If you’re a nerd about production, animation and maybe Mamoru Oshii, this article’s for you.
Ghost in the Shell (1995) is one of my favoritest movies in the whole wide world… however, I’m very iffy about the recently announced 2008 version. Is GitS a film that needs updating? In particular, how does the excessively digital, Innocence-like color recompositing, 3D character animation and flying camera angles serve to enhance the meditative, yuugen-heavy qualities of the film? Won’t it instead prove to be a distraction like in the updated Star Wars classics and (to some extent) the new Evangelion revamp?
Fundamentally it calls into question the purpose of 3D animation. Part of the reason behind all of these revamps is the emergence of this new and powerful animation medium. However, is the medium being used effectively? What exactly is “enhanced” when 3D animation is jammed into otherwise old films?
Star Wars and Ghost in the Shell succeed as memorable works of cinema because they are able to draw from the inherent strengths of their mediums. Much of Star Wars’ charm (in its production) comes from the grimy, material feel of its sets, props, film matte paintings, puppety aliens and robots, along with its characteristic analog effects and film compositing. While they are not realistic, they at least have a believable texture that you could almost reach out and touch. Through that certain “un”-realistic quality, the world of Star Wars comes to life in an immersive, magical way. However, there’s an obvious clash of aesthetics in jamming cartoony CG aliens and modern digital effects into a film that otherwise doesn’t work with those particular qualities.
Likewise, Ghost in the Shell draws heavily from the inherent picture-esque qualities of 2D animation and illustration (in the Oshii/IG style). Many of the visuals in the film are immensely powerful. The sheer force of its meticulously detailed and arranged images are achieved through its eerie yet serene, static presentation, occasionally accented by a slow zoom, pan, or subtle effect animation. Its zen-like presentation is punctuated by sudden outbursts of extreme comic book violence. The two extremes, both of which draw from the inherent qualities of 2D animation, create a powerful contrast. Therefore 3D animation in its “natural state” can only serve to dilute this contrast.
When I say “natural state”, I mean to suggest that all motion picture mediums (including animation) have things that they are naturally “good” at; in other words each medium has a particular feel. When shaping the “feel” of their animations, 2D animators and 3D animators work from completely different starting points.
When a 2D animator lays down his first key frames, the initial result is jumpy, jerky, excessively stylized motion. The animator will then proceed to add more drawings in order to smoothen out the action into a more coherent motion path. Producing an immaculate, silky smooth animation is one sign of a skilled 2D animator (by some cartoon animation standards).
When a 3D animator sets his first key poses, he gets an overly smooth, “floaty” animation. He will then try his damnedest to disrupt the unnaturally smooth motion with more keyframes, meticulously reverse-engineering the impression of force and weight. Making a non-floaty, highly believable animation is the sign of a skilled 3D animator (again, by some standards)
Both 2D and 3D animators push against the natural state of their mediums in order to achieve a certain quality outside of it. Likewise, filmmakers have no trouble capturing the dry aesthetic of reality through a camera lens. However, a high-budget kung-fu action flick has to utilize wire stunts, special effects, creative camerawork, editing, makeup and lighting in order to achieve a larger-than-life comic book style. In other words, filmmakers also have to push against the nature of their medium in order to achieve an aesthetic outside of it.
The Japanese animation industry succeeded in creating an appealing and economical implementation of 2D animation. Anime is often noted for its creative, stylized use of static frames and its similarly stylized jerky motion. Ghost in the Shell is an example of anime’s static frame style taken to cinematic heights. Animators like Hiroyuki Imaishi and Mitsuo Iso (who also animated the first half of the tank battle in GitS) use jerky, dense motion to great effect and typify the ”anime” style. These successes are due to the fact that anime takes advantage of the inherent qualities of the medium, creating maximum impact with minimal resistance.
In contrast, 3D animation is still a very young medium. Until next-gen procedural character animation technologies mature (read: physics engines and other semi-automated tools), “floatiness” will be the baseline of quality that character animators have to push against in order to achieve anything beyond. Productions not in line with this baseline quality, such as the futile attempts at full simulated realism like The Spirits Within and The Polar Express, will only find themselves slipping into the uncanny valley. It’s still very difficult to produce interesting animation that’s “in line” with the natural qualities of 3D animation. Moments of Ratatouille and Advent Children however utilize some of the natural advantages of 3D animation. Ratatouille’s camerawork explores the environment in ways far more dynamic than what any other motion picture medium can achieve, particularly in its various “chase” scenes. Advent Children stretches ‘floatiness’ in some of its battles to a mesmerizing comic book extreme, creating a very videogamey sense of movement, while taking advantage of exaggerated cloth, hair, and particle simulations.
The 3D animations that succeed in achieving non-floaty dynamic realism only do so through meticulous manual labor and/or meticulous motion capture, both of which are expensive or simply time-consuming. Animators for film effects often bemoan the tremendously high standard of quality they must consistently strive for, as their product necessarily must blend in with natural images.
