build a full auditorium of people.

Can you tell us more about the creation of the huge weapon factory?
Most of the shots were made up of practical location plates with cg set extensions. We had a mixture of wide establishing shots and tighter set extensions. The work was in 3 main areas: the firing range and watchtower, the rail yard, and the approach to the factory. When I first met Chas, one of the first books for reference he showed me was about the Krupps factory, filled with pictures of how the factory evolved over decades, organic in shape and layout, and also descriptive of how it was essentially its own functional town with living spaces, tailors, food stores, etc. Our goal was to recreate this organic feel, to make our factory feel as if it had grown over years, and was a fully functioning small factory and city. So in our wide layouts we had areas that were more concentrated with industrial type equipment, furnaces, smokestacks, large warehouse type buildings, and then areas that were more living quarter type areas. We built about 20 cg buildings to different levels of detail, that we used to dress the different views, both wide and tight, along with many street props, furniture, and signage. Our cg buildings were all based on practical factory buildings that the art department had scouted at Sleaford Maltings, Bollwiller, Bitterweld, Dessau and Chatham Docks. Our MPC photographers photographed them in detail and then we recreated them in cg, some as full 3D models and some as more simple 2.5D buildings.

For the watchtower, we had concept art from the art department, as that building never existed practically in a location, though we had a partial set piece that we had to be faithful to. In general, for the look of the factory, we went through a lot of different permutations of a factory that was thick and dense and overhung with smoke to more thin, vast spaces with overcast semi-oppressive feeling skies. This was an area of exploration that evolved as we worked with all the different location plates, as we had to find a consistent look and feel that would work for al the practical photography. I definitely had a stage where I went over-the-top with smokestacks, the factory had veered into looking like a concrete forest, but it was all part of the experimentation.

Some shots involved impressive destruction elements. Can you tell us more about the use of Kali for those shots?
Our biggest destruction shot was the watchtower falling over after it had been hit by the Big Bertha shell. This was an all cg shot. From the early previs, we had an idea of what the shot needed to be and how it would be framed. We knew we needed to sell the scale and weight of the tower going over as it crashed into the Surgery building. Both the watchtower and the surgery building were modeled to be Kali friendly, and our fx lead Georgios Cheruvim, first started with simple geometry representations to get the timing and blocking right, and then we progressively refined the sim to get more detail. Chas was good in pushing us towards practical reference of the way smokestacks seem to bend over the length before they start to give way locally. Once we were about 90 percent happy with the general motion of this shot, we then started working on the smoke, which I felt would really be the thing that helped give a sense of scale and weight to the shot. We studied a lot of reference of building demolitions and we noticed how strong, violent and dense the first rushes of dust and debris were from impact, so we aimed to recreate that feel in our cg smoke.

How was filmed the impressive slo-mo Forest escape sequence?
The forest footage was shot on a Phantom camera, upto approximately 2500 frames / second. The editor would take the raw footage and give us mockups of the retimes he wanted, which we would match to. For some of the shots where we punch in and out in the same shot, Chas used 2 Phantom cameras mounted side by side on a speed rail, one with a wide lens, and one with a longer lens, and then we again matched the mockups the editor created to mix between the plates and then we smoothed out those transitions in comp. In addition to the raw footage, Chas shot Phantom elements of mortar and dirt and trees exploding, that we then mixed into the plates, which already had some of those effects as practical special effects. We also had practical Phantom bluescreen elements of bullets hitting wood trees, some of which we used, and then we also created some cg bullets for some shots and augmented the destruction with the bluescreen tree elements. Our comp supervisor Mark Richardson handled a lot of these shots.

Did you encounter some troubles with those slo-mo footages and Kali?
One of the slo-mo shots in the forest involved an exploding cg tree, which we used Kali for. The shot went through exploration until we landed on the shot that’s in the movie, but we decided early on that we would simulate and render all of the frames so that we could do repos and retimes in comp to make the turnaround quicker. This meant that we were rendering thousands of frames, most of which were discarded once the retimes were applied. Shots that long usually have issues even if only because they are so resource intensive.

What was the biggest challenge on this project and how did you achieve it?
The most challenging aspect of our work was in convincingly recreating the wide vistas of period London, period Paris and the wartime Factory, both daytime and nighttime. We had to pay attention to all the macro ideas and concepts and more specifically to all the small details that ultimately help sell the shot, such as roof material types and construction, chimneys, city density and layout, lighting sources available at the time, appropriate street and building signage, for example. These types of shots involve first a lot of research, and then continuing research throughout the entire pipeline, as different details become relevant as the shots progress..

Was there a shot or a sequence that prevented you from sleep?
There was a period during post production where we had been working on the Opera House square and trying to do a lot of that work as 2.5D projections, but we realized that we needed to go back and do full 3D builds of those buildings to really get the flexibility and look we needed. That kept me awake a few nights.

