Cover image: a key frame by Shin’ya Ohira from ZZ Gundam
This article is part of the History of the Kanada school series
This is just my personal opinion on the matter, but I don’t think many animators ever reached the same level of genius as Yoshinori Kanada in terms of originality and ability to ceaselessly renew their own style. In the course of the chronological period followed in this series, there is however one artist whose ability to do that rivals Kanada’s: that is Shin’ya Ohira. Ohira is widely considered to be one of the most talented animators ever for the highly idiosyncratic and complex style he developed in the 2000’s. But before reaching that stage, he had already pushed the possibilities of the animated medium further, not just once, but three times: first as a Kanada follower, then as a student of Takashi Nakamura, and finally as a highly unique and idiosyncratic animator. It is these first two periods I’d like to focus on in this article.
From Ohira’s own testimony, his story as an animator started on June, 16, 1982, when he watched episode 32 of Urusei Yatsura at 16 years old. Nicknamed “the library episode”, it was animated by Masahito Yamashita and his team from studio OZ, the vanguard of the Kanada school at the time. According to Ohira, the animation impressed him so much he almost choked on his dinner watching it.
After that, Ohira was properly introduced to the industry thanks to his highschool upperclassman Kôji Ogawa. Coming from a background that wasn’t very well-off, Ohira needed to earn his own bread early on, and it’s Ogawa’s advice that he applied for an animator recruiting advertisement published in the magazine Animage, moved to Tokyo and joined studio Pierrot, which had produced the first half of Urusei Yatsura. He did his first in-betweens on the super robot Sei Jûshi Bismarck, in 1985, quickly rose in the ranks and, after some more time in-betweening on Ninja Senshi Tobikage, debuted as key animator on episode 11 of that show. It’s there that he met fellow animator and Yamashita fan Atsushi Yano, with whom he would later create his own place, studio Break.
It seems that Ohira already had a clear idea of what he wanted to do as an animator when he entered the industry: although clearly influenced by Yamashita, his style stands out for how radical it is. Long gone were the days of Kanada’s fluid liquid fire animation: the effects were as stylized as possible, with triangular explosions and abstract geometric shapes of the fire and smoke. Smears were everywhere, and the shading extremely heavy. Color was there to be expressive, and Ohira used it as much as he could. Finally, the impact frames were so stylized and colorful they would sometimes lose the viewer or mix up with the heavy smears. Every one of Ohira’s drawings was full of information, forming extremely rich, almost abstract, compositions.
His work on just 4 episodes of Tobikage made him noticed so quickly that he soon left Pierrot and stopped working on TV series – he immediately transitioned to OVAs and movies with much higher production values and met many industry legends. The most important one was Ichirô Itano, who had become one of studio AIC’s key assets and directed many of their OVAs. The first was Megazone 23 Part II. There, Ohira was only second key animator, completing and correcting cuts by other animators. Even though it would still be a few years before Ohira switched to realism, it’s interesting to note that the sequences he worked on were handled by two early important realists: Toshiaki Hontani and Kôji Morimoto. It enabled him to further his skills as an effects animator, but also to get his first real taste of character animation and more realistic designs.
Following Megazone, Ohira became a regular collaborator on AIC-Artmic coproductions from his own studio Break: he worked on Gall Force: Eternal Story and Ai City in 1986, then Gakuen Tokusô Hikaruon (where he might have met Kanada), Bubblegum Crisis, Black Magic M-66 and Gall Force 2 in 1987, and finally Dragon’s Heaven and Riding Bean in 1988 and 1989. He also got close to Sunrise starting with the Panzer World Galient OVA in 1986 and became a regular on Gundam entries starting with ZZ. To put it shortly, these were very busy years, and yet Ohira was barely over 20.
After Gall Force, Bubblegum Crisis was the second time Ohira contributed to a Ken’ichi Sonoda adaptation. However, he had little occasion to play with Sonoda’s iconic character designs: in these years, he specialized in effects animation, an art he refined to its utmost in the many fight scenes of the two episodes he worked on.
