Kanada, the first realist?

Cover image: an in-between by Yoshinori Kanada (?) from one of his sequences on Zambot 3 episode 22, in 1978

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

Today, especially in the Western side of the fandom, Yoshinori Kanada’s animation is associated with flashy, angular effects and very stylized and exaggerated motion, of the sort in which Hiroyuki Imaishi and his peers have become experts. However, if this is a valid description of the neo-Kanada style and of Kanada himself at one point, it misses a major aspect of the latter’s animation and why it was so important. Nobody would think of him as a realist, and yet… You need to look no further than the influence he had on such important members of the realist school as Shin’ya Ohira and Mitsuo Iso, or the realist shift of many of his direct students, like Masahito Yamashita, to see that there is something at play. In fact, the hypothesis of this entire article is that, from the late 70s to the early 80’s, Kanada was a major actor in the emergence of a realist kind of animation in anime.

But before starting, a definition of realism is necessary. Disney and Tôei’s animation had already created an aesthetic that can be called ‘realist’, but it had very little to do with Kanada’s and TV anime’s realism. Indeed, it was based on full, very detailed, animation, and made use of a technique anime quickly discarded: squash and stretch. It was largely that which enabled it to convey a sense of presence and of the physicality of characters. Thanks to Yasuo Otsuka’s departure from Tôei and strategic position at A Productions, the character animation in 70s anime managed to remain detailed in spite of limited framerates, although it lost its nuance in favor of comical exaggeration. On the other hand, the concurrent gekiga style had brought in anatomical detail and complex three-dimensional movement, but its focus on a visceral kind of expressivity often took it very far from any kind of photorealism.

Photorealism would, however, find its place in mechanical animation, partly thanks to Otsuka but not only, as we shall see. In the case of Kanada and his synthesis of the two aforementioned styles, he would endeavour to go further in the directions of nuanced acting and volume, but that was not all. He would largely contribute to what is considered one of the main aspects of realist animation today: the sense of weight and presence. In that, his role was vital, and it is yet another thing that lets one say that, without Kanada, anime would be very, very different today.

The birth of mechanical animation: the early years of Kazuhide Tomonaga

Before going into Kanada, it is necessary to evoke one of his contemporaries, probably the only member of this generation that could rival Kanada and who did, in multiple instances: Kazuhide Tomonaga. While his role has been somewhat forgotten with time, Tomonaga could be considered to be, along with Otsuka and studio Tatsunoko’s Masami Suda, one of the inventors of mechanical animation in Japan.

Tomonaga was born in 1952 and, at 20, he joined a small outsourcing studio, Tiger Production. From there, he immediately started in-betweening on Tôei series and did his first key animation on Devilman in 1972. Although they worked on different episodes, he might have had the chance to meet a young Kanada, who was from another studio, on Cutie Honey and Getter Robo. But it was in 1974, on Space Battleship Yamato, that they would acknowledge each other and that Tomonaga cemented his place as a rising star. Kanada did uncredited work on two episodes (episodes 2 and 26), and these seem to have been a turning point in his career. Indeed, a few months after his pioneering work on Cutie Honey, it’s on Yamato that Kanada’s style started to blossom in a personal and unique way. The same could be said of Tomonaga, who held a much more important position in the production, animating on seven episodes.

On the second episode, he animated one of the most important and impactful scenes of the series: the flashback that recounts the sinking of the Yamato during World War II. Being the recounting of historical events, the sequence naturally used real-life models. This doesn’t seem like much, but it was almost unheard of at the time: the only ones to have relied prominently on real-life machines in anime before were Yasuo Otsuka and his team on Lupin III in 1971. Besides this, the sequence was astounding on many levels. First, it was animated entirely on 2s—something very time-consuming, and therefore rare on TV anime. Just by doing that, Tomonaga established himself as a hardworking, demanding animator. With such timing he managed to create fluid motion and a sense of volume unprecedented for TV animation. In the first shot, the way the planes slightly oscillate before going down indicates the probable use of real-life footage for reference and instantly places the machines in a real-seeming, three-dimensional space. Throughout the scene, they are drawn in painstaking detail. This becomes very apparent at the moment when one of the planes explodes: all its parts go flying away, and you can still see the shape of the pilot in his cockpit. Yamato was a revolution in effects and mechanical animation, and Tomonaga played a large part in it.