What do you keep from this experience?
I learned a great deal about set extension work. I had previously done lots of creature and character work and some cg set recreation, like the city work on SPIDERMAN 2, but for this show, there was more variety of set extensions and they were also not meant to be present day, so it was not as immediately familiar as other set extension work. Also, it was great to work with Chas and Laya Armian, the show’s VFX producer, as they made it a very enjoyable experience for me.

Which branches of MPC have worked on the show?
MPC Vancouver, MPC London and MPC Bangalore

How long have you worked on this film?
I worked on the film for a little over a year.

How many shots have you done?
MPC worked on approximately 400 shots across 10 sequences, the main sequences being Baker Street, Paris Opera House, and the Factory Complex/Forest Escape.

What was the size of your team?
The main production team was made up of Doug Oddy, our in house producer, Danielle Kinsey, our in-house production manager, and Jennifer Fairweather, our show coordinator. Our compositing supervisor was Mark Richardson. The artist team consisted of approximately 30 compositors, 10 lighters and look dev artists, 5 texture artists, 12 DMP artists, 5 modelers, 5 fx artists, and approximately 75 people comprising roto, rotomation, matchmoving and paintouts, spread between Vancouver, London and Bangalore.

What is your next project?
Thats confidential, stay tuned!

What are the four movies that gave you the passion for cinema?
JURASSIC PARK, first and foremost. I was working as a mechanical engineer at a product design firm, and I went to see JURASSIC PARK and had that moment of « Holy s**t, how’d they do that? » And so I quit my job shortly after that and eventually found an entry level job at a VFX company. Otherwise, as for movies, it’s probably the original 3 STAR WARS movies. Again, they were something that people hadn’t seen before, and on such a grand scale.

A big thanks for your time.

// WANT TO KNOW MORE?

MPC: Dedicated page about SHERLOCK HOLMES: A GAME OF SHADOWS on MPC website.

© Vincent Frei – The Art of VFX – 2012

IMMORTALS: Danielle Plantec – VFX Supervisor – Scanline VFX

Danielle Plantec is in the VFX for over 16 years. She has worked in studios as Digital Domain, Sony Imageworks, or ESC. In 2004, Danielle joined Scanline VFX and worked on films like THE DAY AFTER TOMORROW, 300 or HEREAFTER. She received a VES Award for this film.

What is your background?
I’ve been in Visual effects for 16 years. I was an effects artist at Digital Domain, Sony Imageworks, The Orphanage and at ESC for the MATRIX films. I joined Scanline Munich in 2004 as a CG Supervisor, and specialized in « Flowline » (Scanline’s Fluid effects system). I then opened Scanline’s Los Angeles Office in 2008 and our Vancouver Office this July. I am currently Vice President and VFX Supervisor at LA and Vancouver facilities.

How was the collaboration with director Tarsem Singh?
Tarsem has a wonderful eye and isn’t afraid to stray from the norm. He had some fantastic ideas which he planted and set us free to bring them to fruition. A perfect example was the Tsunami. Tarsem described it as a cross between an atomic bomb blast and a macrophotography slow motion droplet that generates a Tsunami in the Oil (contaminated ocean). He had a clear idea of what he wanted, but also gave us a lot of leway to design the simulations, the look and even what the shots would be, from composition to introducing additional shots to tell the story. I found him to be very supportive and easy going with a wonderful sense of humor.

What was his approach about VFX?
Tarsem chooses a team he respects, gives them the description of what he is after and then lets them use their creativity to execute that vision. He even said things like « do what you do best ».

What have you done on this show?
We did Mount Olympus, the Tsunami, The Interior Tomb environments and destruction and the Mountain collapsing and destroying the army.

Can you tell us more about the free fall of Poseidon and his impact on the water?
We created a digital version of Poseidon to intermix with the Close up practical shots. The environment was mostly matte painting with 3D volumetric clouds and full 3D rock terrain for everything close up, because there was a lot of parallax, that could not be painted.
For the Impact we were after something that was a cross between an atomic bomb and a macro photography droplet made out of a thick oil. We started with droplet idea and expanded on it by adding in shockwaves and thick mist of oil atomizing from the impact to add the atomic bomb feel and sense of scale.

How did you create the Tsunami?
We handled the wave in two different ways. For the wide shot we started with a continuation of the macro-photography idea that we used for the basis of our impact shot, and continued the simulation. This gave us our base, and of course we did it on a huge scale so it was also much slower. Then as we developed the look of what it would be in the close up shots we realized we wanted to have the swirling oily surface, so we simulated another fluid simulation on top of the base droplet style simulation. Then as development went on people started to want to see some of the additional elements we use in our water waves, like breaking spray and mists, so we added all of these on top of the base.

Did the black oiled aspect of the water caused you some troubles?
Tarsem had originally chosen black oil because he didn’t think CG could do realistic enough water. I remember meeting with him in Montreal and he laughed about it because we had recently seen our water work from 2012 and HEREAFTER and said he was trying to make it easier for the fx guys by making it oil. They had already coated the sets in the thick gloppy oil substance to show the aftermath of the tsunami. Water is one thing we definitely have down. So for us it is easier to make something match reality than to design and art direct something that doesn’t exist in nature. That being said, we had a lot of fun with it. We got to come up with something new.

