Advertisement
Current Issue
September 2014
Issue: January 1, 2012

Cover Story: 'The Adventures of Tintin'

By: Daniel Restuccio
MANHATTAN BEACH, CA — The Adventures of Tintin is director Steven Spielberg’s first performance-captured animated feature, and it marks his first collaboration with noted producer/director Peter Jackson. 

“One of the main attractions for me on this entire endeavor,” said Spielberg in quote from the film’s press kit, “was to collaborate with Peter (Jackson). I have tremendous respect for Peter, his movies, his storytelling abilities, his big imagination. A lot of the pleasure of working on Tintin was getting to work with him constantly.”

The character of Tintin was created in 1929 by Belgian Georges Prosper Remi, a.k.a. Herge, who gave up a career as a foreign correspondent to become the writer/illustrator of the adventures of the globetrotting boy reporter/detective. Over the years Herge’s creation became a cultural icon through books, magazines and even a 39-episode animated series. The Tintin franchise has delighted millions worldwide.

Spielberg first became aware of the stories over 30 years ago when a French newspaper reviewer compared the character Indiana Jones to the adventurous Tintin. Spielberg inevitably became a fan of the books and even read them to his children. While he never met Herge, the director did speak to him by phone. Spielberg finally acquired the rights in 1983 from Herge’s widow Fanny Remi. 

Fast forward to 2002. Spielberg renews his option and decides to make Tintin into a live-action feature. “The books that I received in my office,” Spielberg recalled, “didn’t have the translation, but I understood the story. Every frame every single panel told a story in cinematic terms, including color palettes, composition of figures and action… very expressive action. The way that Herge would pose his characters almost as if he was trying to squeeze out 24 frames in a single frame, and succeeding. It was a movie. And that was I think the genius of Herge.”

HOW IT BEGAN

Early 2004, Weta visual effects supervisor Joe Letteri was in Los Angeles when Tintin producer Kathleen Kennedy called him about making a CG version of Tintin’s dog Snowy. Letteri and Weta co-founder Peter Jackson, already a big fan of the property, did some tests dressing an actor like Tintin and adding a digital Snowy. Jackson even added a clip of himself auditioning for Captain Haddock,  interacting with the CG dog, who comically falls off a dock as Jackson dives in after him. 



The question then became, “‘Who’s going to play these parts,’”explains Letteri. “Who’s Tintin? Who are the twins? Who’s Captain Haddock? You start to go down that line... do you put the actors in prosthetics or do you just ignore the fact that they don’t look like the characters? Is that going to feel enough like the world of Herge? As all this was happening we were doing King Kong and we were starting to do more and more of these big virtual worlds. Then we’re done with King Kong and starting up Avatar with Jim Cameron, and we’re doing even more virtual worlds.”

“Steven and I just started talking about Tintin and the characters and about the idea that if it wasn’t live action would it be possible to take Herge’s drawings (and make a movie) using complete CGI animation,” explained Jackson in press materials. 

In 2007 Jim Cameron was prepping Avatar on his performance capture stage in Playa Vista. Cameron let the Tintin team use the stage for a couple of days to create a proof-of-concept. Spielberg and Jackson performed  a test with actor Andy Serkis and some extras. 
Weta took that footage back to New Zealand and worked on things like look development and art direction. Jackson said to his concept development team, “What would the world of Tintin look like if it were real?”

CAPTURING PERFORMANCES

“The allure of performance capture,” says Matt Madden, VP of Manhattan Beach, CA’s Giant Studios (www.giantstudios.com), “is that it provided a director/actor experience similar to live action, but with the creative freedom of an animated feature. 

“Unlike traditional live action, they didn’t have the typical down time for things like lighting adjustments and set changes, because ‘previs-level’ virtual environments had already been built by Weta. There was some physical stage building required to facilitate the performance, but those set-ups took only a fraction of the time of a practical set build.”



In early 2009 after Avatar had completed their performance capture work, the Tintin filmmakers converged on the Giant stage. 

The 36x65-foot space holds over 100 motion capture cameras capable of discerning even the most subtle body gestures. “The natural shape of a volume is lot like a circus tent, which peaks down the middle,” describes Madden. “The set builders would take advantage of this, aligning the sets accordingly to get the maximum height for the shots that needed climbs or falls.”

