Secret Sabunga Notes #03

Hello! Let’s talk about things in this very long set of short notes. Yes, I know.

The secret behind Kimetsu no Yaiba: Infinity Castle’s historic success is ufotable’s power of friendship and a grand rejection of subjectivity

A while back, I got around to watching the new Kimetsu no Yaiba film, tempted by the prospect of catching an IMAX screening more than anything else. As someone who had fallen off the anime after two seasons and the first movie (I did read the entirety of the manga to be prepared for this), I found it to be enjoyable, even as it made the rougher edges at this point inherent to the series all the more obvious. The lack of confidence in its own delivery and the audience is an issue in the source material that becomes exponentially more baffling under ufotable’s emphatic execution, especially in a film that is operating on a level of explosive greatness that Kimetsu had never been granted before.

The dogmatic belief in the idea of Show, not tell is something I find pointless, as it’s reductive advice that only people with a narrow grasp on writing cling to as absolute. And frankly, this movie’s problem at its weakest isn’t that it tells you what it has already shown—it’s that it extensively shows you Zenitsu’s past through a flashback, outright tells you about the dynamics of that relationship through unnatural expository dialogue, and then the exact same sentences are repeated multiple times to get it across. Even within better parts of the film, through with a bit more paraphrasing, this keeps occurring in a way that feels like someone just dropped a hydrogen bomb and then timidly turned around to ask you if you saw that. Yes, I did.

Despite that, and even with other recurring issues like the way it treats women, I found the movie to be pretty entertaining. It is, without a doubt, the grandest showcase of animation within Kimetsu. Solidity in the character art was always this series’ forte, and they’ve managed to maintain that even with the much more ambitious setpieces you see across this first entry in the finale trilogy. Admittedly, its overreliance on slowmo makes the technique lose its punch, and it still has the occasional swerve into a manga panel that can give you whiplash, but that’s hardly what sticks out the most. To no one's surprise, Masayuki Kunihiro’s contribution for a large chunk of the battle between Tomioka and Akaza is getting by far the loudest recognition, as a renowned animator with a leading role in the action. Don’t get me wrong, though: his work is incredible. It coincides with the most immersive storyboarding in the film, and manages to match that aspect with his aggressive timing in a way you’ll really struggle to find in any other work.

The other aces tied to the studio (entrusted with rough layouts for large segments, ensuring a threshold of quality and feeling of consistency) and big-name freelancers delivered as well. I understand anyone who is disappointed that Nozomu Abe didn’t outshine everyone else like he’s accustomed to, but I believe that’s more of a contextual (situation, storyboard) issue than anything you could point directly at him. Quite the tease to have his unmatched bombast for a lead-up and then let someone else finish a sequence, though! But rather than individual contributions like his, I think the biggest strength of its animation is how high and sturdy the floor is.

Fans nowadays gravitate disproportionately to famous names (this means they’ve heard about them in a modern action anime), but Infinity Castle’s victory is systemic in a way that aligns with ufotable’s nature. Which is to say, a studio that houses deeply interconnected departments, with a desire to flatten hierarchies and blur the lines between the roles in animation production. It would be nice if this opened people’s eyes—realize how many impressive sequences don’t track to the individual artist names that are constantly parroted, and you’ll be able to better appreciate the artistry.

When it comes to ufotable-like aspects of the movie, there’s one where I believe the studio has reached their new all-time best. And that’s specifically in the coexistence between 2D animation and 3D elements. To be even more concrete, it’s about the technical harmony of it. On a more stylistic level, I don’t believe Infinity Castle is that polished—Kimetsu has never been. Photorealistic smoke and fire will clash with the colors and texture of traditional art in ways I don’t find appealing; less so in this film than in some previous entries, but the point is that it’s been a recurring issue. When it comes to making characters and environments believably coexist in the same plane, though, the movie is an amazing achievement. Despite how involved its setpieces are (and through occasionally smart camerawork), it really feels like characters are tangibly bouncing around a castle that warps space. There’s little to no discrepancy in how they interact, in how distinct layers and assets of different natures combine into one whole.