How did you create the huge environment that is flooded by the wave?
We used a combination of 3D geometry and matte-painting, and in some cases full 3d geometry with a nice gloppy oil coating so we could get the oozing interactive feel on the rocks.

What were the references and indications you received for the Titans Temple?
The production had a wonderful art department, and they had supplied us with some beautiful concept art to define the look and mood of the temple. Most shots were also predefined by detailed previs that we kept as true to as possible.

Can you tell us in detail the impressive Temple destruction? How did you manage so many elements?
We handled many of the simulation elements separately and put them back together in comp. For example, when the large piece of basalt fell from the ceiling we handled that piece separately, and then did a secondary fine scale rock and dirt pass, followed by a detailed a dust fluid simulation.

Did you used some live action footages or is it full CG?
No the Temple destruction is full CG. Zeus is real, but he’s very tiny in frame.

Did you shared some assets with other vendors especially for the Temple and the Mountain?
The Mountain set was designed by a matte-painting company in Canada, for several sequences earlier in the film. We then took their paintings and built out underlying rock geometry so we could destroy the mountain. In order to get a realistic feel we needed to go from what in a still could be a matte painting to full shaded geometry.

Inside the temple we provided the environment work for the Amazing titan battle sequence from Tippett. We had a couple of shots where our destruction squashed some of their titans. In those cases Tippett provided us guide geometry of what would hit the titans, and then we replaced it with real destruction pieces. and dust elements. They also provided with everything we needed to be able to integrate their characters into our environment in comp. Because everything was so nicely separated out we were able to get lot of depth and destruction and dust elements wrapping around and between their characters.

Was there a shot or a sequence that prevented you from sleep?
Going into it I thought designing Mount Olympus would be our easiest sequence. I was very wrong about that. This was the one that had some designs, but really wasn’t defined. We went through dozens of completely different styles before we narrowed in on the final result.

How long have you worked on this film?
On and off for a year with a skeleton crew at times, and a complete hiatus for a couple of months while.

How many shots have you done?
About 150.

What was the size of your team?
Our artists all work on multiple shows, so it’s difficult to say. We had a fairly small team with a few stops and starts while the edit was being designed, We worked over the course of a year on and off with a team of about. Approximately 30-35 artists.

What is your next project?
We have several great shows in the moment that are delivering in February, but most of what we do is fairly secret until after the movie is released.

What are the four movies that gave you the passion for cinema?
Not so much Cinema, but Visual Effects.. there are so many along the way that re-spark the love of VFX

Early inspiration:
THE LAST STAR FIGHTER
BIG TROUBLE IN LITTLE CHINA
BLADE RUNNER
ID4 – INDEPENDENCE DAY

Rekindling the VFX spark:
THE MATRIX
FIGHT CLUB

A big thanks for your time.

// WANT TO KNOW MORE?

Scanline VFX: Official website of Scanline VFX.

© Vincent Frei – The Art of VFX – 2012

THE GIRL WITH THE DRAGON TATTOO: Sean Andrew Faden – VFX Supervisor – Method Studios

Sean Andrew Faden began his career in 1996 at Digital Domain where he worked on films such as THE FIFTH ELEMENT, TITANIC, FIGHT CLUB or RED PLANET. In 2004 he joined the team of Asylum Visual Effects and participate on films like CHARLIE AND THE CHOCOLATE FACTORY, THE CURIOUS CASE OF BENJAMIN BUTTON or DOMINO. In 2009, he joined Method Studios and oversee the visual effects of films such as LET ME IN, GULLIVER’S TRAVELS or CAPTAIN AMERICA: FIRST AVENGER.

What is your background?
I began my career at Digital Domain in 1996, straight out of engineering school, and worked on numerous projects using Prisms, and then Houdini to create complex procedural animations and fx animation. Eventually, after leading several fx and scene setup teams, I began CG Supervising, and VFX Supervising on both features and commercials at Digital Domain and then Asylum Visual Effects. I have been at Method Studios for almost 3 years now, taking a lot of the commercial production tricks learned at Asylum and applying them to both our feature and commercial projects.

How did Method Studios got involved on this show?
Our show’s lead matte painter, Wei Zheng, had worked on THE CURIOUS CASE OF BENJAMIN BUTTON and ZODIAC with David Fincher and he asked Wei to help out on DRAGON TATTOO. Of course we were very excited for the opportunity – I was especially happy to start what would be my 6th David Fincher project.

How was the collaboration with director David Fincher?
In the beginning, we were sending up mock-up frames for each sequence that Wei and our comp supervisor, David Rey, would create using a combination of Photoshop and Nuke. Fincher was incredibly available, responding at all hours of the day to our postings and guiding us to his vision for each setup. As the post progressed, we began meeting with Fincher once a week at the DI house, reviewing our shots. This was a relatively quick process, as he would generally keep his comments brief and very specific. By the end of the job, we were showing up twice a week for reviews. David is a rare filmmaker that really gets what we do and was always very clear with his direction, while still being very open to our suggestions.