Alex Rutterford, animator of the experimental short Gantz Graf, recalls in an interview how some of the aesthetic inherent in the medium somewhat subverted the months of meticulous manual animation that went into the short:
“Wire magazine said Gantz Graf was ‘a pretty slick affair’, which I find funny because it think the video isn’t at all. It’s really dirty, grungy and fucked up looking, but because it’s computer graphics people instantly say that’s all really slick. Unless you deliberately take an extreme lo-fi organic approach and do it in a non-photo realistic way, by the mere nature of it being computer graphics most people are going to look at it and think ‘That’s pretty slick’. That said, I like an aesthetic that does that.”
2D animation also has an achillies heel– in realistic human “acting” performances. Animation is very good at conveying comic book worlds and stylized motion, but when it comes to a nuanced, believable human drama sans comickey stylization, animators find themselves pushing against a natural barrier. Entire portions of the human brain are dedicated to reading dense and subtle facial cues, placing the standard of quality for realistic dramatic animation very high. Any animator who wants to wants to convey realistic human emotion in a sequence of keys has a colossal engineering task ahead of him. Jin Roh, directed by the near-realist animator Hiroyuki Okiura, serves as a perfect example of 2D animation falling into the uncanny valley during its dramatic scenes (film spoiler). The failures of The Polar Express in other words are not limited to the 3D realm.
The best productions are aware of their medium’s natural strengths and weaknesses, and the extent of resources they can afford to allocate in overcoming those weaknesses through brute production force. However it’s not as if, say, 2D animation is incapable of conveying realistic human drama. It just needs to do so in ways that don’t involve a closeup of a realistically designed character in an intense dramatic scene. Fortunately Japanese visual aesthetics are packed with ways of conveying emotion through the environment and other symbolism (e.g., falling cherry blossoms), so anime aiming for levels of human realism still has a variety of effective ways to achieve its emotional goals. Most anime however just sticks to the comic book route rather than aiming for levels of realism, which works fine too.
Similarly, video games that attempt to create dramatic scenes through meticulous motion capture and realistic 3D models tend to land in the uncanny valley harder than anything. Cartoony, iconic characters and storytelling systems such as those found in games like Final Fantasy 7 and similarly visual novels purposely avoid the valley caused by the current limitations of technology and the inherent weaknesses of animation. Abstract text boxes and iconic representation leave the fine details to the imagination, not unlike literature in essence.
With all that said, what makes Oshii an interesting case study is in how he purposely breaks the “rules” by jamming the styles of different mediums into each other for highly awkward effects. For starters, Peter Chung (Aeon Flux) observed that much of Oshii’s style is drawn from the Kabuki stage. Sometimes it is used to create a sense of detachment, like in Ghost in the Shell, Avalon and Talking Head. Other times it is used for sheer comedic zaniness like in Gosenzosama Banbanzai.
Kabuki aside, Akai Megane does manga comedy in live action. Mini-pato was created with motion-captured illustrated stick puppets in 3D animation. Tachiguishi Retsuden re-animates photos of actors in papery CG dioramas. In Innocence, 80s action hero Batou spits out paragraph after paragraph of literary quotes and references like he’s in a Jean-Luc Godard film.
As usual, Ghost in the Shell is one of the better Oshii flicks to examine. The film blends many odd elements in order to cast a disembodied feel over the film– the Kabuki stage-inspired blocking and heady philosophical monologues; a soundtrack that bends synths, taiko percussion and a faux-archaic choir; futuristic skyscrapers against Chinese-inspired residential slums and markets; a high-tech tank battle in a run-down museum of evolution; a picture-esque visual presentation with videogamey fight scenes; characters communicating without speaking, and subjecting their bodies to extreme violence without feeling pain. All of these disparate elements create a very detached atmosphere — a very deliberate “uncanny valley” in a sense — allowing you to ponder a world where the boundaries of the body and the individual are breaking down. It allows you to feel disembodied and out of place as the lead protagonist increasingly questions the nature of her existence, eventually leaving her cybernetic body for a “higher plane”.
In this way, Oshii is able to break the aesthetic rules in order to deliberately achieve an “uncanny” feel of detachment and disembodiment. Waking Life and A Scanner Darkly similarly blend the aesthetics of different mediums in order to disembody the audience and mesh the realms of dreams and reality. Satoshi Kon achieves the same with his semi-realistic, highly cinematic 2D animation style in rides like Paprika, which wouldn’t be the same in live-action.
Productions need to be aware of the natural aesthetics of their mediums and the limits of their technologies and budgets in order to be successful. Until then, we’re going to see more film/anime-wannabe CG crapfests in both films and games all over the world, along with misguided, unjustified revamps like (probably) Ghost in the Shell 2.0. 3D animation should not be used blindly for its own sake. It’s an exciting medium but, like all other mediums, it has its limits. I’m hoping that some day soon, like the Japanese industry has done with traditional 2D, 3D animation can find a comprehensive aesthetic niche– a “style” that works efficiently, taking advantage of the natural benefits of 3D and elevating it to new heights. At this point, the tools and technology are not yet at the advanced, integrated, automated state they need to be in order to facilitate this revolution.*
*(Full 3D animation is still a messy unintegrated process filled with snags, archaic workflows and lots of meticulous fine tuning and correction. It’s very unnatural and unintuitive in other words compared to animating in 2D.)