Directing Kanada

Cover image: a layout from Nausicaä of the Valley of the Wind, by Yoshinori Kanada

This article is part of the History of the Kanada School series

One of the most notable aspects of Kanada’s career is that, while he never directed anything by himself, he was closely associated with major directors: first Yoshiyuki Tomino, and then Rintarô and Hayao Miyazaki. His relationship with the latter two is what I’m going to research here. More precisely, I’d like to see how animator and directors worked together and reciprocally pushed each other in new directions. The goal will be to explore Kanada’s animation in detail, to investigate and try to uncover what was his, what were his innovations, and what must be credited to other people: directors, animation directors, and other animators.

The period I’m going to study is quite long and extremely dense: it goes from 1979, with Galaxy Express 999 to 1992, with Porco Rosso and Download: Namu Amida Butsu wa Ai no Uta. In these 13 years, Kanada went through major stylistic shifts, and produced some of his most important works even aside from his collaborations with Rintarô and Miyazaki. Therefore, this piece will not attempt to give a full account of Kanada’s career during the 1980’s. I will only focus on 6 works, 3 by director: for Rintarô, the two Galaxy Express movies in 1979 and 1981, and Download in 1992; for Miyazaki, Nausicaä of the Valley of the Wind in 1984, Laputa Castle in the Sky in 1986, and Porco Rosso in 1992.

Galaxy Express: Rintarô and Komatsubara

Let’s begin by putting Kanada’s collaboration with Rintarô back into context. Galaxy Express 999 was the first time they worked together, and it was an important step in both men’s careers. For Rintarô, it marked the peak of his time in Tôei Animation and the beginning of his career as a movie director. For Kanada, the year 1979 was complex and difficult to read. It was mostly marked by the change between studios: it was when he left the place he had created, studio Z3, to rejoin his mentor Takuo Noda in studio N°1.

The exact chronology is hard to follow. Kanada’s most infamous work for the year was on episode 1 of Mobile Suit Gundam, something quite normal since Z3 had become a regular subcontractor for Sunrise. However, Kanada’s contribution to the episode was notably small, and he never worked on the other episodes that Z3 did on Gundam. According to his own testimony, although he was attracted to the world and characters, Kanada felt that Yoshikazu Yasuhiko’s control over him was a bit too heavy: indeed, the animation director did by himself all the layouts and some key animation for the opening episode. It’s possible that there had already been some disagreements between the two men on Zambot 3 two years earlier and, even though it’s hard to tell, Gundam was probably the point when they realized they weren’t made to work together.

It’s around the same time that Kanada received the offer to work on Galaxy Express 999, and this apparently motivated him even more to leave Gundam’s production. Kanada and Rintarô didn’t know each other yet, and my guess is that Kanada was invited by the movie’s character designer and animation director, Kazuo Komatsubara.

Alongside his friend and comrade Shingo Araki (Kanada’s first teacher), Komatsubara is probably one of the most important figures of 70’s animation, and most notably one of the pillars of Tôei Animation’s TV shows during the decade. He was the co-founder of one of the most important subcontracting studios in the history of anime, Oh! Production, and started animating on Tôei’s TV series around 1965. He quickly rose to prominence in 1969, on Tiger Mask, for which he was animation director on 17 episodes, and did his first character designs on Genshi Shônen Ryû in 1971. From then on, he had many opportunities to meet Kanada, as he was animation director and/or character designer on the Tôei series the latter did his first key animation on: Cutie Honey, Getter Robo, and Getter Robo G. More generally, they were in the exact same circles throughout the 1970’s. As for Rintarô, Komatsubara met him as the character designer and animation director of Captain Harlock, that the former directed.

Komatsubara stood out at the time for his unique approach to animation direction. Since the number of animators per episode on TV series was very low (between 2 and 6), the animation director always had a very important role and it was often their style that prevailed the most. It is this context which allowed for the emergence of very controlling animation directors in the 70’s, like Yoshikazu Yasuhiko and Hayao Miyazaki. But Komatsubara seems to have adopted the opposite philosophy: it would be him adapting his own style to that of the animators rather than the opposite. It has therefore been said that there is no “Komatsubara style”, while animator Atsuko Ishida has said of him that he is representative of an approach centered around “attraction”: his greatest skill was to attract talented animators and to offer them the best opportunities to express their respective styles rather than impose his own.

