Academy governors Richard Edlund and Bill Kroyer weigh in on how to acknowledge and reward the growing number of stellar CG character performances.
Producing photorealistic digital humans, animals and other types of creatures has been a one of the holy grails for computer scientists and digital artists since the nascent days of the CGI technology revolution. More recently, in the last 10-15 years, many filmmakers have made significant strides toward achieving the goal -- some more convincingly than others -- of bringing believable photoreal human and lifelike digital characters to the screen.
From the first full-length photorealistic animated film, Hironobu Sakaguchi’s Final Fantasy: The Spirits Within (2001) to Peter Jackson’s ground-breaking Lord of the Rings trilogy (2001-2003), Robert Zemeckis’ innovative but not hugely popular Polar Express (2004) and Beowulf (2007), David Fincher’s mesmerizing The Curious Case of Benjamin Button (2008), various wildly popular Marvel movies starting with Iron Man (2008), and of course, seminal films like James Cameron’s Avatar (2009) and Rupert Wyatt’s Rise of the Planet of the Apes (2011), filmmakers have continually raised the bar, and stakes, with a series of increasingly complex digital characters that don’t just wow audiences with their stunning visuals, but capture people’s hearts with their believably emotional performances. The growing list of such digital performances is a testament to the unending audience appetite for visually engaging storytelling produced by a new generation of talented filmmakers more and more adept at embracing and making use of sophisticated production technology.
Filmmakers have also been creating younger versions of older actors through the wizardry of CG de-aging, or recreating actors entirely who are no longer living. Recent examples, to name just a few, include Jeff Bridges in TRON: Legacy (2010), Michael Douglas in Ant-Man (2015), Robert Downey Jr. in Captain America: Civil War (2016), Carrie Fisher (prior to her passing) in Rogue One: A Star Wars Story (2016), Kurt Russell in Guardian of the Galaxy Vol. 2 (2017) and Sean Young in Blade Runner 2049 (2017) as well as the late Paul Walker in Furious 7 (2015) and Peter Cushing in Rogue One.
Advances in CG technology have provided VFX houses with unprecedented creative tools and computing horsepower, which studios have harnessed more and more effectively in the production of realistic fully digital characters, including photorealistic humans. Which brings us to the inevitable point in time where the entertainment industry finds itself confronting the very definition of an “acting performance” and whether or not digital character performances can, and more importantly should, be judged by the same criteria as human character performances are when awards season comes calling.
This year’s five Oscar nominees for best visual effects each include significant performances by realistic digital characters, both human and non-human, that audiences and critics around the world have found compelling; Andy Serkis gave us both Caesar in War for the Planet of the Apes and Snoke in Star Wars: The Last Jedi; we have Bradley Cooper’s Rocket and Vin Diesel’s Groot alongside a de-aged Kurt Russell as Ego in Guardians of the Galaxy Vol. 2; Blade Runner 2049 includes Ana de Armas’ Joi and Sean Young’s Rachel; and Toby Kebbell was central to the performance of King Kong in Kong: Skull Island.
Determining how to categorize and recognize these digital character performances -- including who, in reality, actually created a given performance -- is a source of considerable industry debate. For two artists steeped in animation and visual effects production as well as motion picture history, this debate goes right to the very heart of how the Academy of Motion Picture Arts and Sciences defines animation and how acting performances are recognized.
Richard Edlund is a four-time Oscar winner for best visual effects who has also won three Academy Scientific and Engineering Awards, an Emmy, and two BAFTA Awards. He is the Pankey Distinguished Artist/Filmmaker-in-Residence at Chapman University in Orange County, California. Edlund has also served as a governor of the Visual Effects Branch of the Motion Picture Academy for twelve years, chaired their Visual Effects Branch since its inception, and, for eight years, chaired the Academy’s Scientific and Technical Awards Committee. Oscar-nominee Bill Kroyer is the director of the Digital Arts Program at Chapman’s Lawrence and Kristina Dodge College of Film and Media Arts. He is also a governor of the Short Films and Feature Animation Branch of the Motion Picture Academy and serves as co-chair of the Academy’s Science & Technology Council.
