Cover image: a key frame from Kill la Kill by Mai Yoneyama
This article is part of the History of the Kanada school series
Hiroyuki Imaishi is no doubt the most important artist to rise out of the Kanada school in the last 25 years: the renewal he contributed to trigger with the Neo-Kanada style completely renovated what Kanada-inspired animation would look like in the 21st century. As one of the major figures of studio Gainax and then studio Trigger, he has also managed to create an environment with a peculiar and recognizable aesthetic, that could hopefully foster new generations of Kanada school artists. Finally, Imaishi is also a famous director, one of the major artistic figures of the last two decades. Having already partly covered Imaishi’s work as an animator, it is precisely this last aspect that I’d like to study here: what Imaishi directed.
The question would then be to interrogate if there exists a Kanada style of direction, or how can direction make room and open possibilities for a certain kind of animation – here, Kanada-style animation. I have already asked those questions, although less directly, in my previous artist spotlight on Kazuhiro Ochi. But there are many differences between Ochi and Imaishi. The first was an interesting and important figure, but didn’t go far beyond episode direction and storyboarding, and as unique as his style was, it doesn’t seem to have had much influence. On the other hand, Imaishi thoroughly cultivates his image as auteur not just as series and movie director, but also as studio leader. The most important distinction between the two men, however, is context: Ochi was in Kanada’s closest circles, and contributed to the original expansion of the Kanada style in the early 80’s, whereas Imaishi and Trigger develop an aesthetic of their own, very different from that of the anime world at large, which has largely evolved beyond the Kanada style.
The other problem at hand will therefore be to try and understand how it is now Imaishi’s direction that brings out new potentials of Kanada-style animation. For this, I will mostly focus on two of Imaishi’s works, among his most recent: the 2013 TV series Kill la Kill, and the 2019 movie Promare. I will show that each one explores very different aesthetical options: one that is turned towards the past, and the reproduction of analogical production and effects; and one towards the future and the adoption of digital workflows and visuals.
Tone matters: parody and the Kanada style
Before that, however, it is necessary to remember that animation is not just a formal device that exists on its own. It is first and foremost a storytelling tool, and the storytelling context is what gives the animation most of its meaning in the first place – with the animation in turn contributing in an essential manner to the storytelling in a mutual relationship. In any case, what I mean by this is that the meaning of the animation in Imaishi’s works is largely dictated by the tone and plot of said works.
It is probably no secret that Imaishi’s approach is largely parodic and rarely takes itself seriously. With the partial exception of Gurren Lagann, everything that Imaishi directed has such an over-the-top, irreverent tone, that even when it reaches emotional climaxes, these must always be taken with some distance. Even in the case of Kamina’s death in Gurren Lagann, it is surely not meant to be welcomed with laughter; but how much is it possible to take it at face value when it is so strongly mediated through another work (Ashita no Joe) in what is as much homage as it is parody? What I mean here is that the generally comedic and parodic tone of Imaishi’s storytelling threatens to dictate the meaning of the visuals themselves and to limit their expressive potentialities. In other words, Imaishi’s overreliance on references to Osamu Dezaki and Yoshinori Kanada (among others) strips them of their initial expressive intent to transform them into parodic, comedic devices. The fault doesn’t lie just with Imaishi: Kanada himself, in many of his late works, increasingly used parodic references and was known for his nonsensical and absurd sense of comedy. Moreover, beyond Imaishi, the Neo-Kanada boom of the 2000’s was largely fueled by self-referential, comedy shows aimed at an otaku audience that would quote what would have been perceived as a “classic anime style” – that is, Kanada and Dezaki’s work.
As time went by, however, Imaishi has come to embody this kind of approach. The clearest example of it is probably one of his early works as episode director, episode 3 of Gainax’s Abenobashi Mahô Shotengai. The episode is Imaishi’s parody/homage to classical anime SF, from Space Battleship Yamato and Captain Harlock to Gundam and many, many other famous or lesser-known series. In terms of animation, it’s a constant display of Kanada-style animation, and sometimes an impressive one. But this episode also displays an incredibly crass and unrefined sense of humor, with tons of fanservice and scatological jokes. This, in itself, isn’t problematic, but what must be noted is that both registers are constantly at the same level in the episode, and that the animation style is the exact same. In other words, the omnipresent Kanada-style animation seems like it’s only there for comedy: the talent displayed by the animators doesn’t seem like it’s there as a show of skill, but only as a parody of itself.
It’s of course perfectly alright for Imaishi’s work to lie in what is essentially a satirical register. The question is, however: if the animation tries to be anything but satirical, can it succeed in doing so? My answer would be no, and I think Imaishi is very well aware of it: what he’s looking for is more amazing spectacle and creating a sense of fun than an emotional reaction from the audience. It is a respectable choice, but the overall relationship it creates to classical anime aesthetics (those both of Kanada and Dezaki) is in the end one of distance rather than understanding or renewal: it’s as if these stylistic traits are only good to be parodied and mocked. With this in mind, it might be possible to say that, as much as Imaishi renewed the Kanada style, he might have (willingly or not) closed it up as a highly formulaic kind of animation. However, it must also be said that, within that formulaic context, Imaishi has been experimenting in extremely new and interesting directions. In other words, he might have reduced the expressive potential of the Kanada style, but he considerably expanded its formal and technical possibilities.