This is certainly visible in both Galaxy Express movies, and especially in the second one, which I’ll cover below. On the first movie, there were 15 key animators. Many of them were Tôei veterans, but there were also charismatic figures that Komatsubara had worked with previously: besides Kanada, there was Kazuhide Tomonaga, who had joined Oh Pro a few years prior (Galaxy Express 999 would be his last work in the studio). The other major names were the three members of Studio Bird, Yoshinori Kanemori, Hiroshi Oikawa, and Yoshinobu Inano. All three had come from Tatsunoko and were experts in character animation – they were probably the ones behind some of the most delightful moments of acting in the movie. Inano, most notably, was in charge of the entire opening and ending scenes of the movie.

As the low number of key animators might indicate, each artist was given a large chunk of the movie, as would be a constant in many Rintarô films. These are but rough estimates, but I believe that Kanada animated between 15 and 20 minutes, Tomonaga around 15, and Inano 7. It is with this in mind that we must approach the animation of the movie: it was a lot of work, especially for Kanada who handled very complex scenes, most notably the apocalyptic finale, one of his most famous and important works.

To put it into more general terms, Rintarô seems to have been mostly interested in choreography and movement; however, motion, that is the way objects move rather than the directions they move in, was entirely left to the animators. This is visible when we look at two scenes animated by Kanada, taken from the collapse of Planet Maetel. 

As for the rhythm, the timing is key for giving the fall a feeling both of surprise and fluidity. We begin mostly on 2s, as Maetel suddenly loses her footing. But then as she really falls down, her head turns from left to right; this is a sudden move, conveyed by a shift to 3s and a much wider spacing between the two frames. After that, the timing gets much more erratic: it oscillates between 1s, 2s, 3s and 4s without any regularity, to give the viewer time to take in each pose but also to keep it dynamic.

The second moment I want to analyze is Kanada’s most famous on the movie, the final destruction of the planet. What’s very striking is that the storyboard contains very little detail, even though this is such an important moment. There are a lot of annotations, but Kanada’s animation is so baroque and inventive here that he was probably improvisating and animating on his own ideas. Since the storyboard was most probably made before each animator was given (or chose) his cuts, it’s not like Rintarô specifically left everything in the hands of Kanada; it’s more probable that he simply wasn’t interested in the detail of the effects. Though it must also be noted that everything isn’t in the storyboard: the key animation was checked by both the director and animation director, and they might have given suggestions.

As things stand, however, I consider that this scene can be considered as one of the most representative instances of Kanada’s creative power. In the first step of the destruction, he added red lines to the core of the planet, adding a great amount of texture. Then, in the second shot, there is obviously a lot of care put into the color, as we shift from black and white to purple, then a sudden explosion with yellows and reds, and then an abstract composition made of straight red lines and a white circle. The work on lighting and photography is very important here. Some of it is already in the storyboards (such as the white light and the use of glass to make reflections), but it would certainly be interesting to have access to the original drawings to see how much of the coloring decisions were taken by Kanada himself.

Galaxy Express 999 is no doubt one of Kanada’s most important and influential works. It’s probably what made him known beyond the circle of animation maniacs and super-robot fans: it was the highest-grossing film in the Japanese box office for the year 1979, and one of the major successes of animation in Japanese theaters. In the context of the history of the Kanada school, it is the style Kanada used in this movie that would be the most associated with him and copied by his first students: the peculiar body shapes and “liquid-fire” effects of Masahito Yamashita, Kazuhiro Ochi and Kôji Itô’s early work all come from here.

All these elements would be even more pronounced two years l ater, when Tôei tried to bank on the success of the first movie by producing a sequel, Farewell Galaxy Express 999. The top team was the same, with Rintarô reluctantly being brought back to direct and Komatsubara as animation director. However, the animation team was completely different. Many of the most important animators of the first movie, like Kazuhide Tomonaga or Yoshinobu Inano, had left for other projects: Tomonaga was in Telecom, working with Hayao Miyazaki and Yasuo Otsuka, while Inano had become a Sunrise subcontractor and a central animator on Space Runaway Ideon.

To replace them, Komatsubara made full use of his contacts: most of the animators on the movie were from his studio, Oh Production. As for Kanada, he brought with him the two other aces of Studio N°1: his teacher Takuo Noda, and his first student, Osamu Nabeshima. Their presence was central, as Kanada-style effects would be recurrent throughout the entire movie.