However, what’s notable about his sequences on Bubblegum Crisis is that Ohira didn’t just stop at simple effects, but mastered the ability to animate complex layouts and pull off impressive moments of background animation. In this short sequence, you find some of his characteristic effects, as well as complex mechanical animation in the bike’s transformation. Although they didn’t meet on the series (Ohira worked on the first two episodes, Obari on the last two), this kind of very detailed motion was close to what Masami Obari was doing at the same time. But what really interests me here is the moment at 0:14, as Sylia jumps behind the Boomer and the camera rotates around them. The attack is unexpected, and the sudden burst of background animation also is. What’s most notable is the rhythm: it starts very fast, with wide spacings, then slows down around the moment of the kick, and goes back fast when the Boomer falls down.
This was, however, only the beginning of Ohira’s attention to detail. Its most famous example is this sequence from Gall Force 2, on which he is said to have spent 3 months. A large part of the sequence is on 1s, and it is also said that he made up to 300 drawings for 5 single seconds of animation – which means, if you do the count, that he didn’t just do it on 1s, but also animated multiple layers at the same time and on the same framerate. On Project A-Ko VS, he became the stuff of legend, with a sequence that reportedly necessitated around 1000 in-between frames (but their number was reduced by the animation director), and his own key animation scattered on 14 layers – it wasn’t just the possibilities of animation that he was pushing further at this point, but those of photography and compositing.
Even without this information, the scene is impressive, and some of Ohira’s best effects animation. Even though it’s the only object on screen during the start of the sequence, the beam manages to establish its own space and depth through its extremely fluid and dynamic motion. Every explosion is very detailed, with lots of debris and smoke effects with their own snappy movement.
The amount of work Ohira was ready to put into a single cut, and the extremely fluid result, show that he was progressively distancing himself from the Yamashita style he had first adopted. He kept it in the angular shapes of many of his effects, but they stopped being as flashy as on Tobikage. Rather, his extremely detailed debris work on Black Magic M66 betrays the growing influence of Hideaki Anno (Wings of Honneamise had just come out) and Takashi Nakamura, with their own unique effects animation.
It would not be long before Ohira and Nakamura worked together, in 1988, a key year in both men’s careers. Before Akira, Ohira contributed to one of the greatest animated movies of the decade, Char’s Counterattack. In some impressive mobile suit battle sequences, he yet again showed his ability to choreograph and animate complex three-dimensional movement. The effects completely follow the general aesthetic of the movie and have nothing Kanada or Yamashita to them anymore. What dominates is the dense and fluid movement of the robots, with the mobile suits on 2s and the funnels on 1s.
However, before he definitely did the jump and shifted to realistic animation, Ohira delivered one last stunt of impressive Kanada-style animation, along with Kôji Itô and Masayoshi Tano, on the Captain Power training video. He was animation director, but busy with many other things, so he did little corrections on the other animators’ cuts, mostly adding effects and some animation here and there. This video isn’t very interesting in itself, but it is without a doubt an impressive artistic feat: shot from the point of view of a fighter jet, it is uninterrupted background animation for almost 10 minutes. It’s hard for me to appreciate it beyond that purely technical level, but what a technical level it is!
Ohira’s animation in his first period can only be characterized by its excess. But it is precisely what made him so special, already then: a capacity to go always further than the limits of what one believes it is possible to do in animation. In its glorious decadence, Ohira’s animation was the most beautiful swansong the then repetitive and monotonous Kanada style could have dreamt of. It would take another 20 years, and the work of animators such as Jun Arai, to pick up things where Ohira had left them.