Tomonaga’s next major contribution on Yamato would be on the impressive episode 22—the most ambitious episode of the series by far, and one of the most ambitious undertakings of TV anime’s first 11 years of existence—on which he is supposed to have animated most of part A. There, Tomonaga again demonstrated his ability for three-dimensional movement in impressive cuts where the ships fly, turn around, come closer and then away from the camera. If I call him one of the inventors of mechanical animation, it is because he was one of the first animators to pay attention to the material details of the objects he was animating, and to make their movement feel convincing. In other words, these weren’t toys, but real planes and spaceships that were fighting each other. He used background animation to the same effect and was, in this respect, far ahead of his time. In crafting such impressive scenes, Tomonaga was no doubt indirectly inspired by Yasuo Otsuka, just as Kanada had been: by that, I mean one of the first great scenes of mechanical animation in anime from 1969’s Flying Phantom Ship, by none other than Hayao Miyazaki, the sequence that reportedly inspired Kanada to become an animator in the first place.

In the final episode, Tomonaga also showed that his skills went beyond mechanical animation, and that he could handle characters. In the second shot of this sequence, the Yamato crew are animated on 2s and 3s, and the slight irregularity of the timing emphasizes the key frames and the fact that they have to settle on their feet to avoid falling down. This is an attention to the weight and volume of subjects that you find later in the same sequence, as the two sides start shooting: you can see the effort on Kodai’s face, as he’s aiming for the enemy, and feel the recoil of the gun after he has shot. Each Gamilus soldier makes a different movement as he falls, and extra attention is put on the last one who turns around while shooting before going down.

After such a strong start, Tomonaga was quickly scouted by the two leaders of the foremost animation studios of the time: Yasuo Otsuka, who was still the unofficial leader of A Productions, and Kazuo Komatsubara, from Oh Production. In 1975, Tomonaga first joined Oh Pro—which made Kanada say that if he didn’t join it in the 1970’s (probably after he left studio N°1, in 1977), it was precisely because Tomonaga was there and he felt he couldn’t compare. This is  understandable when you look at Komatsubara and Tomonaga’s work on UFO Robot Grendizer, as it combines very realistic and three-dimensional mechanical animation and effects that seem more inspired by Kanada. From Oh Pro, Tomonaga then worked on Lupin the Third: Part II, where he met Otsuka in person. He was convinced to join Otsuka’s new place, Telecom, in 1978, just after the production of Future Boy Conan. He would stay in Telecom for the rest of his career, but the connection with Komatsubara is probably what allowed him to work on the Galaxy Express 999 movie in 1979 and contribute to one of the most famous collaborations in anime history, which I analyze in more detail below.

Robots have feelings, too: Kanada as a mechanical animator

In the same years (1974-1979) Kanada, too, mostly worked on super robot shows, first for Tôei and then for Sunrise. His approach, however, was the complete opposite of Tomonaga’s, thereby creating another different kind of mechanical animation. Whereas Tomonaga (like Otsuka and Suda) made machines move as real machines would move, Kanada animated them like people. If Tomonaga, Suda and Otsuka are “pure” mechanical animators, Kanada drew robots like a character animator—which is precisely what made him so unique.

The best example of this is Kanada’s “late” mecha animation from this period and some of his greatest masterpieces, on Zambot 3 and Daitarn 3, in 1977. The characteristic of these two shows is that they have a very carefree protagonist and sometimes have (especially Daitarn, more of a comedy) a very light-hearted tone. This was a great opportunity, since it meant that the robots weren’t just toys, but they also weren’t (yet) war machines either. It was the perfect fit for Kanada’s playful animation, and no doubt a challenge for him: the problem he faced was, how do you make robots fun?