Kanada himself probably did much less work on Farewell than on the first movie; however, what he did was even more idiosyncratic: the film probably represents the peak of his liquid-fire effects style, and contains some of his most iconic moments. In terms of effects, it therefore represents one of the major steps of the evolution of Kanada’s style, and the point where his experimentation with colors becomes more and more distinctive: after the abstraction reached in the first movie, the effects start to be used for figuration and start prefiguring the fire dragons of Genma Taisen.

In terms of character animation, what’s the most striking is how little corrections Komatsubara made on Kanada’s scenes; to put it bluntly, it looks like he didn’t make any, and gave him total freedom. In this sequence, all characters are wildly off-model and their look has nothing in common with the rest of the movie. Later, on the climax, Kanada felt free to include multiple easter eggs, most notably one of his mascot characters, Kabonen from episode 6 of Zukkoke Knight: Don de la Mancha, and none of them was taken away. It therefore seems like Kanada was given almost complete creative control over his own scenes, making them some of his most powerful and creative work.

Meeting Miyazaki: Nausicaä of the Valley of the Wind

The early 80’s is when Kanada started his meteoric rise in the anime industry: he got more and more work as animation director on Tôei movies such as Queen Millennia and Future War 198X in 1982, and then made one of his most iconic works on Genma Taisen in 1983. It’s of course hard to tell, but had it not been for the failure of Birth, Kanada might have risen even higher and maybe would have directed something of his own. But what interests me here is the other prestigious project he worked on in 1984: Hayao Miyazaki’s second feature film, Nausicaä of the Valley of the Wind.

Now that Miyazaki being a central figure of world animation is something we’re used to, it’s hard to go back in 1984 and picture how much of an event Nausicaä was. Miyazaki had already been something of a star in the industry since the moment he had left Tôei, and the success he got with his first TV series, Future Boy Conan, and movie, The Castle of Cagliostro, made him known to a wider audience. And now, he was to direct an adaptation of his own work, an already popular science-fiction manga. His reputation preceded him, to the point that all the animators who first collaborated with him on the movie confess having been intimidated at first.

It wasn’t the same on the producers’ side, though. The movie was given a small budget and, through some curious alleyways of pre-production, its animation was largely handled by the small and unknown studio Topcraft, that had been created in 1972 by ex-Tôei producer Tôru Hara. It was a curious decision because Topcraft specialized in subcontracting for American studios, most notably Rankin Bass. Their approach to animation was wildly different to that of other Japanese animators, and it seems that most of their work on Nausicaä was heavily corrected. As Benjamin Ettinger points out, “it must have been a real eye-opener to the studio’s animators to work on that film with figures whose approach was much more individualistic and focused on creating movement that felt good.”

The other specificity of the movie is that the animation director wasn’t Miyazaki himself, but someone he had never personally worked with previously: Kazuo Komatsubara. It is most probably Komatsubara who brought on board many “charisma animators” whose philosophy was very different from Miyazaki’s: the rising star of effects animation Takashi Nakamura, the ex-Oh Pro alumnus and studio Z3 member Osamu Nabeshima, and of course Kanada himself. Kanada and Miyazaki had already met a few times by then, and Kanada had even done some of his first in-betweening work on a Miyazaki episode back in the 70’s. But it’s on Nausicaä that began a relationship that would last for a decade.

With Komatsubara as animation director, the movie’s animation is probably the most diverse among all of Miyazaki’s catalogue: it’s much easier than in his other movies to spot the different animators and their sensibilities. The level at which Miyazaki exerted his control was therefore not the animation itself, but the storyboards. Especially compared to Rintarô’s, they are very detailed and sequential: every movement and expression is recorded. To put it in raw numbers, Nausicaä’s storyboard is as long as Galaxy Express’ (around 550 pages for each) even though the first one is shorter by 30 minutes and Galaxy Express’s storyboard is already notably long. Stylistically, one of the most striking elements in Nausicaä compared to other Miyazaki movies is the use of impact frames. You would think these were implemented by Kanada himself, since they were one of his frequent techniques. But in fact, the way explosions are drawn in the storyboard seems to indicate that they were planned at that level.

Kanada animated many sequences of the movie, but let’s just focus on the most famous one, the air battle scene. As can be expected from Miyazaki, the layouts are fairly complex, as the airships chasing each other move in all directions across the screen. Most of it is already present in the storyboard, so even though it’s perfectly fair to credit Kanada for being able to pull the scene off, it’s not his own inventivity that’s behind it. The only place where he could really do things his own way was in the actual drawing style, and especially in the shapes.