It’s in 1988 that came what can be considered the major shift in Ohira’s career, Akira. As I noted, his transition was already noticeable in his earlier works. But considering that he had only begun key animating two years earlier, this was a very quick change: no one would have expected 1986’s Ohira, Yamashita’s most radical follower, to work under Takashi Nakamura on what quickly became the flagship of realist animation. And yet, that’s exactly what happened. Ohira ended up there a half by chance, and this is why, as impressive as his work was, it doesn’t seem to have been major: he was mostly helping out on unassigned cuts here and doing second key animation there.
Despite that, it was a turning point. His cuts were extremely impressive, and although they were mostly effects, he started to get closer to character animation and slowly began to focus himself in that direction. It was the right place for it: on the movie, he of course met Takashi Nakamura, but also, on the ending on which he helped out, people like Yasuomi Umetsu, Tatsuyuki Tanaka, or Ghibli animator Makiko Fukaki. It’s also where he met his friend and collaborator Shinji Hashimoto, with whom he would often work during his second period. It was probably in part thanks to this collaboration that Ohira’s shift to realism could happen.
Ohira might have known Hashimoto before that, since he had been working with his younger brother, Kôichi, since Black Magic M-66, and then on to Char’s Counterattack. It was Kôichi, the assistant animation director of Gosenzosama Banbanzai, who then brought Ohira on to the production. He worked on episodes 1, 2, 3, 4 and 6, and it’s the OVA that finally enabled him to try out character animation and a wildly different style, with complete freedom.
What characterizes Ohira’s animation from then is probably the playfulness (to be expected in such a production) and the snappiness. Although this cut is mostly on 2s, the spacings sometimes get suddenly very wide. The father’s acting part, from 0:10 to 0:16 is extremely modulated, with sudden switches from 1s to 4s. This moment is probably the most expressive of the cut, with lots of little movements of the hand, a heavy use of smears and some other cartoony deformations.
The second shot in this sequence, a close-up on Inumaru’s face, is also very interesting. It’s the first time Ohira had the opportunity of handling such detailed character acting, and he pulled it off splendidly. What’s notable is the amount of detail, mostly on the mouth and eyes. This is no simple lip-flapping, but real animation of the mouth, often on 1s. As for the eyes, we don’t yet get Ohira’s characteristic blinking, but there’s a lot of care put into their expression, especially in the eyelids.
After Gosenzosama, Ohira found himself involved in many high-profile productions of the developing realist aesthetic, most notably studio Ghibli movies: his first collaboration with them was on Takahata’s Only Yesterday in 1991, but just after that he worked on Porco Rosso and My Neighbors the Yamadas. Even though he went through a major stylistical shift in the late 90’s, he would keep working with the studio on one of the Ghiblies shorts and all following Miyazaki movies. I think it’s fair to say that Ohira effectively took Kanada’s place in Ghibli – that of a highly charismatic and individualistic animator whose work kept standing out even under Miyazaki’s corrections.
Besides those prestigious works, however, what’s by far the most interesting in Ohira’s career between 1990 and 1994 (a very short period) are the more obscure but highly experimental projects he found himself involved in as a director and/or animation director. There are four of those, which I’ll cover in some detail: The Hakkenden episode 1, in 1990; “The Antique Shop”, in 1991; the Junkers Come Here pilot in 1993; and finally The Hakkenden episode 10, in 1994.
I already mentioned The Hakkenden as one of the most important works of the early 90’s, a key to understand the transition from realism to flow animation, the two dominant styles of the decade. And it so happens that Ohira was one of the major actors behind this transition, as the joint animation director (with Shinji Hashimoto) of the first episode. It was probably the most crowded one, with animators as important as Mitsuo Iso, Hiroyuki Okiura and Osamu Tanabe. It’s where Hashimoto produced the seminal sequence of the flow animation style, but even Iso was going in the way of more spontaneous and liberated motion, while Okiura was putting out increasingly complex fight scenes with more and more characters moving simultaneously.
