Eight years ago, a reason why Andy Serkis was denied an Oscar nomination was a lack of recognition for how the computer, the actor, and the animator created Gollum’s performance in “The Lord of the Rings”. Last month, James Cameron brought up the same issue when “Avatar” actors Sam Worthington and Zoe Saldana were left out of Oscar consideration. The issue about nominating motion captured actors has persisted because despite advances in mocap technology, the use of animators has always remained. Numerous video and print interviews have featured James Cameron and Jon Landau, (Avatar’s director and producer respectively) insisting that the motion capture performance was derived from the actors with little work from the animators. Such high profile public declaration will not help resolve this question as both detractors and promoters of mocap often inaccurately assess the process of mocap cleanup.
Most movie pundits have rejected, supported, or hesitated over crediting mocap actors. The Hollywood Reporter’s article, “Oscars Snub ‘Avatar’s’ Motion-Capture Actors,” featured quotes by acting critic James Lipton, film professor Richard Brown, and editor Peter Ranier that perfectly echo these three views. Lipton, interviewer of acclaimed actors, illustrates the distrust that acting would be successfully performed through the computer.
James Lipton, host of Bravo’s “Inside the Actor’s Studio”, was not sure that motion capture equated to acting:
“What an actor is doing when acting is not just looking like something but expressing something going on inside. I’m not sure that motion capture, while it captures the flicker of an eyebrow, the twist of a mouth, a gesture of a hand, equally captures emotion.”
Richard Brown felt no distinction between mocap actors and live action actors:
“This is very much the first film of the 21st century. What we need to do is expand our concept of what the word actor means. It’s unfair to take performances as good as these and not designate them as actors.”
Wearing motion capture suits has never been analogous to wearing a costume as the “costume” has been controlled independently by people other than the actor. Yet calling motion capture a form of puppeteering has not been accurate either. Peter Ranier, critic of the Christian Science Monitor, felt this fact should be acknowledged:
“On one hand, it is a performance, but on the other, it is so aided by technology. If I were ‘King of the World,’ I would create a separate category.”
Of the three experts Hollywood Reporter quotes, Ranier has the most accurate assessment of motion capture. The great appeal of motion capture is having actors portraying characters, physically unlike themselves, interacting in environments unlike the audiences’. However, the less a digital character resembles the actor, the more work must be done to ensure the digital character’s movements appear natural. This is accomplished by creating rigs, a “skeleton” that dictates the range of 3D character movement, and instigating clean-up, a process that can involve editing mocap data and adding key frame animation to replace unacceptable mocap data or add movement not recorded from an actor because it was physically impossible to perform.
Since the taller and thinner Na’avi also possess tails and pointed ears it is not surprising that Weta Digital employs both steps. The show “Hollywood Dailies” featured an interview with Richard Baneham, Animation Supervisor for “Avatar,” with scenes of a mocap session, including Weta filming the actors as reference for animation staff. It would be easy to assume that HD filming was part of the motion capturing as the camers point at the actors since no other camera was shown. Actually Baneham cited the filming as a need “ to create a data set for the animators…” in order to “…to take Zoe’s, Sam’s and Sigourney and attain the specific idiosyncrancies of the performance.” Even if this fact was missed, the flying sequences shown in the video were only possible with animators.
Although the actors performed with props, animators work would have to make the Mountain Banshee fly and to ensure the Na’avi mounted, rode, and dismounted convincingly. The crash against the cliff seen in both mocap session and film, also showed the great care the animators took to adjust Worthington’s performance to the new animation without deviating from the actual session. It could be argued that Worthington’s performance was not edited, but if so, then Weta would not have needed to film the actors for animation reference. Watch the aforementioned clip below:
A Look at the Animation Techniques of “Avatar”
Mainstream journals and even entertainment articles have not always explained the role of animation for mocap cleanup as it has become easier to see digital characters register the mocap session. Two years ago Robert Zemeckis had a camera to view the 3D stage and check the actors’ work in his characters after each recording for “Beowulf”. Cameron now has a 3D camera that he pioneered to immediately see what the characters look like in real time from the mocap data. In numerous interviews Cameron has mentioned this new camera as instrumental for immediately checking how the digital characters register the actors during the mocap sessions. Unfortunately, relying on the director or the producer to explain the role of the animators and mocap actors has allowed for a great deal of hype and very little understanding of the entire mocap process.
Upon Worthington and Saldana’s failed Oscar nomination, The Hollywood Reporter quoted Landau’s frustration:
“We made a commitment to our actors that what they would see up on the screen were their performances, not somebody else’s interpretation of what their performance might or might not be.”