The actors were prepped in custom suits with 50 markers and upgraded head rigs by Glenn Derry and Video Hawks to capture facial motion. “Camera resolution is very important,” says Madden, “but the real issue is the effective resolution, which is ratio of resolution to visible area covered by each camera. “You certainly don’t want any marker to move within a pixel and have it not be detected, so we make sure that camera only sees a certain amount of space to prevent any motion detail from being lost.”

One key system upgrade from the Avatar shoot was in the realtime performance playback. Part of it was leveraging faster processors, part of it was more advancement in Giant’s software to allow realtime capture of up to seven characters. 

Madden emphasizes that for performance capture it’s not just about the capturing marker data. “It’s the solving and retargeting software that creates the character, the editing software and artists who clean up and manipulate the motion without compromising the integrity of the original performance. It’s building an accessible database, integration of virtual camera and editorial; so there’s lots of pieces to the process.” 

With Avatar, Cameron would capture multiple performances and meticulously edit together the best performances into a scene build. Then he would enter that virtual world and do his camera work. For Tintin the workflow was adjusted to Spielberg’s style of live-action shooting. 

Giant was able to target the motion capture data to rigged character models in the CG environments in realtime. The models, built by Weta (www.wetafx.co.nz), are low polygon, with no facial articulation. As the actors perform on stage the motion data is retargeted in Autodesk MotionBuilder and synced to character models that move in virtual environments. 

Spielberg and his operator shot “virtual camera” coverage similar to how he would on a traditional film. Only in this scenario he had three viewpoints: wide and close-ups from reference cameras and, through his hand-held viewfinder, the moving character in the virtual environment. Every take was also slated and recorded into video assist just like any live-action movie. 

“I had like a game controller in my hands,” described Spielberg, “with a TV screen with buttons I could crane up and down just with my thumb. Another button I could dolly right and left with my other thumb.”

Once a take was approved, Spielberg would often explore different angles with the virtual camera while directing the actors. Spielberg and Jackson could then evaluate the performance and circle the approved takes based on a combination of the CG footage, the synchronized close-ups on the face and the actor’s physical performance. 

Once a scene was completed there was no shipping film off to the lab or even transcoding. Selects were made and the motion capture data was immediately processed by Giant’s internal software and sent to MotionBuilder for the New Zealand-based Weta team to QC and render. The CG render was then directly sent down the hall to Tintin editor Michael Kahn’s Avid editing workstation.

“Once (Spielberg) was finished coming off the stage,” says Letteri, “he had everything he needed to cut. Now you’re just traditionally making your selects and putting them in the film. So Michael Kahn actually had everything he needed to cut with.” 

Giant and Weta created an expedited workflow to convert every shot as quickly as possible into a game-level representation of the scene... enough for Spielberg to evaluate the timing and blocking of the action, as well as shot composition. “This cut provided the foundation for Weta to build upon with their magnificent animation, textures, lighting and effects,” describes Madden.



The main hero comedic and dramatic beats were done in Los Angeles. The more complex actions shots, crowds, like second unit production, were primarily done in New Zealand on Weta’s Giant-outfitted performance capture stage.
 
Early on for the 3D version of Tintin, Letteri took a couple of scenes from the movie and presented them to Spielberg with different styles of convergence. “You could stay focused on convergence, which is what we did for most of Avatar, we showed him parallel convergence, which is IMAX-style, a couple of other styles, just so he could wrap his head around what the 3D space really is.” Working from that the animators got a feel for what Spielberg wanted. “Once we got renders we’d get stereo versions going on early on as well,  just in case there were some final changes.”

Once the performance capture edit was locked, Weta worked for over two years finishing the movie. At the peak of production during the summer of 2011, over 900 people were working on the film. “Usually at the end of a film it’s all hands on deck,” says Letteri.

Park Road Post in Wellington completed final color correction, sound re-recording and final sound deliverables.

“I never considered, in the 1980s, to do the film animated,” said Spielberg, “but as technology began to evolve I realized that there was a medium that was going to be the perfect tool and the perfect simile between Herge’s art and what the art of Tintin could be as a motion picture.”