Indeed, all the effects are very easy to recognize: they’re characteristically angular and the color work is fairly unique. What interests me the most in this sequence is the missile part, from 0:14 to 0:17. The storyboard details the movement, but the actual motion is quite striking. First, it must be noted that the animation is entirely on 1s and 2s with close spacings – the kind of combination that’s usually used to convey detailed, nuanced and realistic motion, one seldom used by Kanada. The amazing sense of action therefore doesn’t come from the timing but from the rhythm of the motion itself: in the first shot, each missile is launched in sequence, and they then enter in the next shot one by one, something that doesn’t seem to be clearly indicated in the storyboard.

Kanada in Ghibli

1984 was a turning point in Kanada’s career both because of Birth and Nausicaä. The failure of the former and the success of the latter profoundly changed the kind of work Kanada would do. Except for the Tôei movie Odin: Kôshi Hansen Starlight in 1985, for which he had probably already been recruited before Birth or during its production, he did almost no animation direction until the end of his career, and only some storyboards here and there. Most of his work outside Ghibli was minor, and can be boiled down to him helping out some of his friends and students on the shows or OVAs they directed or animated on.

It’s hard to tell what was the precise status of Kanada in Ghibli. He participated in every Miyazaki movie, and had his own desk in the studio until 1992. However, in all the lists of Kanada’s works, he is still credited as a freelancer. Even from there, he doesn’t seem to have had an exclusivity contract with Ghibli: Kanada kept his strongly independent spirit. After Takuo Noda had joined Madhouse in 1983 and N°1 was dissolved, Kanada might have joined or become strongly associated with Kaname Pro; but Birth put an end to that. Then, in those years, besides Ghibli, one would often meet Kanada in studios One Pattern (the successor of Masahito Yamashita and Shinsaku Kôzuma’s studio OZ) or Z5 (created by ex-Shingo Araki students who were close to Kanada).

The Miyazaki-Kanada collaboration is often considered paradoxical, and it’s widely thought that the latter was somewhat on the losing end of the partnership: his idiosyncratic style didn’t really fit in with Miyazaki’s, and he was corrected to the point of being almost invisible. My analysis of Nausicaä should already have nuanced this idea, but I’d like to go further and show that it misses some important facts: Kanada’s stylistical evolution cannot be understood without his works on Ghibli movies. If you just think of them as a strange parenthesis between his extremely fluid liquid-fire style effects of the early 80’s and the very snappy, angular animation of the 90’s, there’s no way of making sense of that transition. But this transition partly happened in the Ghibli period.

Nausicaä is already a first sign of this: the fire effects are remarkably more angular than anything Kanada had done before. This would go on to become even more remarkable on Laputa. The first example of that is Kanada’s most famous contribution of the movie, the storm when lightning transforms into dragons. Because it’s a case of figurative effects and features dragons, this scene tends to be read in the continuity of Genma Taisen’s finale. But, without denying the similarities, I’d relativize that, because there’s a central element missing: the use of color to create motion. In this sequence, the dragons are monochrome, and the motion is only created using lines. In that sense, Kanada’s effects work must be understood in relation to two things: first, Masahito Yamashita’s effects at roughly the same time which use lines in a similar, although more baroque, way, and Kanada’s previous work on beam shapes, which had featured this angularity for a long time.

The same applies to character animation. The other standout moment that Kanada did in Laputa is the encounter/fight between the inhabitants of the mining town and the pirates in the first part of the movie. What’s striking in this scene is the rhythm: we get long moments of anticipation and then sudden, very fast motion on the hits. The spacings suddenly get very wide at some moments, and the limbs are always extended or making stark angles. What this looks like is, in fact, less Kanada than early Yoshiyuki Momose or Yoshifumi Kondô. In other words, this is a return to Kanada’s 70’s inspirations, probably spurred by Miyazaki who had been close to the A Pro school. This is absolutely essential because, as I will show in a later article, Kanada’s late style in the 90’s can in fact be read as a return to the A Pro fundamentals. And this sequence from Laputa shows that this had already started in 1986, under Miyazaki’s supervision.