More generally, the fantastical setting and the atmospheric direction were the perfect fit for what Ohira and his team seem to have been trying to achieve: realistic and detailed movement, but also a feeling of uncanny, of something so natural it becomes strange and somewhat frightening. This is probably best achieved in Osamu Tanabe’s sequence, in which the ugly face of the crying child takes on an almost terrifying aspect, with its overly large tears and expressions.
It was no surprise, then, if Ohira’s next project was a horror story. “The Antique Shop” is the last part in an anthology OVA coordinated by Ichirô Itano, Yumemakura Baku Twilight Gekijô. An obscure production (for reference, it hasn’t been licensed or subbed), it nevertheless involved major figures: Yasuomi Umetsu on the first short, Shinji Hashimoto on the second one, and Ohira, as writer, director, storyboarder and animation director on the last one. As Benjamin Ettinger noted, this work was then exceptional in Ohira’s career: this would be the first in the extremely rare occasions he did direction, storyboards and character designs. Like Kanada, he remained a pure animator, but the work he did on each of the things he directed and storyboarded is exemplary.
“The Antique Shop” reunited most of the team that had met on Gosenzosama, all people for whom Ohira had unquestionably become something of a leader: Tanabe, Iso and Hashimoto, but also two young prodigies that had made themselves known on Akira, Tatsuyuki Tanaka and Kazuyoshi Yaginuma. Those latter two would become closer to flowing and deformed animation in the vein of what Hashimoto had done on The Hakkenden, and “The Antique Shop” was the occasion for them to show off their talent and how fitting their liberated style was for Ohira’s photorealistic designs.Indeed, it is the designs that must first be noted. In stark contrast with Satoru Utsunomiya’s work on Gosenzosama, and even with Umetsu’s first short in the anthology, Ohira didn’t choose simple, Nakamura-inspired designs, but instead went towards photorealism. What he was looking for wasn’t the “clean” look that most realistic animation still adopted, but the visceral energy that comes from the real presence of bodies and their physical expressions. Benjamin Ettinger conveyed this feeling better than I could:
Why is it that we can never see people who actually look like people in animation? That does not mean having to be photorealistic. It means being honest about physical features, thinking honestly how to show a face that can convince any viewer that a soul inhabits it, and not simply adhering to a style out of lack of courage to tweak convention. In the face of the woman on the porch, Ohira had created what for perhaps the first time in animation to me struck me as looking like a living, breathing human being. She was not prettified. She was homely. But she was one of the most beautiful characters I’ve seen in animation because it was a face I could believe in. It was as if I experienced a sense of relief finally being able to see that kind of face in animation.”
In more technical terms, what did this mean? That, as Ettinger noted, characters wouldn’t look pretty or attractive in any way: they would have small eyes, big hands and wrinkles all over the face. One of the most impressive details about the animation is that characters blink – animated characters never blink, but Ohira insisted that they should. There is a strong sense of embodiment, conveyed not only through the animation, but also by the camerawork: in the first part of the short, when the main character visits the titular antique shop, there are many first-person sequences in which the camera moves wildly, following the drunk character’s staggering gaze.
The direction didn’t make anything easy for the animators, as this sequence by Tatsuyuki Tanaka illustrates. The layouts and camera angles are among the most complex for an animator to put into motion: in the first shot, a character moving into depth; and in the second shot, a stark high-angle perspective. And yet, the animation is perfectly fluid and full of little motions, especially in the hands and arms. At 0:07, when the man brutally pushes the woman away, you can see the muscles on his back. The woman’s fall, on 2s, is painfully detailed: going back on it in slow-motion, you see her grabbing her belly and her face distorted in pain, before the second shot when her head slightly bounces off the ground.