The first answer Kanada found was in the timing. One of the consistent principles of his style until the late 80s is that to make something funnier you have to make it jerkier. The idea is that the movement is so irregular, unnatural and mechanical, it will become funny. In other words, it’s not just the drawings that are fun, but the motion itself—something that’s absolutely not intuitive or easy to pull off. For example, in this sequence, there are a lot of holds and still frames, which emphasize the silly faces the characters are making. But when they move, it’s in sudden bursts of strange, irregular and unexpected poses. Kanada just had to do the same thing with robots to reach the same expected effects.But just that on its own wouldn’t always be enough to sell it. The second innovation Kanada brought in was at the same time much bolder and much more natural: it was to give the robots expressions. What he did was just bring to its logical conclusion the idea that the mecha is a symbolic extension of the pilot’s body. With this in mind, why not give the mecha the same expressions as the pilot, especially when you can’t see the latter because they’re inside their cockpit? The best instance of this is the following sequence from Daitarn 3: as the enemy robot gets beaten up, it keeps making sillier faces, transforming the tense fight into complete slapstick.

This is very far from Tomonaga’s very detailed kind of realism, and a completely different approach to mechanical animation. But that is yet another illustration of all that Kanada owes to the gekiga animation of Araki and Kimura: the point is not to imitate reality for the sake of it, but instead to make the animation so powerful that the viewer is affected as if it were real. Blending mechanical and character animation together might seem like a strange move, but it  was precisely what made Kanada’s robots so memorable and influential. You need to look no further than Mitsuo Iso’s work on Evangelion to see the power that animating robots like living creatures can have.

Is nuance possible when you’re on 3s?

If Kanada’s way of animating mechas boiled down to animating them like characters, what was his approach to animating characters? He tried out different things, with various levels of success, but two things he seems to have aimed for above all else are complexity and nuance. By that, I mean that he tried as much as he could to have people express different expressions, as many as possible, in a limited amount of frames. In that, he was clearly taking inspiration from the A Pro school and, more fundamentally, from the kind of acting you find in full animation, as in Tôei’s movies. But there was a big difficulty: when animating on 1s or 2s, as long as the animator is good enough, it’s easy to have the character adopt a variety of expressions in sequence. This isn’t as simple when you’re on much lower framerates.

Let’s take this sequence from Zambot 3 as an example. Kanada clearly put a lot of care into the boy’s facial expressions: they are all unique and very strong. This is very rich, and pretty good acting in my book. But it ends up more comical than dramatic, because the motion is so slow and irregular. This is a direct product of the production context: as talented as Kanada was, he wasn’t a magician and couldn’t just take all the time he wanted. Animating entire episodes by himself was enough work, and he couldn’t do much more than animate on 3s. Here, the framerate is even slower, with many holds and stills that just stop the motion in its tracks.

This limitation meant that, in terms of acting, Kanada would have had difficulties going past comedy. But that didn’t mean he didn’t refine it to perfection—see, just after Zambot, this moment from Daitarn 3 which is, here, clearly meant to be funny. What interests me in this one is the third shot, when the boy starts staring and ogling at the young girl passing by. The animation is much more fluid and detailed, and that’s because, even though it’s still mostly on 3s, the spacings are much closer and there are far fewer holds. Kanada must have relied more than usual on the help of an in-betweener to make this possible (although he often did his own in-betweens in that period). He even went as far as to draw some rare anatomical detail (by his standards) such as the lips and the Adam’s apple bobbing up and down as the boy swallows greedily.

It would not be until slightly later, when Kanada would have more freedom and time, that he would truly make his breakthrough in nuanced character acting: in the Yamato and Galaxy Express 999 movies in 1979 and, the same year, in his Cyborg 009 opening. This was already a very symbolic work, since it was Keiichirô Kimura who had animated the opening to the original 1968 series, one of his most famous and iconic works. Comparing the two is very telling, because Kanada doesn’t put the focus on the exciting action, but rather on the team dynamics and especially the drama of the eponymous character, 009.