How was the collaboration with Production VFX Supervisor Ruslan Ogorodnik?
He was not involved in the work we were doing- we dealt directly with Fincher.

What have you done on this show?
Method completed about 110 shots in total, mostly replacing green screens on interior shots to create BGs of downtown Stockholm, Hedestad, and London with varying times of day and weather conditions.We did all of the BGs for all of the Millenium Newspaper interiors and office window comps on the exteriors, the Milton Security interiors, Martin’s house, and Surveillance van interiors.One notable all CG shot was the wide shot following the train across a snowy landscape towards Hedestad.

Can you tell us more about the wide shot showing the train and the Swedish landscape?
That shot took quite a while (about 3 months) to complete. It was a lot of CG snow and fog elements layered upon a terrain that was multiple matte paintings projected onto a landscape geometry. The projected matte paintings were helped by having thousands of tree cards generated in Houdini as well as some procedural tree shapes to make convincing edges of mountains etc. Our Matte Painter (Wei) also created a few custom trees which he projected on multiple cards for hero bits. The train and tracks were modelled, animated and lit in Maya using Vray. Interactive snow for the train was generated in Houdini. The final composite was done by David Rey in Nuke, with one of the greatest challenges being to bury the environment in snow without losing detail and realism.

Can you tell us more about the matte-paintings? How did you create them?
Matte paintings were always sourced in some photography – we had some production reference of Sweden, but a lot came from our own research – some shots of Stockholm’s harbor were shot by one of our compositors when he was visiting family in Sweden. Each sequence required a look frame to be made which we would go back and forth with Fincher until we had something we agreed on. From there refinements were made to make sure it would hold up for a sequence and additional perspectives were painted where necessary.

Have you filmed some elements for the backgrounds?
Not specifically, although production did provide photographic reference when available.

What were the advantages of using both Nuke and Flame?
We were able to use Nuke on a lot of shots where there were similar setups, greatly benefiting our overall efficiency. Our Nuke artists were able to easily share keying setups and color setups between similar shots within a sequence. There were a few difficult one-off shots which stayed in Flame as well as some self-contained sequences. The Flame gave us a very quick feedback on shots where we were making a lot of small adjustments such as the nighttime Milton Security BGs, where we added moving lights etc. to give the matte painting life.

Can you tell us more about the CG snow?
We created CG snow for 2 shots – the big train/landscape shot and an additional landscape shot that did not make the final edit. The snow, created in Houdini and rendered in Mantra, was comprised of several layers including: multiple layers of instanced snowflakes on particles, a corresponding volume layer, fog for BG atmosphere, and heavy sheets of particles instanced into the distance to create additional texture in the atmosphere. For the train, we generated volumetric snow getting kicked out by the train wheels as well as volume snow blowing off of the train roof. These elements were generated using Houdini’s DOP volumetric sim tools.

What was the size of your team?
At our largest we were 6 compositors (we averaged 4 however), 2 matte painters, and 2 CG artists

Your compositing team is pretty small for so many shots. What is your way to be so efficient?
We tried to focus our attention on specific shots within a sequence before getting ahead of ourselves trying to do too much at once. Once we had David Fincher’s blessing, we would go full force to bring the rest of the shots up to that level.

What was the biggest challenge on this project and how did you achieve it?
The big train shot was the most difficult since it literally began from a concept sketch. It was a slow process, but we did it properly and were able to work with Fincher to come up with something that in the end is pretty convincing.

Was there a shot or a sequence that prevented you from sleep?
Not really, I was always confident in our team, and had fun every day going through the shots, looking for new ways to push the realism.

What do you keep from this experience?
We had an amazing team on this one, and we really tried to be efficient despite tireless efforts to hit the look and quality that David Fincher would expect.

How long have you worked on this film?
Method worked on the film between mid August and mid December (about 4 months).

How many shots have you done?
110.

What is your next project?
I’m currently busy on a few Super Bowl spots and reading a lot of scripts for potential new feature work.

What are the four movies that gave you the passion for cinema?
RAIDERS OF THE LOST ARK, THE TERMINATOR, THE FISHER KING and THE SHAWSHANK REDEMPTION.

A big thanks for your time.

// WANT TO KNOW MORE?

Method Studios: Official website of Method Studios.

// THE GIRL WITH THE DRAGON TATTOO – SHOT BREAKDOWN – METHOD STUDIOS

© Vincent Frei – The Art of VFX – 2012

THE GIRL WITH THE DRAGON TATTOO: Tim Miller – Creative Director & Co-founder – Blur Studio

Tim Miller is the co-founder and creative director of Blur Studio. He directed the DANTE’S INFERNO TV spot for the Superbowl and more than 10 minutes of cinematic of this game. He has also directed the game cinematic for DC UNIVERSE ONLINE.