It seems that Kanada tried out his new philosophy of character acting in My Neighbour Totoro. Testimonies from the movie production seem to indicate that the animation of the bath scene was extremely striking and idiosyncratic. Maybe it went a bit too far – although the finished product retains a great sense of liveliness and fun, it’s not as bold as you’d expect. In fact, Shinsaku Kôzuma told that, on Totoro, Miyazaki flat-out rejected some of Kanada’s layouts and had him use an enlarged version of the storyboard instead. A clue that supports this is the fact that after Totoro, Kanada was seldom given pure acting cuts – Miyazaki might have been afraid of him going too much off-model. Off-model was something Kanada was an expert in, and his Ghibli layouts, most notably from Laputa, seem to indicate how playful and free-spirited he was with his work and Miyazaki’s characters.

Despite all that, Kanada was no doubt a central figure in Ghibli. On Laputa, he was credited as “head key animator” (原画頭). Some accounts claim that this honorific title is a relic of a new kind of organization Miyazaki wanted to try out: like it had been the case in Tôei Animation in the 60’s, it would have been a two-tier system with head key animators doing first key animation having a series of assistants doing the second key animation. However, this experiment was soon abandoned, and Kanada’s credit was mostly an in-joke within the Ghibli staff, who called Kanada “chief” (頭) because of his prestige and prominence. On all Ghibli movies, he was generally given the most complex scenes on the movies: on Totoro, he had to animate one of the most impressive Catbus sequences, with the very intricate movement of all the legs and a long moment of background animation. On Kiki, it was a lot of flying scenes, most notably the one when Kiki is attacked by a pack of crows, and the finale of the movie. Finally, it was on Porco Rosso that Kanada had the most opportunity to showcase the range of his talent.

I’ve argued previously that Kanada’s animation in the 70’s represented a key step in the development of realism, and that under the influence of Kazuhide Tomonaga, he slowly turned towards a realistic approach of mechanical animation. Porco Rosso probably represents the apex and conclusion of that evolution: in charge of many flying scenes, Kanada revealed his ability to convey the volume and weight of machines. In this sequence, the hardest would have been to make the movement of the plane believable, and to convey both the danger and speed. He established the second through amazing bits of background animation, but he also managed to make the plane evolve naturally in such a quickly-evolving environment. 

The machine is animated on 2’s, often on 1’s, and there’s something impressive about its consistency. Even with such uniform timing, Kanada created a sense of rhythm, mostly through the bursting water splashes. An especially impressive moment is around 1:11, when the plane just rebounds on the water. It enters the frame on 2’s, but after the first bounce, it goes on 1’s until the end of the shot. But as it goes further away from a camera and bounces again, the spacings get closer and closer, creating both the feelings of distance and speed. Everything is carefully studied, and Kanada’s work on the movie deserves to be counted among the masterpieces of mechanical animation.

A new shift: Download

I tend to consider that Porco Rosso is Kanada’s last work in Ghibli. He did do some animation on Princess Mononoke in 1997, but his contribution was notably short. The movie’s production report indicates that Kanada wasn’t meant to work on the movie in the first place, and that he ended up on it because Ghibli was understaffed. I believe that Kanada distanced himself from the studio on Porco Rosso following its internal reform: during the production of the movie, Miyazaki and the rest of the studio’s management decided to raise the salaries of the employees and made a mass series of recruits. Following this, Ghibli started to rely less and less on the services of freelancers such as Kanada. There might also have been some personal disagreement between him and Miyazaki; but most importantly, there was the fact that Kanada’s style was taking a new, major turn, that absolutely didn’t fit the Ghibli aesthetic anymore. This turn took notably shape in Rintarô’s 1992 OVA, Download: Namu Amida Butsu wa Ai no Uta.

The OVA must be understood in context. It is one of the many works Rintarô directed in his OVA period, which roughly goes from 1987 to 1994, during which he tried out, with uneven levels of success, very different works and styles. Download was, as strange as that might sound, a cyberpunk comedy. It was one of the late meeting places of the oldest members of the Kanada school: Takuo Noda, who had been Kanada’s master, had become one of Rintarô’s closest associates since Genma Taisen in 1983, and he invited not only Kanada, but also Kazuhiro Ochi and Masahito Yamashita.

Kanada was therefore in a comfortable environment, and had the opportunity to do the character designs himself. It might seem strange that he didn’t also do the animation direction; it’s possible that it was offered to him. But he was likely already busy on other projects (Porco Rosso was in production at the same time) and handed the position down to Noda. It had been a long time since the golden days of their collaboration, but Noda and Kanada were old friends: it’s not like the animation direction was given to entirely foreign hands.