Kill la Kill: unexpected crossovers
As I mentioned, Imaishi’s style mostly takes cues from two figures: Kanada and Dezaki. While these two have come to retrospectively embody a sort of “classical anime aesthetic” (and Imaishi probably played a part in that), the two men have never worked together, and direct contacts between the two styles have been rare at best. Considering the reputation each man has – Kanada as a highly individualistic animator who would do anything to stand out, and Dezaki as a visually-driven director who did without elaborate animation – such a state of affairs might not be surprising.
Taking both styles and using them as prominently in a single work is therefore one of Imaishi’s most original ideas. Doing this in both Gurren Lagann and Kill la Kill, he revealed that the parallel between Kanada and Dezaki wasn’t just a retrospective one, made between two important figures of anime history: in Imaishi’s hands, there could be a strong and unexpected complementarity between the two approaches.
In my mind, the series that best exemplifies this is Kill la Kill, not only because it pushes as far as it can its leaning on Kanada and Dezaki, but also because it often tries to reproduce a rough, organic approach to animation that’s the best possible homage to classical animation. Indeed, it manages to replicate the same impression – the pure, unmediated sensation of feeling the brush itself under the movement. The way blood is represented is a perfect showcase of this. No need to say that blood is both narratively and thematically important in Kill la Kill, and that it’s simply everywhere. And when it is projected all over the screen, it is a red, bright presence, made of what seems like wild splashes of paint. There is a sort of raw materiality here, which works especially well in contrast with characters with “cleaner” designs – such as Satsuki, Ragyô or Nui.
Imaishi’s work as animator, in continuity with Kanada’s late style, was itself generally clean. It did make use of rougher linework sometimes, but the lines themselves remained clear and there was very little of all the ornamental little bits of line and drawing that Kanada had used so prominently. The neo-Kanada style in general used a somewhat minimalistic approach, aiming for clarity and modernity. But in works like Kill la Kill, Imaishi has changed this; most of the time, the animation itself retains that simplicity, but it is inserted into a greater whole. It is most notably the art direction which reimplements roughness in the drawings – the blood was an example, but afterimages and brushwork are everywhere to be found in the series.
The use of Dezaki-inspired techniques, chief among them the “harmonies”, can be understood in such a context. They work as both an homage and parody of Dezaki himself, but also take place in the more general movement of Kill la Kill on the formal and technical level: reproduce the feeling of cel-era techniques and use digital and 3DCG techniques to their fullest potential. In the hands of a less-talented team, these two sides would have either not worked at all, or only in contrast with each other. If Kill la Kill is a success, it’s precisely because it goes beyond that: it shows that there’s a complementarity between the two.
Perhaps as a logical follow-up to the simplification initiated by the Neo-Kanada style, the animation in Kill la Kill also perfectly oscillates between dense, flashy action scenes and minimal movement that’s all the more expressive. The two characters that seem to embody that are Mako and Nui: Nui is most times not even animated, but moves across the screen via pull-cels and related techniques. This kind of motion, which might feel cheap in other contexts, makes perfect sense here: it traditionally creates a wacky, absurd atmosphere, but also conveys Nui hybrid, not quite definite and dangerous nature. The same applies to Mako, a character whose animation is always extremely modulated and irregular. In some sequences, Imaishi himself corrected the animation to remove frames and create an even bigger sense of contrast between some very short fluid movements and others that are very stiff, with poses being kept for a good dozen frames. In some fight scenes, such minimalism expresses power, as if the character didn’t even need to be animated much to win a fight. In Mako’s case, this is the exact opposite: Mako is powerless in a fight, but it’s her personality that comes through.
In other words, the contrast between complex and simple animation isn’t just the kind that one might find in any other show, where the animation is naturally distributed according to the intensity of each scene or the talent of each animator: it takes on a narrative and thematic significance. This is of course only possible because the animation staff of Kill la Kill was good enough to pull off impressive sequences whenever needed, pushed by a show that very rarely slows down its pace. But Imaishi’s approach to movement as a director also needs to be acknowledged. It could be argued that he conceives of everything that happens in terms of intensity: the objective is to create a spectacle that’s as strong and overcoming as possible in a sort of constant excess. Although it works out very differently, it is precisely in the same terms that Dezaki’s direction could be described: a work that almost tries to exhaust the viewer by pushing every emotion to its maximum. It is also a way that one can understand Kanada’s approach to animation: making each little sequence a work of art, an intense experience that showcases virtuosity. Relying on both at the same time, Imaishi reveals this similarity and perfectly exploits it; in that, he doesn’t just put together two styles that have only little in common: he uses them as natural complements and creates a vision of his own.