What is your background?
Originally I was an illustrator and animator at the dawn of the digital age. Currently I’m the creative director and one of the owners of Blur Studio in Venice.

How did Blur got involved on this show?
David has been a friend of the studio for years and we enjoy working together. Early on he had ideas for what he wanted this title sequence to be and thought we’d be a good fit for it.

How was the collaboration with director David Fincher?
First of all it’s an honor; in my opinion he is one of the greatest directors working today and his creative sensibilities are unmatched in my opinion. But less documented is just how great a collaborator he is; very open to ideas and new ways of approaching things. He’s a great guy and one of the funniest people I know.

What references and indications did David Fincher gave to you for the main title?
“Imagine James Bond if he was a 22 year old, bisexual cutter.” What more could we possibly need? (kidding) We also knew we were using Trent Reznor and Karen O’s cover of “Immigrant’s Song” which gave a lot of structure to this right from the start. But finding the look of this piece was a very collaborative effort. He’d give us his thoughts… we’d collect reference and ideas and toss them to David… he’d cull what he liked from those… we’d further refine… etc. And relatively quickly – we honed in on the key visual concepts.

What was your approach for the title?
David wanted to tell the entire story of the trilogy in 2.5min. He wanted it to be all CGI and he wanted it to be a dream/nightmare sequence. Those were the big creative prerequisites at the start of the film. And of course we had a budget and schedule as well.

Can you tell us step by step the creation of the title?
Overall I want to point out that most of these steps were heavily overlapping, none were done in isolation. It’s important to think of the process as having structure but also an organic nature that allowed for creativity and flexibility. The steps included:
1. Discussion of the general ideas in the broadest possible ways.
2. Visual reference gathering; tear sheets, samples, examples, etc.
3. Define the key moments and ideas from the trilogy.
4. Come up with a host of abstract ideas to represent those key moments.
5. Storyboards/visual development/animatics all started at the same time and overlapped heavily.
6. Modeling started first on known elements and then other assets as they became relevant.
7. R&D on fluid and VFX was ongoing.
8. Animatics and editorial started early and was adapting throughout the process.
9. As models were rigged animation was getting started. We tried to handle this as “camera agnostic” as possible so that we’d have the flexibility to choose any angle at any time.
10. As animation was finished, fluid simulations in RealFlow began in earnest.
11. Final lighting and rendering was set up and then made final as the various sequences worked their way through from other departments.
12. High-fives all around.

Have you prevized the whole sequence?
Yes, previs started almost right away and continued deep into the project. Once we had a working previs edit we continued to refine it with better camera angles and then final animation, FX and renders. The animatic edit was a living thing throughout the process.

Can you tell us more about the shooting?
There was no shooting. The titles are 100% CGI.

How did you create the various CG elements?
The characters were scanned and then finished up in Zbrush and 3ds Max. All other elements – like the dragon – were created in Max and zbrush. We rig and animate in Softimage. RealFlow was used for all of the fluid simulation.

Have you develop specific tools for this project?
All the software used was off the shelf.

How did you create the fire and the different liquids elements?
the liquid was RealFlow, the fire and smoke was all Fume FX and the rigid body simulations were RayFire and a variety of particle systems.

Did you have the soundtrack since the beginning to help you with the rhythm and the final edit?
Yes, which gave us an amazing head start.

What was the biggest challenge on this project and how did you achieve it?
The fluids were by far the biggest challenge. Our FX team uses RealFlow for other projects regularly but not to the extent that we needed for this. So getting that sorted was a huge priority. Secondly it was just an organizational issue. The average Blur cinematic is about 15-20 shots per minute. This piece has over 250 shots in a 2.5 minute duration.

What do you keep from this experience?
The same thing I take away from every project – how lucky we are to have such an amazing group of artists here at blur.

How long have you worked on this film?
The project took about 3.5-4 months from start to finish.

How many shots have you done?
252 shots.

What was the size of your team?
About 40, but it ramped in and out.

What is your next project?
Top secret!

What are the four movies that gave you the passion for cinema?
BLADE RUNNER, ALIENS, GLADIATOR and CYRANO DE BERGERAC.

A big thanks for your time.

// WANT TO KNOW MORE?

Blur Studio: Official website of Blur Studio.

// THE GIRL WITH THE DRAGON TATTOO – MAIN TITLE – BLUR STUDIO

// THE GIRL WITH THE DRAGON TATTOO – MAKING OF – BLUR STUDIO

© Vincent Frei – The Art of VFX – 2012

J. EDGAR: Ollie Rankin & Geoffrey Hancock – VFX Supervisors – Method Studios

Geoffrey Hancock has been evolving for over 10 years in visual effects, he worked at Gajdecki Visual Effects, Rainmaker and CIS Vancouver, the latter of which is now Method Studios. Geoffrey has worked on films like I Robot, VANTAGE POINT or NIGHT AT THE MUSEUM. This is his third collaboration with Clint Eastwood, he oversaw the effects of CHANGELING and INVICTUS.