With this 13-minutes short, Ohira had effectively pushed the boundaries of the medium further, and it would be easy to think that there would be nothing to equal “The Antique Shop”’s superb animation. But 2 years later, in 1993, Ohira and a slightly different team (Tanaka and Yaginuma had been replaced by Manabu Ohashi and Akihiro Yuki) produced what is probably one of the greatest masterpieces of the animated medium: the Junkers Come Here pilot film. This 3-minutes short, a simple compilation of scenes that were supposed to be featured in the movie, long had somewhat of a legendary status: it was not publicly released after its production, and it got its reputation through the word-of-mouth from the amazed animators and producers that had had the luck of seeing it. Since then, it has been made available, first on the movie’s DVD boxes, but also on the animation website Catsuka.
One of Ohira’s original design sheets for the Junkers pilot
Ohira’s designs were simpler than in The Antique Shop, but still a step more complex and detailed than those of the final 1995 movie. They were less pretty, with more focus put on the mouth and eyes of the main character. The backgrounds, too, were slightly different: even more abstract and washed out than in the final movie, they made the animation stand out that much more. And what animation that was!
Perhaps it comes out most clearly in this short sequence by Osamu Tanabe. First, it is almost entirely on 2s, with close spacings: this is what creates the sense of fluidity and ease. It is very important here, because it’s what makes the movement feel so spontaneous, perhaps the greatest quality of Ohira’s animation. But it’s not just monotonous timing either: when the girl bends down and goes back up, the spacings get wider and the timing slower for just a few frames, making it feel more sudden and irregular. For such a short cut and apparently unimportant motion, it’s also impressive how each little element is animated: the girl’s pigtails, the fabric on her clothes, and the little figurines attached to her backpack. This care put into the detail and the secondary movement in every moment of this short is probably what makes its animation a masterpiece of photorealism.
But in my mind, there’s more to it than this. Ohira’s animation is not just important because of its physics. I think the Junkers pilot is one of those very rare moments in animation when the realism is so strong that the word doesn’t even apply anymore, when realism as an aesthetic implodes. The first reason for that is what I’d call the spontaneity of the motion: it feels natural, both intentional and free. In this, the short took a direction radically opposed to the one the other realists had been starting to open on movies such as Patlabor 2. This division was conceptualized by Satoru Utsunomiya, who noted in an interview a difference between two kinds of realism – the one that tries to emulate live-action cinema’s acting, and tends to suppress this kind of small, almost unnoticeable action, and the one that tries to reproduce spontaneous, daily life movements.
This goes beyond simple photorealism, as both kinds of animation mentioned aim for it in some way – that is, a detailed reproduction of objects and motions as they would be if they were photographed or filmed. The difference is in the kind of thing that’s animated. And the mastery of spontaneous action is what sets Ohira apart: his approach is for example quite different from Hiroyuki Okiura’s work, which tends to consider photorealism as an end in and of itself and tends to emulate acting performances. Okiura’s animation tends to feel cold, almost mechanical, as a result. I don’t see that in Ohira’s animation: there’s emotion, a kind of soul behind each little movement.
The other thing would lie, I think, in the shapes. It’s hard to put into words, but in that regard, Ohira’s work is close to Iso’s and what I call the “trembling” quality of the latter’s early work. Even though the motion is very fluid, there’s a kind of instability to it, as if the shapes of the characters were about to give up and explode into blobs. But they don’t, and it’s this state between perfectly realistic physics and an impending sense of total collapse of all forms that makes Ohira’s animation so powerful.
This is the point where you see how much Ohira’s character animation owes to his pre-Akira work and inspirations. Indeed, even though it is much more cartoony, you find a similar approach to shapes in Shinsaku Kôzuma’s early work, which had probably reached Ohira through Yamashita. I’d argue that this leaning was a major drive of Kanada-style character animation, that only Kôzuma and Ohira managed to perfect, one in the direction of expressionism, and the other towards realism. Moreover, it must be noted that Ohira’s post-Akira smoke animation has the exact same unstable, but also animistic feeling to it where each little element has its own individuality and gives way to some new shape in an organic and constant evolution. Whereas most effects animators, like Kanada or Nakamura, brought character to effects animation, Ohira did the opposite: he brought the density and spontaneity of effects to character animation.