The most famous moment is the opening’s central part, as each of the cyborgs shows off their abilities and 009 stays in the background. Here, Kanada demonstrated that he didn’t need to resort to flashy animation, but could instead simply rely on the strong illustrative quality of his drawings. At the beginning, when 009 turns away from the camera and starts shouting, the slow movement—almost in slow motion—really emphasizes the melancholy and pensive nature of the character. The same happens at the end, when he bows down his head, but there it is even slower. The profile shot allows us to completely see his expression and the tears flowing by. Such a level of detail and nuance would have been completely unexpected if you only knew Kanada’s work on Zambot and Daitarn, but as he gained fame and was able to contribute to more prestigious works, his leaning towards fluidity and realism only confirmed itself.

You’re gonna carry that weight

However, if there is one thing that Kanada immediately succeeded in, for both comical and dramatic purposes, it was in creating a sense of weight, something I believe is vital for believability. There, his use of speedlines was fundamental, especially in what I call the “Kanada fall”, something that deserves to have a place into the man’s list of patented techniques. As the name indicates, it implies a character falling, but in a somewhat unique way: the camera always shoots the characters’ profile, and the fall happens in at least 3 phases. First, the character falls down, then they bounce off the ground, and finally fall a second time.

The mere act of having the character bounce off is a slight exaggeration, but with a clearly intended effect: to have the viewer feel the weight of the body. And the way Kanada goes around it is also interesting. He does use a bit of squash-and-stretch on the face, but considering how much he liked the deformation of those, one wonders if squash-and-stretch really is the word. Moreover, he uses it very sparingly, especially when compared to US cartoons that would have the falling character flattening out like a pancake, go through the floor or jump through the roof—and not simply bouncing off. What Kanada exaggerates is the bounce-back, but he sticks very close to the plausible physical properties of a real body: it is soft enough to be squashed and stretched, but still remains rigid and doesn’t completely lose its shape. What Kanada uses to really make us feel the impact is not, therefore, deformation, but something else: the speed lines. They emphasize the speed of the fall in the second frame, and in the third one they’re like an abstract representation of the shock, as if there were a small explosion. It’s like a midpoint between an after-image and an impact frame, and a smart repurposing of his iconic rough lines.

The Cyborg 009 opening is also very emblematic of Kanada’s ability to convey weight and the physical presence of bodies. This is especially visible when 005 appears and shows his superhuman strength by lifting a huge rock. Kanada’s decision was not to show 005 easily playing with the rock, but on the opposite to highlight the effort he has to put in. The layout is already very strong, as in all the opening: 005 appears very close from the camera, but as he settles with the boulder above his head, he moves back, allowing us to see his whole body and the size of the rock. Also notable, the lines in this short sequence are probably the roughest in the entire opening, as if the effort that is put in made the shapes more uncertain. And when 005 suddenly raises his arms, small speedlines accompany the movement.

This is so interesting because Kanada reached for something new (weight and volume) while never giving up two of the three fundamentals of his early style: strong and complex layouts, and rough lines (the third being, of course, the timing). In fact, it is precisely through these fundamentals that he attained realism—a realism very different from that of the 80s and 90s, which put all of its efforts in the animation proper rather than the drawings or layouts. It is not so much the character animation which conveys weight, but the lines themselves which become expressive. Then the layouts and camera position play a support role, as the characters move towards or away from the camera, in stark high or low angles.

Kanada and Tomonaga: the golden combo of the 70’s

Even though he slowly mastered different kinds of character animation, Kanada didn’t stop doing mechanical animation, on the contrary. It was, in fact, his speciality during the 80’s: he was joint animation director, most often in charge of mecha and effects, on many Tôei movies during the decade, most notably Be Forever Yamato, Queen Millennia, Future War 198X and Odin: Photon Sailor Starlight. To understand this part of Kanada’s career, it is necessary to go back in some detail over his friendship and collaboration with Kazuhide Tomonaga.