Ollie Rankin worked at Weta for the first two LORD OF THE RINGS and then joined the teams of MPC and has worked on films such as KINGDOM OF HEAVEN, WATCHMEN or HARRY POTTER AND THE HALF-BLOOD PRINCE. For the occasion of INVICTUS, Ollie joined CIS Vancouver.

What is your background?
Ollie Rankin (OR): I’m from New Zealand, where I studied computer science and specifically artificial intelligence, a background which got me my break working on the team that developed the Massive crowd pipeline at Weta for THE LORD OF THE RINGS. Since then I’ve worked in California, Tokyo, and the UK before joining Method Studios in Vancouver three years ago as a CG Supervisor on INVICTUS, which was our most recent collaboration with Clint Eastwood and Michael Owens.

Geoffrey Hancock (GH): I began working in Vancouver in 1994 doing architectural visualizations and commercial work. Subsequent years of fast paced television series work and smaller films requiring a wide range of effects has given me a solid background for today’s more specialized effects required by larger feature films.

How was the collaboration with director Clint Eastwood and Production VFX Supervisor Michael Owens?
OR: Once again it was a great pleasure working with these two. Clint pretty much entrusts the entire responsibility for his films’ vfx to Michael, so we know that when we have completed a shot to Michael’s liking, it will almost always be blessed by Clint. This is the only director/vfx supervisor pairing I have ever worked with where such trust exists.

GH: On the three films (CHANGELING, INVICTUS, J. EDGAR) that I’ve had the pleasure to work with the Malpaso Production crew, I’ve been struck by the professionalism and dedication they exhibit both on set and through post production. We primarily deal with Michael Owens on a day-to-day basis, both remotely from his home and in the studio when he comes to Vancouver. Through the early part of post we are often in close contact with their editorial team, as sequences often evolve once the visual effects shots come together and options for expanding the story are discovered. One of the cornerstones of our work together is the trust between all parties involved and the excitement of working on great stories.

What was their approach about visual effects?
OR: For these filmmakers, visual effects is always used in service of the story. Since Clint is very focused on the human drama aspects of the story, it often falls to the vfx team to create the larger shots that provide a broader context and scale.

GH: Visual Effects seems to have become a comfortable tool for Clint Eastwood. As he’s had previous success with Michael and us using visual effects to expand his films, we have seen more types of work being suggested and relied upon.

How did you create the wide shot of burning Chicago?
OR: The shot of Chicago during the 1919 race riots is a good example of this. Having tried a number of different camera angles during the concept phase, we arrived at a particular view of the city that satisfied two key requirements. The shot had to be a broad enough vista to take in the scale of the riots in relation to the city as a whole, while remaining close enough to the action to convey the threat. Our matte painter pored over period photos and postcards to compose an over-sized image of the city, which we brought to life with footage of real fire and smoke. Then we added further small details like dancing shadows and window reflections and a slow pan to allow the audience to take in the scale.

GH: A lot of time and thought went into the way the audience’s focus would travel and shift across that shot. We wanted to know that key moments and elements would be absorbed. The timing of the explosions, the composition of light and shadow and the way this shot tied in with the preceding shot’s composition were all used to direct the audience’s eye.

Can you tell us in details the creation of the two motorcades sequences?
OR: The most poignant way that Hoover’s personal political power is conveyed in the film is through his relationships with the various presidents during his reign at the head of the FBI and the two inauguration parades serve to contrast his early influence against his later decline. It was therefore important that the two parades echoed each other visually. The film crew was able to get onto the actual balcony of Hoover’s old office in the Department of Justice building and shoot a couple of reference plates of the Pennsylvania Avenue parade route as seen from that perspective. Based on that viewing angle we set about creating two different period versions of the intersection below and the view beyond. Specifically we had to research the ages of the various buildings to determine which ones belonged in each era, create trees whose size reflected the appropriate age, build roads and traffic signals and shop fronts appropriate to both periods, populate it with crowds and vehicles and then decorate it all with moving flags and bunting. We specifically motion-captured an actor performing the iconic Nixon victory pose, with both hands raised above his head in “V” signs and we applied this motion to a digital character built to look like the actor who plays Nixon in the film.

GH: Both motorcade shots are entirely virtual, not having used the reference plate as a starting point. This freedom of composition and timing, compounded with a near limitless amount of details to be art directed, is always a long process with many people involved. It was especially interesting in this case to be playing the two shots off each other for contrast of emotion. A lot of thought went into bridging the lighting direction between surrounding shots to allow for the desired mood in these shots.

How did you create the crowd?
OR: We developed a proprietary crowd system and a library of reusable crowd assets for INVICTUS which we have extended on subsequent projects. This served as the basis for our CG crowds on J. EDGAR. Our crowd system, which is built on top of Massive and Houdini, is optimized to handle what we think of as the ‘bread and butter’ shots that constitute the majority of CG crowd work. That is, those shots in which a large space needs to be populated with an assortment of people, in a mixture of clothing styles, mostly sitting or standing on the spot and exhibiting a variety of behaviors. By pre-caching these type of behaviors and designing our crowd characters and wardrobes in a modular mix-n-match type fashion, we are able to very quickly populate a space with people, art direct their demographic and clothing distributions and choreograph their behavior.

