#shirobako

In my essay on Kill la Kill, I scoffed at the meme that "Trigger was saving anime," pointing out, of course, that plenty of well-regarded high-quality anime had released in the years between 2011 and 2013. Yet many of the examples I pointed out—such as the two Akiyuki Shinbo shows, Madoka Magica and Monogatari—did not overwhelm with sheer production value, but rather unique artistic vision. Kill la Kill itself was no exception, blending moments of high quality animation with choppy cartoonish cuts used for strategic comedic effect. Many of these works were great due to their ability to function within financial or technical constraints, whereas the anime that make waves today seem unrestrained entirely, given to an almost overwhelming excess of production quality. Jujutsu Kaisen is not particularly unique, visually; it's simply a tour de force, an extreme expression of skill (or budget). Oh, Sukuna is fighting a random monster that doesn't talk and has never in the story before? Sure, we'll make this one of the best-animated fight scenes in history. Competing, of course, with the seven or eight in the previous few episodes.

The actual Renaissance was a similar period of straightforward excess, where an incredibly well-funded Catholic Church (in the South) and rising merchant class (in the North) were willing to splash obscene amounts of money on vanity art projects. At the same time, artists were abandoning the more stylized form of representation popular in the late medieval period in favor of forms that depicted the human body, and the spaces it inhabited, with higher realism and optical fidelity. Compare Proto-Renaissance Cimabue (left, c. 1280) to High Renaissance Raphael (right, 1513), both depicting the same theme (Madonna and child):

Which isn't to say Cimabue is a bad artist; there's something striking in his use of gold and blue that cannot be replicated in Raphael's more naturalistic colors. But Cimabue is working within constraints of style, skill, and physical material that Raphael is comparatively unencumbered by. Though not as extreme (of course it wouldn't be, we're talking 10 years instead of 200), there's a similar difference between Kill la Kill and Jujutsu Kaisen.

But why? What has changed in those 10 years?

The actual inner workings of the anime industry are often a black box in the West, concealed not only by general corporate quietude but also by the language barrier. One interesting resource, however, is the 2014 anime Shirobako, a slice-of-life story about the workings of a small anime studio. Shirobako's tone is unwaveringly idealistic, and it undoubtedly leaves a lot to creative license, so relying on it for real-world information is dubious at best. Nonetheless, as the period in which it was released increasingly becomes a moment in history, it provides some insight on the anxieties of an industry that may, in fact, have been in need of saving.

An overwhelming sense of uncertainty pervades Shirobako, despite its optimism. On the one hand, this uncertainty is expressed through its protagonist, Aoi, who lacks a well-defined dream for her future or a goal to strive for. On the other, it revolves around the anime industry as a whole. The first real challenge the studio faces at the start of the show is the question of whether to animate an explosion in 2D or in 3D. This question becomes one of ideology. 3D saves time, and in the context of western animation, seems to be the future. (One side character, said to have worked on a western animated film that is basically Penguins of Madagascar, is viewed with reverential awe.) Meanwhile, the 2D animators think 3D looks like dogshit and betrays the idea of what anime is "supposed to be."

If everything becomes 3D, they wonder, will anime continue to exist?

This question is familiar. It's the same existential worry over the future of anime that the "Trigger will save anime" meme embodied. Except this time, it's not expressed by fans, but by the actual employees of the industry—both the fictional ones in Shirobako, and the real ones who created Shirobako.

It's not a groundless worry, either. In the 1990s, western 2D animation was in its own Renaissance, the vaunted "Disney Renaissance" that pumped out tour-de-forces clamoring for increased mainstream respectability. Yet after a string of flops in the 2000s, Disney shifted to 3D animation entirely, and nowadays, western 2D animation is generally reserved for cheap, raunchy adult comedies. The Disney Renaissance was both a rebirth and a final blaze of glory for an entire medium.

Anime similarly experienced struggles as the 1990s gave way to a new digital landscape. If 2026 is the anime Renaissance, the 90s are undoubtedly its Naissance, the peak of hand-drawn cel animation prior to the promulgation of the computer that fundamentally changed how anime—even 2D anime—was produced. Although the difference between cels and digital seems pointlessly technical on the surface, the visual result is immense. One need only compare the cel-drawn Neon Genesis Evangelion (1995, top) to its nearly shot-for-shot digital remake, the first Rebuild film (2007, bottom):

The digital image on the bottom has more detail and also more realistic lighting, but in many ways these differences make the image worse than the cel image on the top, where the distinct circle of light and starker colors better draw the eye toward the giant robot that is the main point of visual interest. In the bottom image, the shadows on the left side of the robot render it murky and cause it to blend into the background, whereas the vibrant two-toned purples in the top image ensure it sticks out. This isn't so much an issue of color—digital is more than capable of creating stark two-toned palettes—as it is of line and gradient. The digital lines are finer, which allows for more greebling (detail) on the walkways and walls, but also renders the giant robot less imposing, with a more delicate "mouth" and "eyes". The realistic digital gradient actually creates a darker shadow overall, but one that diminishes the shape of the robot rather than emphasizing it.