Download featured Kanada’s first character designs since Birth, and it’s hard not to see the two in continuity. The design of the female characters is very similar in both OVAs, although Download went even further in exaggerating their curves. All the characters were simple and animation friendly, and expressed a constant feeling of energy and joy. In that, they were probably the perfected realization of what had been tried out in Birth’s character animation: the possibility of a Kanada-style character animation that would put aside any form of rigidity in favor of liberated, spontaneous movement – in other words, expand to character animation what Kanada had done to effects.

In concrete terms, this meant three things: a prioritization of motion above all else, a disregard for staying on-model, and a liquid approach to shapes. These are the characteristics of what I call “flow animation” which would become more and more important in the 90’s and triumph in the 2000’s. All this was already present in Birth, and would be pushed further by two artists throughout the 80’s. The first is Kanada himself, notably on The Chocolate Panic Picture Show in 1985 which, just in-between Birth and Laputa, features both a flowing, deformed kind of animation, and snappier timings and cycles. The other is Shinsaku Kôzuma, one of the few students of Kanada who didn’t completely turn to effects animation in the second half of the 80’s. His work made a heavy use of shape deformation, squash and stretch and stark slow-in/slow-outs, creatin a yet unprecedented feeling of rhythm.

But it was Download that really showed what this approach to animation meant. This was partly due to the generally better production conditions compared to Birth, Kanada’s even more perfected character designs and a new philosophy he had probably agreed on with Noda. But there’s also a determining element: Gosenzosama Banbanzai. This 1990 OVA was one of the masterpieces of realistic animation, but it was also, through the pioneering work of Satoru Utsunomiya, Shin’ya Ohira and Mitsuo Iso, one of the major steps in the development of flow animation and the new philosophy of character acting.

It is therefore no surprise if two of Download’s foremost animators, Kazuyoshi Yaginuma and Tatsuyuki Tanaka, were pure products of the realist school. The two men were intimate friends and had both started their career in studio Telecom. Close friends with Takashi Nakamura and Shin’ya Ohira, they were at the center of the realist circle, and Tanaka even made his debut as key animator on Akira, before working on Gosenzosama Banbanzai where he met Iso. Tanaka’s animation is probably one of the highlights of Download. It is highly modulated, switching freely between 1s and 4s, close and wide spacings. This creates a sense of frenesy, furthered by the multiplication of movements of all parts of the body, clearly inspired by Iso. The other Iso inspiration is probably in the smears: whereas Kanada school animators like Shinsaku Kôzuma used wide, deformed smears that exaggerated the motion, following Gosenzosama, Iso pioneered smaller smears that would only distort the outlines of objects, most notably of the hands: they were in the service of dynamism rather than deformation for its own sake. All in all, there’s a very fun and cartoony aspect to it, that’s the perfect crossing point between Gosenzosama’s expressive animation and Kanada’s crass, body-centered kind of humor.

What makes this animation so unique in Download is its contrast with the more angular Kanada-style effects, and Kanada’s own animation. He animated by himself the long and impressive final scene. What little acting there is here is as energetic as Tanaka’s, but the shapes are still more rigid. The thing that really stands out, however, is the effects animation. The star-shaped beams are directly taken from Kanada’s work on the Odin movie from 1985, and the simplicity of the fire and smoke effects is clearly descended from the smoke and fire of Nausicaä. But what changes is the return of speed lines, which had totally disappeared from Kanada’s animation in the 80’s, and the intensive use of straight and geometrical shapes. Kanada’s use of a ruler had never been more obvious, and what you see here are all the elements of his later style that would become so iconic in the 90’s and 2000’s.

There would still be a lot to be said about Kanada’s work in the 80’s. But this article had a somewhat reduced scope, and its goal was not to be a complete retrospective. What I wanted to explore were two things: one, what was the relationship between Kanada and the directors he worked with, and where exactly were the places of innovation and invention. And, two, how did the transition from his apparently realistic style of the late 80’s to his heavily stylized animation to the 90’s happened. In fact, I think it is a conclusion of the research I made here that there was no transition – Kanada held the two concurrently, and there was never any contradiction between them. This period is so interesting precisely because Kanada’s style isn’t fixed, and in 1992-1993, Kanada’s character animation could clearly have gone in two directions: the liquid, free-flowing style tried out in Birth and Download on one hand; and, on the other, the solid, rigid, A-Pro inspired animation that he began to use on Laputa.