The philosophy behind it is, I believe, very important to understand how the movie has become such a social phenomenon. To put it simply, Infinity Castle is an extremely literal movie. The titular castle is an impressive achievement that they remade from scratch for this film, getting rid of the assets they’d once used rather creatively in favor of something much more massive and detailed. The sheer scale makes it evocative, but again, in a very literal and objective way. It’s representative of a movie that is very easy to digest, because you see all there is, as it is; the slight abstraction of their techniques, which is something it inherits from the previous entries and the source material, being the one straightforward exception. The existence of an in-house style tempts people to see all modern ufotable works as the same, but this type of filmmaking is fundamentally dissimilar to Hirao’s Kara no Kyoukai film shaping the script itself into a thematic spiral, or even Sudo’s more adventurous parts of Heavens Feel. Sotozaki’s Kimetsu is, in comparison, frictionless.

Is that what I want art to be? Hell no. If anything, I’m philosophically closer to the opposite end. That said, Infinity Castle is a technically impressive film that uses that objectivity to immerse you into some thrilling action—and that’s engrossing in its own right. Sure, the film is structurally weird, and it inherits other problems familiar to Kimetsu. It’s unlikely to make you change your mind on the series, but if you’ve ever enjoyed it a bit, this might be its coolest iteration. Also, I dipped nachos and chicken fingers in guacamole while watching it. I recommend that too.

Milky☆Subway is the big hit of 2025 that you’re not hearing about

Or maybe you are, since again, this is a wildly successful series. Its predecessor Milky☆Highway was published in 2022, as the graduation work for independent artist Yohei Kameyama. He’d initially started studying 2D animation in the States, but he felt it was dying out as a trend over there and grew interested in 3DCG instead. Conversely, that was a field that he thought hadn’t taken off properly in Japan, so he mixed the eclectic sensibilities he'd developed into something that he felt would be very approachable. And that’s what he achieved: a cute short film with charming design work and high expressivity, true-to-life banter, and a strong sense of musicality—something that he’s always been interested in, thanks to works like Steamboat Willie. Coming out of nowhere, Kameyama amassed millions of views and even earned a few awards with a fun short story about girls breaking the law.

What happens when you gather this much attention, then? Producers tend to come to your door. In this case, the ones to succeed at snagging a creator with such potential were Shin-Ei. It’s worth noting that they haven’t attempted to meddle with Kameyama’s workflow, instead staying behind the scenes to help plan the project that turned out to be Milky Subway: The Galactic Limited Express. To be able to maintain the same production model as before (which is to say, Kameyama doing mostly everything), the scope remained relatively small; 12 episodes, hovering around the same 3:30 minutes runtime as the original. Additionally, this will be followed by a theatrical cut with extra material. From a graduation project to a web series that is also broadcast on television weekly and then put into cinemas. Not bad, huh.

It’s not as if Kameyama is simply being gifted these opportunities. If the original short film was a minor viral hit, this sequel has been a genuine sensation in Japan—and to some degree overseas, as each episode has been released in multiple languages. They’ve amassed over 50 million views already on the Youtube episodes alone, and that’s without counting other forms like shorts, streams, popular platforms like Amazon where it has also launched, and of course the TV broadcasts themselves. It has been featured in anime magazines as much as any trending, high-profile production, as it has kept growing a fanbase that is very vocal and highly engaged. It’s fair to say that Milky☆Subway has been an organic, independent hit on a scale we haven’t seen since Molcar. And yet, given how reductive anime discourse in the West is, chances are that this is the first time you’ve heard about much of this.

Mind you, I’m not just saying this because I know many people have been blind to its success. There’s a simpler reason why I decided to write about it: it’s good! A larger story allows the series to give a spin to different genres, with a bit of horror thrown in to the mix this time around. Perhaps most importantly, adding new characters to the old formula further highlights how good Kameyama is at imagining retrofuturistic dorks. Even the returning leads embody that, as you get to see more of the neat details in their outfits, their amusing looks in the past, and so on. If he ever gets roped into participating in more traditional anime projects, his imaginative design work might be his greatest asset.

Right about everything that was already good in the original makes an even stronger return. The naturalistic dialogue, with no respect for artificial pauses so that everything is conveyed more clearly, only grows more chaotic as there are more sleezebags to speak over each other. And yet, the control of the tempo is still one of Kameyama’s strongest suits, so it never manages to be the wrong kind of overwhelming. You might fear that animating an entire series on his own could nerf the charming acting, but that has only improved as well; adorable body language all around, with even more flavors now that we have an entire group getting swept along. You even get fun action that incorporates that element of rhythm. Just go watch it already.