In a video interview for the Discovery Channel, Cameron talked about how the new facial capture “proved to be the Holy Grail to do CG faces, not the stuff that that we’ve done before, marker based or what we called imaged based.” Unfortunately Cameron immediately followed that statement with the animators’ role in the film as limited:
“We got the best animators in the world to take all this performance from our captures and limited their options to things that were value-added like the ears and the tails. So they took a human performance with no diminishments and added to it. So when people ask me so what percentage did the actor’s performance come through the final character I say 110%. Because you had an increase of the emotionality of what the moment was.”
Motion capture performance has always been derived from the actor. This has been cited by animators and film crews as the reason not to consider mocap the same as key frame animation as the digital character could take action solely from the actor. Yet if Cameron stated no mocap cleanup was needed on the Avatar characters, then the animators’ contribution was not required. If Cameron meant that the new facial mocap system was so revolutionary that it increased the amount of data from the actors, then the animators’ contribution was effectively zero. With either interpretation of Cameron’s comments by touting how the technology captured the actors’ performance, Cameron has disregarded the animators’ contribution.
The Los Angeles Times article, “Avatar’s’ Animated Acting” did mention that “Cameron points out that it took a team of twenty or more animators at the Weta Workshop in New Zealand nine months to fully animate each ‘Avatar’ character.” However, unlike the beginning of this article, Cameron did not detail what the animators did. Until a Weta animator states otherwise, the actual work of actor and animator shall remain debatable: perhaps the nine months were spent on just on tails and ears, but, in light of Baneham’s interview, perhaps not.
“Avatar”: Motion Capture Mirrors Emotions
Ironically the one member of Avatar’s production team who has worked with motion capture the longest has admitted ambivalence to the issue. In a recent interview with Animation World Network, Joe Letteri, senior special effects supervisor for “Avatar”, detailed how the technical animators delivered facial rigs every two weeks to ensure the best mocap sessions possible, but responded to the question “Are people beginning to understand more about the role of the animation and the significance of the performance capture?” with the following:
“In this medium, you’re really looking at the actor or actress to drive the heart of the performance, but then it has to be realized through some other medium. And I’m not sure if some people are saying if a computer does all that and no one touches it, that’s fine, or if someone touches it, it’s not fine, which doesn’t make sense to me. I think people are trying to understand it. And it’s hard to quantify all the things that actors bring to it, but I think they’re worth a lot… So what if an animator had to fix the toes so they touch the ground properly. That’s not really what we’re talking about here.”
Above: This little mocap piggie went to market — wait a minute. Gotta wait for those animators!
The goal for Letteri, like other Weta members from Gollum’s production team, has been to make a credible character with the best tools, actors and animators possible, with both tools and animators drawing heavily from the actor. Unlike Cameron, Leterri has listed the specific roles actor and animator played and has cited scenes where Gollum was completely animated, (scaling down a wall), or what was a combination of mocap and key frame, (hands and face were animated) because of mocap limitations. Yet some proponents still claimed Gollum was fully mocapped while some critics claimed that an actor cannot be nominated if the role was not wholly performed by one person. To be involved in “Avatar” that invoked the same controversies must be passé for Letteri.
It has been easy to analyze poor motion capture and poor cleanup. The often maligned “Polar Express”, whose producers proudly reported it did not have mocap cleanup, presented numerous examples of character limbs lacking weight, feet failing to make proper contact with the floor and eyes failing to convey emotion. This problem was most noticeable with Polar’s children because some adult actors were mocapped and the animators were not allowed to ensure the performances would fit the characters. “Coraline” director Henry Selick, voiced confusion about classifying “Avatar,” in the Los Angeles Times “Avatar” article about the failed Oscar nominations, but he was very adamant about the animators’ role in motion capture.
“Is it animation? Is it a new category? I’m like the academy. I don’t know where it fits. I will tell you this, animators have to work very, very hard with the motion-capture data. After the performance is captured, it’s not just plugged into the computer which spits out big blue people. It’s a hybrid.”
These days no one would claim that performing traditional photography and film are alike although both utilize cameras. Given that film reviewers and animation directors feel motion capture is a cumulative effort, why not create a separate category that recognizes the work done by actor and animator? If “Avatar” has heralded a new era of filmmaking it should not seek the strict classifications of acting by the Academy of Motion Picture Arts and Sciences. It has been rumored that “Avatar” will have two sequels, so Cameron would have two more attempts to enter the Oscars. If the “King of the World” should seek to create a new category, perhaps a dearth of films using mocap will appear and more mocap actors and animators will get deserved recognition.
Written by Contributor – June Chi [Freelance Artist], [2D & 3D Animator], [Past Lecturer and Teaching Assistant of 3D Modeling, Texturing, & Lighting at University of Texas at Dallas], [MFA included study of mocap technology]