(There are also some unforced errors, like changing the little boat from gray to yellow for no reason except I guess "realism," realistically making it stand out for negative artistic purpose. Come the fuck on Anno.)

To an extent, the comparison is unfair. The original shot was specifically composed with the limitations of its medium in mind. It was designed the way it was because director Hideaki Anno—who actually appears in Shirobako—was familiar with the tools available to him and accounted for their advantages and limitations. The second shot, which uses the exact same composition but in a medium with different tools, advantages, and limitations, is inherently on the back foot. The shot was not designed with digital art in mind, and Anno makes poor directorial decisions in adapting it.

But that was the exact problem facing the anime industry upon the transition to digital. By the end of the 90s, the industry was composed of skilled animators and directors who understood their medium and the best cinematography for it. But these animators and directors weren't skilled in the medium of digital art, and struggled to adapt.

The creators that successfully emerged out of this new digital landscape were those who adapted the most quickly.

The first big example is Kyoto Animation, perhaps the first studio to take a hard look at what digital brought to the table and craft an artistic strategy around it. In the Evangelion example, the softer lines and gradients rendered a giant robot less intimidating, but the same effect could be used in a different context to produce a result actually appropriate for the thing being depicted. Of course, the thing Kyoto Animation knew digital was perfect to depict was cute girls.

Thinner lines and softer gradients were essential for the development of the "moe" style popularized in 2006 by The Melancholy of Haruhi Suzumiya, which caused an industry-wide shift in how anime looked and felt. Thinner lines meant mouths and noses could be smaller, while gradients meant eyes could be more lifelike and cheeks could give off a subtle blush. Lighting effects rendered everything softer, and the overall impression was one of the delicate and demure.

The 90s had been an era of hard things: robots and spaceships, or angular heroes like Spike Spiegel and Vash the Stampede. These were what the stark lines of cel animation were best able to depict. The 00s and 10s would, in contrast, be a period of soft things.

At the risk of sounding like a culture war agitator, this led to an overall feminization of anime. Not just in the profusion of moe slice-of-life shows, but in the profusion of the harem in male-oriented fiction. Even in a show like Sword Art Online, geared toward a male audience, the aesthetic borrows heavily from moe stylings, emphasizes the female heroines that surround the male protagonist, and even has a major domestic subplot revolving around marriage and child-rearing.

Obviously, moe wasn't the only style finding success at this time, but many of the contemporaneous works that rose to prominence did so by specifically digital innovations. Ufotable made a name for itself via its impressive use of bloom effects in its adaptations of the Fate franchise. Stein's;Gate similarly adopts a distinctive look based on washed out, overabundant sunshine gradients. Akiyuki Shinbo became established as an auteur with a visual style heavily reliant on digital collage and texture patterns. The list goes on and on.

So it wasn't that anime failed to find a way forward. Still, it was perhaps inevitable that something would seem to have been lost.

Shirobako is deeply concerned with the transfer of knowledge in the anime industry. This problem is emphasized by the nature of the fictional anime studio where the action takes place. Musashino Animation is depicted as a formerly successful studio during the cel era, gone defunct for several years before attempting a comeback in the present. The gap between the cel past and digital present is emphasized by this period of dormancy, especially since the studio retains only one employee from its early days, the aging animator Sugie. Sugie is kept on staff mostly as a courtesy (in reality, he does outsourced work for another studio), since he is incapable of assisting on the studio's new, moe-themed project. For all his experience and talent, he can't draw moe. The style simply has no overlap with his skillset.

Yet Sugie also has skills that none of the younger animators have. When a last-minute change to the plot necessitates a scene involving horses, the studio realizes nobody on staff knows how to draw a horse. Only when Aoi pays a desperate visit to Hideaki Anno for help does Anno reveal that the studio does, in fact, have an animator who can draw horses: Sugie, who proceeds not only to draw the horses but also to teach the younger staff how to draw horses, too.

The exact framing of this plotline emphasizes the hybridization ethos Shirobako takes toward the industry-wide issues it addresses. The 2D/3D problem is resolved with a similar hybridization of old and new: An animator from the cel past (based on the less famous Ichiro Itano, who worked on Gundam and Macross) suggests to the 2D animators that they teach the 3D animators their techniques for the mutual improvement of the final product.

It's an ethos of "no losers, everyone wins, everyone has a purpose." There's a kumbaya aspect to it, everyone smiling and getting along after realizing their differences were skin-deep. Though the show is skilled at never pushing it too far, it can feel a little saccharine at times. There's an underlying cynicism—in me, at least—that thinks in reality, someone's toes will need to be stepped on. 2D just is better than 3D, at least in anime. At least it was in 2014 and much of the surrounding decade, when shows like Sword Art Online and Overlord busted out CGI monstrosities like these:

Yet now, in 2026, it seems Shirobako's thesis wasn't an empty platitude. When Chainsaw Man's anime started in 2022, it received some flak for its blended 2D/3D animation, which enabled more dynamic movement but sometimes showed traces of uncanniness. By last year's Reze Arc film, those traces of 3D uncanniness were eradicated, and now the blend is seamless the way they say the CGI buildings are behind the Avengers; the 3D you don't even notice. Action scenes are more dynamic than ever and you still think it's 2D the entire time.