After all that, it’s not so surprising if, just a year after the Junkers pilot, Ohira went through his third major stylistical shift, in which shapes would, in effect, start to disintegrate and the linework go through a paradigm shift. That was Ohira’s fourth masterpiece: The Hakkenden episode 10. It was his last work before he left animation for 3 years to care for his family’s business. It would be tempting to attribute his sudden stylistic shift to this temporary departure from the industry, but the dates don’t add up. Probably as a result of the frustration of being taken off from the Junkers movie, Ohira had this in mind since 1994.
Ohira storyboarded and directed the episode, but he left the animation direction to a young animator he had heard was making waves on the production of Crayon Shin-chan: Masaaki Yuasa. To collaborate with Yuasa and his erratic style, Ohira must have already had an idea of the direction he wanted to take; but their meeting definitely marked the next step in Ohira’s evolution as an animator.
The direction was already extremely bold, pushing to the extreme everything that Ohira had tried out on “The Antique Shop”: lots of first-person shots, warped perspectives with multiple vanishing points, and uncontrollable camera movement. The character designs were completely remade by Ohira and Yuasa, to an even more radical degree than what Nakamura had done on episode 4. Whereas Nakamura had “cleaned” and simplified them, Ohira and Yuasa made them much rougher and complex, adding lines everywhere, bending everything that was straight to achieve a sort of synthesis between the visceral drawings of 70’s animator Keiichirô Kimura and the intense realism of the “The Antique Shop”. The linework was probably the most important: it took the flowing, deformed animation of people like Shinji Hashimoto and Shinsaku Kôzuma, but instead of using it with simplified designs, it would add even more visual information and create an unparalleled sense of organicity.
An incredibly bold work, the animation on the episode didn’t follow any conventions. It never tries to be appealing, and there seems to have been no such things as model sheets: the lines were so irregular and shifting that characters’ shapes and appearances completely changed from a frame to another. The raw feeling that Ohira sought to achieve had been reached to an unprecedented degree through the dreamish – or rather, nightmarish – atmosphere of the episode. And yet, Ohira never relinquished the realism that he had mastered. In this sequence in particular, it comes out most clearly in the impeccable sense of weight, but also in all the motions of the arms swinging wide.
Indeed, even though Ohira’s work here has mostly been interpreted as his first step in steering away from realism, things aren’t so clear, if only because fellow animators Toshiyuki Inoue and Satoru Utsunomiya consider The Hakkenden episode 10 to be one of the crowning achievements of realistic animation. I share their opinion, because for me, this episode shows that the search for a constant and convincing sort of movement so present in Ohira’s previous works had extended beyond photorealistic designs and motion – it now reached his very linework, as if the bodies themselves were crossed all over by some energy that intensified every little detail.
Since then, Ohira has gone through no new radical stylistical evolution and only gone further in the new way he has opened: a kind of animation that would recreate everything from the ground up, reimagining the very nature of animated motion through linework and the complete collapse of the distinction between character and effects.
It seems that Ohira is content with his last approach, a radically raw and powerful one. This is probably because, with it, he has finally achieved what he was always looking for: to evoke visceral feelings into the very body of the viewer. Ohira had succeeded in creating an aesthetic of his own that challenged the very foundations of the medium, and is for that reason one of the greatest animators in the world, without any contest. It might be surprising that, as respected as Ohira is, his influence as a teacher hasn’t spread much further. But this is, I think, easy to explain: much more than Otsuka, Kanada or Nakamura, the “big three” of Japanese animation history, Ohira is absolutely impossible to imitate. There are no fundamentals of the “Ohira style” that you could break down and copy, like the Kanada poses or the Nakamura debris. There is just a consistent search for something new, a personal history made up of continuous experimentation and research on how to push the medium further – and that’s something that’s impossible to reproduce.