Kroyer has seen this issue brewing for years. “I saw this coming in 2001, the year we established the Animated Feature award,” he notes. “When we sat down to write the rules for what an animated feature was, at that time I said, ‘In the definition of an animated feature, we can’t make any mention of how it looks.’ I could see right away, the day will come when animation will look so photoreal you can’t tell it from live-action. So, what makes an animated film? If the performances are created by animation techniques, we decided, that’s an animated film. That’s still true. So today, a film like Avatar could, technically, qualify. You could make a movie that would have a lot of Benjamin Buttons in it, and you could consider that an animated film. It all depends if the performance itself is done with frame-by-frame techniques. Photorealism, we could see that coming from miles away. You could see that coming two decades ago.”
According to Edlund, when you consider digital characters such as Caesar, you must also consider the team responsible for animating the performance. “Well, [the Academy is] studying this whole issue,” he says. “I know Andy Serkis thinks that he was the first one to use motion-capture and all that, and he’s a talented guy. But the thing is, Andy Serkis’s performances are also tweaked by animators, and so rather than giving Andy Serkis the entire award…if he were to be nominated and voted in, you’d have to split the award with the animators. So, it’s a very difficult thing. There’s always a serendipity about performance. There are things not on the page of the script that happen within the performance of the actor. When you animate, everything is intellectual, everything is being created. So, this is the valley that animators have to cross.”
Kroyer also voices concern for the consideration of who really “created” the digital performance. He explains, “The Screen Actors Guild, for its entire history, has recognized, has equated performance, with individual actors. Now you’ve got a situation like Planet of the Apes, where the performance is partially done by an individual actor and completed by an entire team of people. So, at what point do you start rewarding the performance? That performance [Caesar] has as much or more impact on an audience [as a human performance], regardless of the fact of how it was created.”
More specifically, Kroyer goes on to describe how regardless of whether or not the acting is digital, that performance should be worthy of awards consideration. “Well, for one thing, there is nothing good or bad about technology in itself. It’s not inherently evil. It’s what you do with it. There’s nothing bad about motion-capture or reproducing completely photoreal people. As far as the Academy goes, it’s something we’re looking at very seriously. Think of what the Academy honors -- it honors entertainment! It honors skillful entertainment that has an impact on an audience! So, if you have a situation where a fabricated character is having as much influence on an audience as a real character, why shouldn’t that performance be recognized? And that’s the big question we face.”
As far as recent attempts at producing photoreal human characters, Edlund, a pioneer in visual effects going back to the early Star Wars days, minces no words. He notes, “I was discussing Rachel and the Uncanny Valley [the idea where digital humans that are almost, but not quite, photorealistic elicit uncanny, or weird, often uncomfortable feelings in viewers] with John Nelson [overall VFX supervisor of Blade Runner 2049] at the VFX Bake-off. We were talking about the difficulty of actually making a digital human realistic and believable. He said, ‘Well, Benjamin Button kind of beat us there.’ In a certain way, it did. They used numerous techniques in Benjamin Button. He was playing the same character, but in various times in his life. For John, it was very time consuming to do the work that they did on Rachel, and on top of that, her performance that they copied was pretty uneventful…it was basically just a few shots, and she didn’t have to act much. I’ll go ahead and mention the Rogue One characters, which I didn’t think worked, either one…certainly, Leia [a de-aged Carrie Fisher] was a flop in my opinion. Grand Moff Tarkin [Peter Cushing recreated digitally] was pretty good. There was a real dark scene and he’s not the lead character in the movie…a lot of people don’t remember him well. That was kind of alright, but Leia was bad.”