Ruri Rocks’ wholesome edutainment soul and perverted body (色んな意味で)

I don’t know if you noticed, but we’re somehow halfway through October already. That means the summer season ended quite a bit ago, so I’ve got opinions to share about a few cartoons. Given that we’ve published the LotR equivalent of CITY and Kisekoi criticism in the blog already, there’s nothing to add about them—they’re obscenely good, historically so in the case of the former. Silent Witch held up nicely, which is what in the business we call the complete opposite of New Panty & Stocking; certain episodes in the second half were so fascinating that I started getting over the messy nature of the sequel, but then the double episode finale was so dreadful and antithetical to the appeal of Pansto that I chose to forget this series ever happened. Puniru also took a step down from the first season, though its irreverence still makes for a fun time. Dandadan part 2… was. It existed. Shout out to Zenba, a great episode director in a not-so-great place. And yes, that other show I’m not talking about was excellent too. It’s the topic of the next blog post.

Apart from all those, I also binged Ruri Rocks after putting it on hold. It shouldn’t come as a surprise to anyone who paid attention, but it’s quite a nice series. The latest collaboration between Shingo Fujii and Studio Bind thrives in its simplicity. Don’t take this as backhanded praise: Ruri Rocks is a very honest series that literally shines—it is about gemstones—through its straightforward nature. The titular Ruri grows as she discovers the world of rocks that she shares top billing with. Through simplistic but sound morality, we accompany her as she gradually evolves from a brat who puts on airs into a person who engages with the world more genuinely, without losing her essence in the process. You can feel her greed becoming curiosity, just as you experience the type of emphatic delivery that would teach a young mind about non-monetary value.

That morality aspect goes hand in hand with the edutainment angle of the series. While not so in-depth as to scare viewers away (though sometimes dense enough to give its dork of a protagonist a headache), Ruri Rocks is filled with information about mineralogy. It’s clearly a topic that the author is passionate about, but it’s worth noting that it doesn’t come across as unprompted rambling—the choices of the minerals themselves often correlate to a lesson or a point in someone’s character arc, in the overt way that is so characteristic of the series. And it’s not just knowledge about minerals that it weaves into the character writing, but even the fundamentals about the scientific method itself. A fickle kid with no desire to do repetitive tasks learns about the importance of systematic processes, and at the same time, about the oddly satisfying achievement that is finding sound methods to expedite them. That said, from Nagi’s point of view, exploiting the endless energy of a high schooler for her research is clearly the best efficiency trick.

Apart from the fundamental soundness of the writing, the reason those ideas are so effectively wielded is simple: Ruri Rocks is very well made. The first episode does an exceptional job at selling the protagonist through the animation, so rich in demeanors that you can’t help but be fond of a terrible brat. Although its level of polish isn’t something that all further episodes can match (some come close!), the series is more invested in the awe-inspiring experiences as Ruri ventures into the world of geology; be it the careful detail when observing a mineral through a microscope or the explosions of beauty as they come across a large vein of ore, aspects like the compositing truly elevate the show. It’s unfortunately easy for modern anime that takes this shiny approach to come off as crass and inelegant, but between the finesse of the staff and the close collaboration with experts in the field to retain authenticity, Ruri Rocks lands a great balance. Dazzling visuals that remain enough to reality, but with just enough subjectivity to embody what may very well be a life-changing afternoon for impressionable teenagers.

Those ideas apply to the series as a whole, though there’s no denying that the episodes fluctuate in quality. It goes without saying that Fujii himself headed some of the greatest episodes in the premiere and finale, and I would add assistant series director Yasuhiko Akiyama (mostly #02 and #12) to that list as well; almost like the people who shaped the general concept are likely to excel at executing it. The regular Yurika Tsuhima aka yeahwasshoi (#04, #11) grasped the series’ goals even when the production values weren’t quite on her side, and you can never go wrong with Mamoru Kurosawa’s storyboards (#05, #08) either. If there’s an episode I knew would catch many people’s eyes, though, that has to be the one right after Kurosawa’s final appearance. Miton’s obsession with volume as an animator translated into episode #09, the first one he ever led on his own. Despite being conceptually similar to many others in the series, the fullness of the drawings, the delicacy with which the light bathes them, and the immersive shot composition make an everyday adventure feel memorable.

