Issue link: https://digital.copcomm.com/i/100015
cover story Return���to���Middle-earth W By IAIN BLAIR The Hobbit���s VFX efforts are no small task. ELLINGTON, NZ ��� Writer/director/producer Peter Jackson spent years creating one of the most ambitious and technically impressive epics in cinema, the Oscar-winning Lord of the Rings trilogy, which he then followed up with King Kong (2005) and The Lovely Bones (2009). But maybe it was only a matter of time before Jackson and his team, including senior VFX supervisor Joe Letteri, returned to Middle-earth, this time for The Hobbit: An Unexpected Journey, a prequel ��� and the first of a planned trilogy ��� set 60 years before the Lord of the Rings blockbusters. At press time, they were still deep in post production, and here, in an exclusive interview, Letteri, whose credits include The Adventures of Tin Tin: The Secret of the Unicorn, Man of Steel, Rise of the Planet of the Apes and X-Men: The Last Stand, talks about making the eagerlyawaited 3D film, the workflow and creating the cutting-edge visual effects. that the story���s being cut and Peter���s seeing what gaps he needs to fill in. So the second comes out in a year, and the final one in 2014.��� POST: Visual effects have progressed so much since you worked on Lord of the Rings, which must have really helped with CG characters like Gollum? LETTERI: ���You���re so right, and a lot of what we did back then was on faith and intu- studied and has evolved over the last decade. With Gollum, he had to look the same, but yet we wanted to capture all those changes and nuances. So the audience will see him as the same Gollum, but with all this new detail that makes him even more realistic.��� POST: Peter Jackson decided to shoot this at 48fps. How did that impact the VFX? LETTERI: ���The 48fps is designed to mini- POST: What sort of film can fans expect? JOE LETTERI: ���A very big film and a visual Weta���s Joe Letteri headed up a VFX team of 800, working on 2,000 shots. 14 treat. The Hobbit was just one, fairly slim book, and I know there���s been a lot of talk and discussion about the fact that we���re making three movies out of what���s the smallest of the stories. But that���s because Peter and the others (co-writers/co-producers Fran Walsh and Philippa Boyens) went back to all of Tolkien���s source material, all the appendices he used to flesh out the story after he finished the Lord of the Rings trilogy,��� and that story is getting woven back into The Hobbit. So it���s much bigger in scope than people may anticipate.��� POST: What were the biggest challenges of creating this? LETTERI: ���As usual, it���s all about the characters. Anytime you create characters that have to take on leading roles, it���s always so important to get it exactly right. Gollum returns again, although it���s actually his first appearance in the stories, with the whole riddles and dark sequence. We have Bilbo and Gandalf again, and the Goblin King and the dwarves who are captured by the goblins and taken down into the caverns. So we created the Goblin King, and a number of the goblins were all done digitally. Then there are the orcs they ride and the eagles. It���s a lot going on!��� POST: Is it true you shot all three films back-to-back? LETTERI: ���Yes, we���re still in post on the first one and the other two are shot, but there will be pick-up shoots next year now Post���������December���2012��� Grinnah, a CG goblin: Weta���s main 3D tool is Maya. They call on Nuke for compositing. ition. Gollum���s a really good example. We broke new ground with him, like sub-surface scattering and the amount of effort we put into facial animation and motion capture... just to get it all working together ��� the body performance and facial movements and dialogue and the look of the skin ��� and make it believable. In the 10 years since, we���ve really broken down and studied what each of those elements means. ���So motion capture has become performance capture, because we can do it with Andy Serkis on set now instead of having to go back and recreate his performance. We can now capture with the facial camera headrig his facial performance at the same time as we capture his body, and integrating the two is so crucial to believability as the real-world characters talk and act at the same time. So we���re no longer guessing about the way the muscles work under the skin. ���We���ve studied all that to understand how movement relates to form, and there has been a lot more work on areas like sub-surface scattering and how light behaves in eyes and hair simulation. Every element that goes into making up a CG character has been www.postmagazine.com mize motion artifacts. Normally if you���re looking at things in 2D, it���s not such a big deal, because we���re all used to motion blur in cinema and stills, but in stereo your eye expects to be able to resolve things spatially and that motion blur is hard to take. So the 48fps essentially just shortens up the motion blur so you get sharper detail. ���What that meant for us was really a lot of preproduction tests, which are what we need to do to understand and mimic the actual set that they shot on. In other words, the virtual cameras we have to track and the copies of the virtual sets and everything we have to do to lock in our virtual world to the physical world ��� being able to either extend sets or put characters in them ��� all that requires a lot more detailed work as you���re doing twice as many frames. Some things allow you to get more detail, like animation. ���You can really nuance the animation since you have finer frame increments to work with. Other things are not really effective, like lighting. How you set a light doesn���t really matter. So it was this mix and we had to look at it area by area and department by department and see what changes were needed.���