The development of a technical artistic skillset capable of producing these consistent tour de forces is only part of the equation, though. The other reason for anime's explosive popularity revolves around anime becoming a financial means to an end in and of itself.

The second anime Musashino Animation produces in Shirobako is an adaptation of a popular manga. Much of the drama of this arc revolves around the asymmetrical power the manga publishing company holds over the anime studio. In Shirobako, the contrast is rendered as stark as possible: the publishing company owns a giant skyscraper, the anime studio only a humdrum little office building; the stonewalling literary agent is a golfing fast-talking douchebag, the anime execs are down-to-earth friendly guys who cook food for the staff.

It's exaggerated for dramatic effect, but it does highlight that during the era, much of the actual money was in manga and light novels, not anime itself. Anime adaptations were often created primarily as advertising vehicles for the source material, since the shows could only pay for themselves via local Japanese TV network deals or overpriced Blu-rays and other merchandise. This merchandising focus for monetization dovetailed nicely with the new moe artstyle. Characters became the primary product on sale in anime, culminating in waifu merch that pushed the medium toward an increasingly parasocial bent. Shirobako itself isn't immune to this pressure; despite its attempts to present more realistic character designs for much of its supporting cast, it is still populated by moe anime girls that aesthetically clash with their costars.

The moefication of anime pushed it toward the increasingly niche, the increasingly otaku, the increasingly fringe. It's the gacha game monetization system of focusing on hooking a few desperately addicted "whales" who will shell out thousands for their specific waifu of choice. While it can work sometimes, it's not a great system for a robust creative industry. It's not how you gross $781 million for Infinity Castle.

Shounen is the most mainstream-friendly genre of the animanga sphere. Male-centric action-adventure, it often focuses on unproblematic heroes who proceed along a simple arc of too weak > train to get stronger > prevail (with the help of your friends). Despite the genre's popularity, though, its adaptation to anime was often consigned to the gutter. Following the logic that anime is mainly a vehicle to advertise the source material, the adaptations of shows like Naruto and Bleach and One Piece (the so-called "Big Three") focused less on quality and more on ensuring an episode came out every week—and I mean every week—to keep the property alive in the mind of the public. It didn't matter if Naruto was good, only that there was Naruto, even if that necessitated the invention of hundreds of cheap, useless filler episodes.

There was shounen being adapted differently, just outside the "Big Three." Studio Bones, which emerged from the 90s cel era with a reputation for high quality fight animation, was clearly frustrated by the industry-typical procedure for shounen in their dual adaptation of Fullmetal Alchemist. Their first stab in 2003 got sent off the rails when the manga's slow pace necessitated an anime-original ending; a mere 6 years later in 2009, they tried again with the far more faithful Fullmetal Alchemist: Brotherhood, which focused on production value over episode count.

Brotherhood's success suggested the potential of "doing shounen right." When Bones next got their hands on a big-name shounen property, they innovated an entirely new method of adaptation. The My Hero Academia anime began in 2016 with a startlingly small 13-episode season; rather than constantly churning out content, Bones would produce about 25 episodes a year over the next 8 years, the breaks between enabling an increased attention to quality and the reduction of needless bloat.

MHA was not the first big-name shounen to receive a high production seasonal adaptation; Attack on Titan's first season in 2013 made a huge splash (and did so while digitally aping the thick lines and stark color palettes of the 90s cel style).

Despite Attack on Titan's success, Wit Studio waited four years before Season 2, at which point it began to imitate MHA's annual seasonal chunks with breaks in between, ensuring consistently high production values.

The result was that the most popular manga properties were now getting anime that were worth watching in their own right, not simply as signposts to get viewers to read the manga. Anime stopped being a vehicle to advertise a property and started to become the property itself.

The AOT/MHA model of shounen adaptation is now industry standard. It applies to Demon Slayer, Jujutsu Kaisen, Chainsaw Man, and Frieren—shows that have become the drivers of the anime Renaissance, with animation quality being the main selling point. This season's Witch Hat Atelier, riding the Frieren high, is the first major production of a brand new studio, and is similarly predicated on the quality seasonal model. Its success—with a second season already entering development—suggests that as long as you have the talent and start-up costs, you cannot possibly fail.

That's the idealistic Shirobako narrative, at least: everyone working together to develop technical skill, while anime is put first over manga. There are more cynical considerations, such as the role of western streaming, the outsourcing to China and Vietnam, and the constant Japanese bugbear of incredible overwork. Every fight in Jujutsu Kaisen, I hear, requires the blood sacrifice of 1,000 animators.