Kroyer points out that for many digital characters, the very essence of the performance came from a real actor whose efforts were then transformed by a team of animators. He explains, “If you’ve ever seen the making of Peter Jackson’s King Kong, talking about Andy Serkis’s role, you have to realize that in that movie, there wasn’t a single frame of mocap that was ever used. Despite all of Andy’s work. But, I was personally impressed with what Andy Serkis brought to that performance. He brought a unity, a vision and a sensitivity that without a doubt was the foundation of the performance, even though, technically, his movements weren’t actually captured.”
He continues, “On the other hand, when what the audience is seeing on screen is not what the human actor portrayed, where an animator had to go in and make a decision, that poses another legitimate question about who should get the credit for what you’re seeing. As you know, there are many instances where a twitch of an eye or a certain look can end up being the thing in the scene that really has an impact. So, what if it was the CG animator that made that decision?”
Edlund also raises the issue that the CG revolution has fundamentally changed all filmmaking, making it harder and harder for everyone to fully comprehend the extent of a film’s digital elements. He notes, “It was different in the analog era because it was a little bit easier to see what was what. But, it’s gotten to the point now where the work is so transparent that even people in the business, the visual effects practitioners, can’t figure out how it was done, what was done, what’s real and what isn’t.”
Looking back to his days on 1977’s Star Wars: Episode IV - A New Hope, Edlund remarks, “I remember when I did Star Wars, there were maybe 80 people that worked on the film over a period of two years. With a budget of two and a half million dollars. This was 40 years ago, but at the same time, now the visual effects are taking over the industry.”
Regarding how the Academy membership judges VFX-driven films, including the increasing appearance of digital characters, Edlund shares that the issues, which can seem overwhelming, have gotten the Academy’s attention. “The visual effects branch executive committee and the nominating committee wrestle with that every year. And the thing is that now we have these dueling tent poles where the effects budgets are in the tens of millions, and sometimes even more than a hundred million dollars, and the film credits show five across quintuple rows of computer animators rolling by at high speed for a couple of minutes, with a couple of thousand names. It’s worrisome to me in a way because if you like dramas that are made for $20 to $30 million, they are up against a movie that costs $150 to $200 million, and those are the ones that bring in the billions of dollars. The industry has changed dramatically. It’s really mind-boggling if you think about it.”
For Kroyer, the key to any change in the Academy’s consideration of digital characters and new, technology-driven film performances, is patient and informed dialogue. He explains, “Where things are actually heading right now is examination and discussion. I mean, this [area of discussion] is absolutely on the table. Very few people know all sides of these issues. So, what we’re trying to do at the Academy is educate all the pertinent branches on each other’s feelings and disciplines and have this discussion. Because absolutely…the time has come to do it.”
While Academy membership skews younger each year, there are still many older members who are less familiar with recent technological advances in filmmaking that to many, including Edlund, seem to dominate contemporary movie making. For Kroyer, educating these less tech-savvy members is part of the Academy’s examination process. “Admittedly, a lot of the Academy members are older. This new technology stuff is all really unfamiliar to them, But, what we found in other areas, is once you educate people and get them familiar with what’s going on, they’re much more likely to make decisions that reflect the new information. So right now, I see a ‘wait and see, convince me’ kind of attitude. As the Academy’s membership gets younger and more educated [on the technology of filmmaking], I think they’re much more open to this and they don’t out of hand dismiss things.”
As far as if we’ll see a new Oscar awarded to best performance by a digital character anytime soon, Kroyer cautions against jumping the gun. “The time has come to try to see if there’s some legitimacy to moving into different definitions here, maybe even different awards,” he says. “We’re much more conservative than the VES [Visual Effects Society] or ASIFA Hollywood [the organization behind the Annie Awards] in creating new categories. Our thing is to look at what [categories] we have, see how the new technology is affecting them and see if there’s something to be done. We don’t just leap in and make a change overnight. It’s not typically our way. We want to make sure the next step we take is the right step. Right now, though, I can tell you that we’re very aware, and very active, in examining these questions.”
Dan Sarto is Publisher and Editor-in-Chief of Animation World Network.