GH: Having extended our crowd system on numerous films it is wonderful to see it in full swing once a film like J. EDGAR has fed it with new wardrobe periods and new motion capture behaviors. Michael often asked us to stir up the equation and lose the odd balls, which we can do quickly and accurately, allowing for a lot of art direction of the look and behavior of a crowd.

Did you create some CG cars?
OR: The opening shot of the movie, which establishes the Department of Justice building where Hoover worked his entire career, was originally conceived to take place in the 1930s. So the film crew had sourced a collection of vintage cars, then halted all contemporary traffic and filmed a number of takes of the vintage cars driving past the building. However, when they came to edit the film, they realized that they actually wanted to use the shot in a sequence that was set in the 1960s. So we cleaned out all of the modern and vintage cars and filled the plate instead with CG cars built based on models that were around in 60s.

GH: There are also a number of CG cars in the motorcades and throughout the Broadway movie premiere scenes. We had a lot of good success with the Craft Animation tools which helped create believable suspension and road interaction.

Can you tell us about the Broadway shots?
GH: These virtual shots were another case of how we focused on directing the eye to key story points like the premiere sign, and to other marquees and areas of dense population, so we could show instantly that this is a happening entertainment district. These highly reflective night time scenes were primarily rendered in Mental Ray, with the crowds coming from Houdini.

What were your references to create this iconic place?
OR: A very similar strategy to that employed for the parades was used to create a fully CG Broadway. However in this case we didn’t have a specific section of Broadway to match and our only live action cue was the theatre in LA where they shot the actors leaving the theatre. So this was what we had to match to in the first instance. Beyond that, we drew upon period film footage and photos and rather than trying to recreate an exact replica of a recognizable section of Broadway during the period, we instead created something representative of the time and place, thereby sidestepping rights issues.

Can you tell us more about your work on the gunfight?
OR: For the shootout between cops and gangsters, the actors fired machine guns that shot blanks, giving quite realistic muzzle flashes, but all the bullet impacts in the walls and people were added by us including puffs of dust and spurts of blood. For the shot in which a car gets totally riddled with bullets, not only did we have to add bullet holes, breaking glass and smoke, but the man inside the car was added in post along with the vibration of the car caused by the repeated bullet impacts. In this shot we even took advantage of the extra footage available in the soundtrack region of the anamorphic frame, to improve the composition.

What was the real size of the racetrack?
GH: It was at the Santa Anita Racetrack in Arcadia, California, that Michael shot the plates of the horses. They used a one mile oval. This plate was primarily used for the horses, as the surrounding area was redressed in the matte painting.

How did you create the racetrack and the huge environment around it?
OR: The principals and a small group of extras were shot on a partial set, while the horse race was shot separately at an actual racetrack which was modified with matte painting to fit the time period and to add the hills that overlook the location we were recreating. Finally a CG extension was built for the stadium and this was populated with a digital crowd in period wardrobe.

What was the biggest challenge on this project and how did you achieve it?
OR: The biggest challenge on this project was logistical, rather than technical; that it required such a diverse variety of types of shots, taking place in different locations and filled with different types of content. This permitted us very little reuse of assets and methodologies. We overcame this by splitting the CG crew up into smaller teams that focused on particular sequences and types of shots.

GH: Personally, one of my big challenges after facing multiple health issues that took me out of the mix of daily production, was needing to turn over the reins of the project. Luckily I knew I could trust Ollie with the task of completing a great film and it was gratifying to see how well everyone worked together when things got down to the wire.

Was there a shot or a sequence that prevented you from sleep?
OR: The two fully CG inauguration sequences went right down to the wire and involved some rather late nights toward the end of the schedule. So I guess that quite literally prevented us from sleep. The sheer number of layers that went into those shots is staggering, especially when you consider minutiae like the fact that every overhead flag needed to cast its’ moving shadow on every car and person on the road and everything in the whole scene needed to reflect in the windows of the cars and buildings.

What do you keep from this experience?
OR: Several members of our crew were struck down by health problems during the course of this production and in fact it was Geoffrey Hancock’s appendix bursting that brought about my field promotion to vfx supervisor. I was very proud of the way that those of us left standing pulled together as a team and filled the gaps to get the project out the door on time and to our clients’ satisfaction.

How long have you worked on this film?
OR: A little over 7 months.

How many shots have you done?
OR: Around 150.

What was the size of your team?
OR: We had 46 artists, 6 technical support people and a production team of 6.

What is your next project?
OR: I’m currently working with Jim Madigan as supervisor on G.I. JOE: RETALIATION.

GH: Having taken some time to recuperate I’m enjoying a new challenge of effects work in the future as opposed to the past for a change on CLOUD ATLAS.