Promare, Kanada, CG and beyond
This drive and constant research for intensity is beautifully expressed in the constant spectacle that is Promare. However, it was produced by completely different techniques, showcasing Imaishi’s inventivity. Whereas Kill la Kill heavily relied on a “retro” aesthetic, Promare is decidedly modern, whether through its simple character designs and minimalistic background art, its vivid color and, more essentially, its prominent use of 3DCG.
Experiments with 3DCG go far back in Imaishi’s career, and are already very visible in Kill la Kill. Besides the formal and aesthetic drive, this was made possible thanks to the close relationship between Trigger and one of the most important 3DCG studios in the anime industry, Sanzigen. Established in 2006 by former Gonzo CG director Hiroaki Matsûra, Sanzigen quickly established itself as a key figure in digital animation over the 2000’s, and its first collaboration with Imaishi was on Gurren Lagann.
Gainax and Trigger’s use of 3DCG to enhance the 2D work dates back a little further. The real stepping stone can be considered to be FLCL, with impressive rotations sequences in 3DCG handled by Kaoru Matsumoto and Yoshimasa Yamazaki, from the CG department of Production IG. Their work was awe-inspiring for the time, and it could be argued that even now, it has rarely been equaled. Maybe spurred on by this successful attempt, Imaishi must have taken interest in the possibilities opened by the technique, which he explored throughout the 2010’s. The most important moment in that evolution is probably his work on the 2012 Black Rock Shooter TV series: supervising Sanzingen’s work on action scenes, Imaishi was credited as “CG special skill director” and storyboarded as well as directed 5 episodes. The use of CG there is very reminiscent of Kill la Kill, as it combines a very mobile camera and complex fighting choreographies with more organic, 2D hatching and light flares.
The reason this works so well is not just because Imaishi adds his own sense for action storyboarding: it has to do with Sanzigen’s own production process. One of the reasons why 3DCG often integrates poorly with anime’s 2D animation is timing: it can be hard to combine heavily modulated 2D motion with the more fluid CG animation. But Sanzigen does its 3DCG on 3’s, that is, what would be considered “limited” rather than “full” animation, thereby following closely the standard of 2D animation. They also implement specific techniques to integrate their work better to anime standard, such as cel-shading texture work, or highly plastic character models which create an expressivity almost similar to that of 2D animation, with the goal of letting every 3DCG artist develop their own expressions and style – that is, to create something similar to the “charisma animator” culture of Japanese 2D animation.
In Promare, Sanzigen’s work only pushed further the possibilities opened by Black Rock Shooter and Kill la Kill. It pursued constant camera movement to create extremely disorientating action scenes that more often than not stun the viewer with their audiovisual excess. But what’s most impressive is the effects work, which was largely left in their hands except for some specific sequences.
It is thanks to their work that Promare introduced what is probably the first ever full-3DCG Kanada dragon. Kanada’s Genma Taisen dragons were created through very detailed and intense work on curved lines and the interaction between colors. Here, there is something similar, although it was created very differently. Indeed, the curves so characteristic of Kanada’s fire effects have been replaced by mostly triangular shapes; but, in the very same way, it is the relationship between the colors that creates the motion. The geometrical shapes, as well as the unique sense of texture created by the 3DCG convey a very different feeling, but the instability of shapes and the constantly morphing aspect of the dragons remain in what is not just a technical, but also an artistic masterwork.
Beyond all this, it must be noted that experimenting with 3DCG cannot just be formal play for a Kanada-style artist such as Imaishi: consciously or not, it places him in a complex relationship with Kanada himself, who had himself been a pioneer of 2D/3D interaction in his later years. However, as I noted in an earlier piece, Kanada’s use of it was very specific: the 2D characters evolved in a plane of their own, made of flat surfaces, whereas 3DCG was used as an opportunity to explore complex camera movement and layouts. In other words, Kanada relied on the contrast between the two techniques, using them in opposition to craft action scenes.
Imaishi’s approach, probably helped by the evolutions of technology, is completely different, if not at odds with Kanada’s. Indeed, what the former is moving towards is not a contrast between 2D and 3D, but a merging of the two that would try to erase their differences. With this, Imaishi has considerably opened the possibilities of the Kanada style, enabling it to eventually translate to 3DCG and conquer new worlds. Although Imaishi’s future works still remain to be seen, he might have made a prodigious step with Promare, realizing and probably going beyond what were Kanada’s aspirations.
Imaishi’s aesthetic, both as an animator and director, is very much a product of its time. With the general look of animation having moved on from the Kanada style, he single-mindedly exploited the most easy path that was still open to it: homage and parody. In that sense, it would be difficult to credit him with a renewal of Kanada-style animation past his pioneering work as animator. But Imaishi has also integrated the principles of modern action animation, most notably overload, and used them to rework the Kanada-style, opening new ways in both the technological and formal aspect. It will probably be difficult for this to have a significant influence on individual animators, considering how much it relies on Imaishi’s own sensibilities and collaborations. But it is in any case an important step, through which Imaishi constantly renews the identity of the Kanada style, between something ancient, bound only to be parodied, and one of the most modern and pioneering approaches to animation.