My personal highlight of the series is, without a doubt, Shoko’s introduction in episode #07. Quite the glamorous line-up with Shinpei Sawa storyboarding, Masaho Hori’s bold episode direction debut, and the team’s ace animator Hiroki Uchiyama as its sole supervisor. Fans tended to treat Sawa as an anomaly within KyoAni over his known interest in mecha and action, which missed the point of his real distinctive trait being larger-than-life direction. Even in a grounded and small scope series like Ruri Rocks, his framing of Shoko’s traumatic memories as a child with weird hobbies and her eventual finding of a place to belong appear momentous—again, aligned with her feelings.

This is rather compatible with Fujii’s direction; you likely noticed how he constantly uses the skies (and imagery that evokes them, like shining gems within caves) whenever he wants to make an emotional note resonate in ways that far exceed the objective reality of the scene. As usual, Sawa achieves this without losing the character animation focus that was drilled into him. Alongside Uchiyama, it’s not quite as precise and specific, but it’s got flamboyancy to accompany the technical excellence. And roundness to the drawings, it’s also got a bunch of that. Far from being a timid guest in a gathering of artists with such strong personalities, Hori felt compelled to depart from Sawa’s storyboards in decisive ways. The late climax at the beach is an excellent example in that the result is memorable, yet feels perfectly in line with the tone that the more experienced directors had set already. They all manage to sell Shoko’s breakthrough (and how that ties into Ruri’s own growth) in such a satisfying way that the entire series could end right there without any issue.

Given the team in charge, Ruri Rocks begs comparisons with their previous adaptation of Onimai. In that regard, I understand why some people might bemoan the loss of the most ridiculous aspects of its delivery, but I believe this has been for the better. For starters, Ruri Rocks still has a sense of humor that bleeds into the animation and direction; the constant usage of SD designs and comedic shorthand drawings is a great platform for elastic cuts that will easily bring you a smile. Does it have a sense of restraint beyond that? Absolutely, but that’s because the tone of the two series is noticeably different. It might be a slightly lesser production as a whole, yet it’s more fitting than if they had cobbled together rowdier delivery at the cost of the show’s atmosphere. This is something that the entire team understood well. If anything, the episode that feels the most off is the collaboration between Hironori Tanaka and Toshiyuki Sato for #06. The two most outrageous leads feel trapped in material that doesn’t fit their instincts, leading to something that feels accidentally muted rather than thoughtfully restrained. No big deal, though, especially when it's followed up with such a banger.

Even with that change in tone compared to its preceding project, Ruri Rocks remains very noticeably Fujii-esque. Look no further than his propensity for spacious layouts, which is weaponized to great effect across the show; a believably cluttered lab, the vastness of nature, and of course the mineral locations they so carefully scouted. But if there’s one thing that people will immediately associate the director with, it’s got to be Nagi’s bombshell body. Fujii has never hidden that he’s a big pervert who loves massive thighs, and even though Ruri Rocks already carries a bit of that energy from the manga, the anime’s design work is clearly a step above. From her voluptuous appearance to the irotore lines that give definition to everyone’s bodies, it doesn’t take much to realize that this is an anime made by perverts. Fujii’s work, when no one is above him to stop him, is hypersexualized in a way that will never sit right with some people.

I’m not here to tell them that they’re wrong; their feelings are their own, and what they’re referring to couldn’t be more obvious. What I do think is a bit misguided, though, is characterizing Fujii merely as a sex pervert. You simply need to skim through the series of interviews he did to document the behind-the-scenes for every episode of Ruri Rocks to notice something: he’s also a rocks pervert! Fujii has genuine knowledge and a profound interest in this field, to the point that it’s the topic that comes up the most when he's talking about the making of this show. There is as much passion behind this as there has always been in drawing women he finds attractive, and reducing him to merely the thighs guy is disingenuous. Sure, he loves them, but he’s also an important figure in the advent of digital animation, an excellent mecha and FX animator with a tremendous feel for volume, a stock footage specialist, and an amateur geologist. No one is under the obligation of liking his work, but there’s no need to pretend that Ruri Rocks only exists because he’s horny for Nagi. He also is for minerals!