What are the four movies that gave you the passion for cinema?
OR: Like many in our industry, as a kid I was obsessed with Sci-Fi. It’s hard to narrow it down to four films, but I guess the four most influential in my early passion for film would be, chronologically: FLASH GORDON, BLADE RUNNER, THE LAST STARFIGHTER and TERMINATOR 2. After that there was no going back.

GH: JURASSIC PARK made me realize I wanted to be a computer animator way more than an architect. THE FIFTH ELEMENT, IRON GIANT, FORREST GUMP, THE LORD OF THE RINGS, the list goes on.

A big thanks for your time.

// WANT TO KNOW MORE?

Method Studios: Official website of Method Studios.

© Vincent Frei – The Art of VFX – 2012

MISSION IMPOSSIBLE – GHOST PROTOCOL: Kyle Cooper & Ahmet Ahmet – Creative Directors – Prologue Films

Kyle Cooper has designed and directed over 150 film title sequences, the most known is SEVEN. Among his recent work, we find THE WALKING DEAD, FINAL DESTINATION 5 or X-MEN FIRST CLASS. He is the founder of the studios Imaginary Forces and Prologue Films.

Ahmet Ahmet worked over 10 years in the creation of film title sequences. He handled the title of movies such as THE ONE, SPIDER-MAN, IMAGINE THAT or THE TWILIGHT SAGA: ECLIPSE.

How did Prologue get involved on this show?
Prologue was invited to submit ideas for the proposed opening sequence. Three other companies were also invited to tender ideas, as it was a competitive pitch. However, Kyle had worked with editor Paul Hirsch on the first MISSION IMPOSSIBLE main title and the direction was similar to that film. Brad Bird liked our work so we felt like we had a good chance.

How was the collaboration with director Brad Bird?
The collaboration was a very close one. Brad had already formulated the idea to pay homage to the original TV series’ title sequence by using content from the show and investigating interesting ways to integrate the burning fuse.

What were his expectations for the main title?
Brad’s expectation was to create a sequence that paid homage not only to the TV show’s original sequence, but also to the art of the full main title. In this regard, we chose live action elements that expanded on the action of the movie and showed the action from different points of view, using B-roll footage, effects and shots that were specific to the opening sequence. The idea and the challenge was to integrate the fuse into every scene.

Brad also had the idea of scripting the opening moment of the title. The character Ethan, played by Tom Cruise, delivers the opening line: “Light the fuse.”

Can you tell us the indications and references he gave to you?
Brad had eight scenes that he had composed specifically with the main title in mind. He used an IMAX camera and a dolly move through these scenes. The title required many more set-ups, so we expanded upon the idea by creating more environments that could take us on this fantastic journey.

Did you use some previs or animatic to block the animation?
In our initial presentation, we showed the entire sequence in previs. We wanted to show the complete journey in order to do justice to the concept. This obviously evolved with different editorial cuts.

Can you tell us in detail the creation of the different environments?
The first challenge was to recreate the tunnel environments at the beginning of the sequence, and seamlessly transition from the live action into our CG. Using Maya, Cinema 4D and a combination of set extension techniques, we rebuilt many of the scenes in order to accommodate the choreography of the fuse and typography. The whole sequence involved a very tight combination of effects and live action.

How did you create the flames going through the environments?
Our effects team created the sparks and flames. These elements were generated entirely in CG.

How did you manage so many elements?
The number of elements was not the only challenge – our schedule was short to deliver such a CG intense sequence. Organizing the elements to accommodate render times was a key component.

Which software did you used to achieve this main title?
We used a combination of Maya, Cinema 4D and Houdini. For compositing, we used After Effects, Nuke, Flame and Inferno.

What was the biggest challenge on this project and how did you achieve it?
Schedule. Time was of the essence to get the best we could possibly achieve.

The editorial process became critical, since so much of our CG and choreography was totally dependent on the scene order and selection. To change one aspect of a scene meant that all effects, CG, texturing, lighting and modeling had to change as well. It was like a domino effect. We basically worked every hour available to make the deadline.

What do you keep from this experience?
All challenges are rewarding if you can rise to them. Teamwork and some very talented people made this possible.

How long have you worked on this title?
A total of five weeks from our initial presentation.

What was the size of your team?
There were lots of animators and compositors involved, but the core team was Ahmet Ahmet, Kyle Cooper, Adam Swaab, Jose Ortiz, Andrew Honacker, Nate Park, Ian Dawson and Ty Van Huisen.

What is your next project?
That’s confidential at this time, although we just finished SHERLOCK HOLMES: A GAME OF SHADOWS and numerous VFX shots for TWILIGHT SAGA: BREAKING DAWN – PART 1.

What are the four movies that gave you the passion for cinema?
BUTCH CASSIDY AND THE SUNDANCE KID, THREE DAYS OF THE CONDOR, THE DEAD ZONE, CHARIOTS OF FIRE, PLATOON and HENRY V.

A big thanks for your time.

// WANT TO KNOW MORE?

Prologue: Official website of Prologue Films.

© Vincent Frei – The Art of VFX – 2012