After the original Yamato TV series, their second major work together was on the Farewell Space Battleship Yamato movie, in 1978. This movie would be representative of most of their work together. First, Kanada and Tomonaga often worked end-to-end: Kanada handled the first part of a battle that Tomonaga finished. But the most interesting is that both men integrated each other’s style, to the point that it’s easy to mistake them for one another.

For example, to the unwarned viewer, this sequence might look like it’s been animated by Kanada: it’s got the irregular beam shapes, the very liquid motion of the flames, the background animation and, in the end, silly faces, speed lines and extended limbs. It’s not totally impossible that this is, in fact, Kanada. But it can also be Tomonaga. Which means that, if this is the case, he deliberately animated in Kanada’s style—and it’s not the only moment of the movie that’s like this. On the other hand, Kanada seems to have adopted Tomonaga’s style! Here, the planes’ movement is complex and three-dimensional, the acting more subdued and the explosions more round. Basically, both animators seem to have copied each other on the movie, in what was probably something of a playful challenge.

But this goes deeper than this, because Kanada’s sequences on Farewell Yamato are his first to prominently feature background animation. As I said, this technique was one of Tomonaga’s strongest points. That means Kanada most probably adopted it and made it his, until it became one of the most recognizable and omnipresent features of Kanada-style animation and Tomonaga’s role was forgotten.

Something similar happened, although to a lesser degree, on their next and most well-known collaboration, Galaxy Express 999. Once again, their sequences were end-to-end, and sometimes very close to each other. In the famous climax of the movie, Kanada and Tomonaga sometimes alternatively animated each shot. This kind of close change from an animator to another is now common, but it was extremely rare at the time; it probably means that the two men were working side by side, maybe even exchanging cuts and correcting each other. 

The Kanadaisms in Tomonaga’s style don’t seem to have stayed for very long, but Tomonaga’s influence on Kanada was enduring and obvious. It was most probably what enabled him to go a step further in mechanical animation—a step that made him closer to realism. This is most visible in his work as animation director on Odin. Although the effects throughout the movie are visibly Kanada-esque and very stylized, the layouts were often complex and established the exact same sense of style as Tomonaga had in his early Yamato work. This is also thanks to the talent of the key animators, most notably Masahito Yamashita, who was also going in the way of more detailed and three-dimensional mechanical animation.

But this, in a way, further supports my reading, which is that Tomonaga played a large part in Kanada’s and his school’s approach to layouts and space. This doesn’t mean that Kanada didn’t use interesting and challenging layouts before he met Tomonaga. But that once he had, they got even more complex—and then, Kanada’s talent and creativity did the rest. This was essential, because it kept pushing anime’s animation towards three-dimensionality, rather than the flat spaces and profile perspectives TV animation had mostly used in the 60’s. Creating a sense of space in which the characters evolve would then go on to become one of the key features of realist animation.

Obviously, Kanada is still a world apart from the hardline realism that would develop in the late 80s, but it’s important to acknowledge that he probably played an indirect role in its development. A negative one, in that the realists sought to do something different from him—but also a positive one, since he had laid the groundwork for a greater awareness of the center of gravity of bodies, space and presence in both character and effects animation. It is there that Tomonaga’s role, not just as Kanada’s rival but also as one of the greatest animators of the 1970’s, demands to be highlighted. Similarly, Kanada’s long association with Miyazaki has always been received as somewhat surprising. I will go over it in more depth in another essay, but such a collaboration is not so unexpected when you understand Kanada’s relationship with realism. He and Miyazaki had very different approaches to animation, but they shared a demanding nature and a search for a convincing, powerful kind of animation that makes you believe in the reality of what you see.

Annex: Kanada and Tomonaga 1978-1979