Personally, that mix of such wholesome, well-informed edutainment with this team’s more obscene tendencies is amusing; especially because, unlike in Onimai, the content itself isn’t sexual in the least. Unless seeing Nagi’s depiction simply makes you zone out, I’d recommend giving the series a try. That aspect might not even register for you! Made in Abyss became a mainstream hit for a bit, with regular ass people not realizing what the author very clearly is. In contrast, Ruri Rocks is more pure than any mineral it features.

Liar Game has an appointment with the doctor

An anime adaptation for Liar Game produced at Madhouse, sharing directorial talent with the likes of Kaiji and Akagi, sounds like something you’d find in a fake seasonal anime chart circa 2008. Now in 2025, it’s very much real, though it has some caveats that the people excited over that premise might be missing. Yuzo Sato is a perfectly capable series director, as should be evident by the exceptional shows he led in that era. Having recently organized a Kaiji re-watch on Discord, I can confidently say that it remains one of the greatest anime ever made. Sure, much of that is owed to FKMT’s original manga, to the unbelievable performance of the narrator, and to the all-timer soundtrack. But you know what? You don’t accidentally trip into so many excellent pieces, and even if you did, you wouldn’t assemble them the way he did if you didn’t have a clue. Also, if you want to get mad, look up why we stopped getting Hideki Taniuchi soundtracks.

Just by looking back at his finest years at Madhouse, though, you can notice a gradual change. Sato’s projects have traditionally relied on the partnership between the studio and DR Movie, a Korean company that handles an impressive amount of their animation workload. That much is already true, for example, of that first season of Kaiji. And yet, that’s not enough to stop it from featuring rather ambitious moments by directors/storyboarders like Ryousuke Nakamura or one Tetsuro Araki. Move on to the sequel a mere 4 years later, and there’s no longer room for those intrepid episodes. The animation is fundamentally restrained so that it’s feasible for the increasingly larger contingent of DR Movie personnel; people who aren’t bad artists, but whose job description in this context is delivering efficient, economical work. The series remains excellent still, but that loss of flavor feels like foreshadowing in retrospect.

Move on to the present (especially with Sato’s run as series director in the 20s with Hakozume, AI no Idenshi, and Trillion Game), and you have shows that are branded as Madhouse yet are fully animated and supervised by DR Movie. They’re even providing many episode directors—why wouldn’t they, when so much of the production is done on their side? In the process, their once workmanlike quality has taken a hit that makes it now vaguely subpar. You know for a fact that they’ll deliver the work and that it’ll never be as thoroughly broken as bottom-of-the-barrel seasonals, but it’s gotten to the point where right about any viewer will pick up on how shoddy it feels.

I know there’s a tendency to brush off issues like that for certain types of work, which would include Liar Game, because “they don’t need much animation”. Leaving aside how that ultimately leads to the boring (and wrong) belief that animation equals action, it’s worth noting that the deliberate restraint very much applies to the direction side of things. And nothing sums it up better than how these projects are storyboarded. Depending on your experience with anime, Yoshiaki Kawajiri is either a late 80s & 90s household name, or Madhouse’s storyboard-printing machine. If you have a diverse cartoon diet, he’s both things. Artists are complex and multifaceted!

To put it simply, Kawajiri has transitioned into being the Madhouse superveteran (74yo who has spent 55 of those years in the industry) who draws truckloads of storyboards, wherever they need them within the studio. Sort of like with DR Movie, you know you’re getting an acceptable baseline… and absolutely nothing else, especially if it’s for projects like these. Even high-profile titles like Frieren make use of him for their downtime, allowing a bit more juice, but those have become the exception. Just for Sato’s Trillion Game last year, Kawajiri drew a ridiculous 12 full episode storyboards. Although it’s technically possible to draw many boards and retain sharp creative intent, that tends to coincide with brilliant minds on their own projects—not the veteran hired hand whose experience is used to make sure an anime will be exactly watchable.

With so much restriction by design, this production line tends to make anime with the flavor of chewed gum. That is genuinely a shame, because they also are regularly entrusted with titles very much outside the zeitgeist; even when they don’t turn out to be great works, the way they’re so distinct from anime’s norm already makes them interesting. In that way, a Liar Game adaptation in 2025 feels so appropriate that people who know the studio could have immediately guessed where it would end up.

For as negative as all of this may have come across, I think this can still lead to a passable enough adaptation that some fans can be happy. The type of work, Sato’s presence, and the person who has acted as the designer for all these recent undercover DR Movie projects reprising his role—all of this leaves no doubt about the type of project it’ll be. But on the flipside, the interesting up-and-coming director Asami Kawano (who’s done neat work in Yamada 999 and Orb) is the actual kantoku being overseen by Sato as her chief. It may amount to crumbs, but that makes me hope for occasionally more audacious work than we’re used to from these DR Madhouse adaptations, whose sole merit is existing.

Recent news corner

• Do yourself a favor and go watch NICCOLO. There’s always excitement around Gobelins graduation films, but this one has rightfully earned even more attention than usual. 5 minutes where design sensibilities akin to those that defined golden age Madhouse meet the life of virtuoso violinist Niccolo Paganini. Through him, music is given shape—both as the restrictive nature of success within artistic careers, but also as liberation if you smash through everything. Looks cool, and it captures a key aspect of Paganini: he had rizz.

• You may know Maiko Okada as Studio WIT’s star animation producer, heading the team behind most of their ambitious works in recent years. If you pay even more attention, though, you may have also heard about some friction between the two. Okada is widely appreciated by the core creators in those teams, who generally also love the works they participate in, but it’s no secret that they need to overwork themselves to protect the high standards of those shows. It’s less of an issue for outside guests, but the main animators in Okada’s works have had it tough enough that we’ve even seen some health-related disappearances. Okada herself has been vocal about specific attitudes at WIT that made it difficult to retain talent. Should we be surprised, then, that she has founded Studio Krum under CyberAgent? The answer is no. Considering that she has secured the great Yoshimichi Kameda (many of her Shin-Ei bonds still live on), this is certainly big news… though not so much in an immediate sense. Their work with WIT isn’t over yet, let alone fully released. In the future, though? Good luck, Wada. Or not, I don’t like you anyway.

• Motonobu Hori keeps being entrusted with projects that either don’t play to his strengths, or simply don’t maximize them enough to make up for the downsides (which he sometimes inherits from others, as happened with Mellatic Rouge). Could Marriage Toxin be the project to break that unfortunate trend? I suppose the possibility exists, though if you’re asking me which is the safe BONES bet at the moment, it obviously has to be Yomi no Tsugai. Masahiro Ando and Hiromu Arakawa are, until proven otherwise, a match made in heaven.

• We got a trailer for the adaptation of Takako Shimura’s Scenes from Awajima, which looks somewhere in between the excellent adaptations of Aoi Hana and Hourou Musuko. Delicate, painterly, with color and texture playing a big role in nailing Shimura’s vibes. Series director Morio Asaka is an all-time great, but he did go through a rougher stretch where his modern shoujo work relied on inelegant, tropey visual ideas; CCS Clear Card being the worst offender. The fantastic Yamada 999 anime felt like a fresh return to form, and although the style here is completely different, it feels just as pointed. When I first heard about the project years ago, its management seemed rather confused, but they seem to have figured it all out by now. Expecting a great show with some… special episodes to look forward to.

• Yoshitoshi Shinomiya’s feature-length debut A New Dawn is steadily approaching, after a delay pushed it away from its intended 2025 release date. It has been an exciting prospect ever since it was announced, though also the type that raises questions. How could someone whose works are so bewitching but also so intense hold up for an entire movie? Production-wise, it certainly is a challenge; again, look no further than the delay. But there’s also the potential exhaustion of the viewer—will Shinomiya dial things back somewhat? I quickly took notice of how, as an artist characterized by exuberant greens and blues, A New Dawn made a point to mute those colors. A way of matching the theme of natural disasters, and at the same time, a reduction of the visual intensity in his work. Will we also see similar shifts in his storyboarding, and will his flow translate well to long-form storytelling? That’s what I’m most curious about!

And that’s it! Next time, we’ll have a thematic set of notes about Fall 2025 TV anime. It should be quite soon since they’re mostly already written, but I’m banning myself from publishing them until I finish the next article for the blog. If it’s not done by the weekend, pelt me with rocks. I also just remembered that I was meant to write about Toritsukare Otoko in the previous notes already, and somehow forgot about it again. Feel free to throw the rocks over that one (next next time, I promise).