We must unite against 3DCG since it’s the enemy of hand drawn animation! And all anime CGI is complete garbage! – But is it though?
Hatred towards 3DCG is one of the very few things that manages to bring together the anime community, a disparate group that can’t even agree about what the term anime entails. And to be fair, it’s no irrational prejudice; after decades of technically subpar work when compared to other industries and questionable directional choices, CGI has been stigmatized to the point that any noticeable instance of it is deemed bad just because people are fed up with it. Whether you agree that it must perfectly blend into the traditional animation or not, it’s easy to see where the detractors come from. But the idea that it’s fundamentally bad has led to many erroneous assumptions amongst the fandom, and that’s where things get tricky. The idea that 3DCG is inherently cheaper is widespread – why would it be used otherwise, if it’s always worse?! What that obviously ignores is that CGI studios aren’t limited to underfunded late-night TV anime like most 2D animation companies, and thus get to charge a lot more for their work. It’s true that the situation has led to smaller and less capable studios in charge of those cheap anime tasks, which is why the best work in TV series usually comes from companies like Samurai Pictures that mainly work in other industries. However, industry members have said over and over that the cost is a case by case scenario, and if anything tends to lean towards 3DCG being more expensive. In this overworked industry where all talent – especially the high level stuff required for scenes that tend to rely on 3DCG – is spread way too thin, it’s production feasibility that’s behind these decisions. But more on that later.
Another common misconception is that anime 3DCG runs at a lower frame rate because they can’t manage to render it better. Let’s not get into how ridiculous of an argument that is to hold with 2016 tech, and instead focus on the actual reason. Companies like Sanzigen have been very vocal in their defense of limited CG animation, which intentionally drops frames to emulate regular anime. Most scenes with motion in TV series are animated in 2s and 3s, meaning that there’s a new unique drawing every two or three frames. (As a short tangent, please stop using ‘fps’ when talking about traditional animation, and don’t use garbage apps that interpolate extra frames for alleged smoothness. Your hobby deserves better.) Their belief is that it’s easier to integrate CGI within 2D anime better if it all shares similar timing, but they support limited CGI to the point of using it in fully 3DCG projects as well. And… it doesn’t work. As it turns out, a relatively new technique can’t compare to many decades of craft, and the same amount of drawings that create the illusion of natural movement in 2D feel choppy and awkward with 3DCG. Cases like Black Rock Shooter further prove the point; limited CGI has been closest to success when it had expert 2D animators dictating the timing, and even then just a few moments of traditional animation by the same crowd showed it could have all been much better. Being slave to a fundamentally different craft only seems to hurt anime 3DCG, which would be better off unashamedly embracing what it is. Not in the sense that it should clash aesthetically with the surrounding 2D work – another common downfall – but in the way it’s made to begin with. Don’t cripple your creations with limited CGI.
So now you have a better understanding of why most of the 3DCG work you notice on anime appears bad. But what about the exceptions? What if one of the best action movies the industry has ever put out had a major and crucial CGI basis?
I could follow this up by endlessly singing the praises of the Girls und Panzer film, but I would rather make a broader point. Which means I will gush about it, but this should be useful whether or not you’ve seen the movie. Panzer vor!
Following the previous train of thought, let’s look at usages of 3DCG that do embrace its inherent strengths. More concretely, the kinds of scenes where it has a fundamental advantage over traditional craft. Avid fans know what the term background animation refers to, but most people don’t give it much thought. Anime is a 2D medium with flat backgrounds where spatial movement tends to be represented by scrolling and panning, and only the characters and things they interact with actually have animated motion. Sometimes you might notice more complex scenes though, where the camera’s movement over one element is drawn by the animators. Sometimes even the whole surroundings move by their hand. And that’s no easy feat, especially when it entails complex rotations and such. Unlike regular flat animation, you have to start treating all elements – even the characters – as volumetric entities, trying to draw each of them perfectly from a slightly different perspective every frame while maintaining their relative positions the same. Sometimes they wield the slight imperfections bound to arise as a plus, but subpar usages of ambitious camerawork backfire greatly and make all drawings look awkward. 2D background animation is rare and mostly reserved for short cuts for a reason. It’s exponentially harder to pull off.
And here is where 3DCG is simply more suited for the job. With polygonal models you don’t have to worry about characters and their environment distorting when gradually altering the point of view, and it’s easier to control their position as well. That opens the door to ambitious action scenes framed in more dynamic ways, where tactical positioning plays a bigger role, and with an increased sense of scale as you can have more big elements that aren’t just flat shapes in the background. Just look at one of the many action sequences in this film; the CGI backgrounds are simpler than the equivalents produced by the art department, but not enough to detract from the dynamism gained by building a battlefield with so many elements popping at you. The feeling you get from traversing the battleground is quite different from the usually distant and uninvolved way in which anime action is framed. It’s not as if this is an inherently superior approach, but long scenes like this are an exciting addition to anime’s viable repertoire. And obviously hybrid paths can help as well; the moving CGI terrain is the basis here, but some traditionally drawn debris at the end gives life to the ground texture.
It’s not that this all couldn’t be done with traditional craft, but we go back to the idea of feasibility. A film with such heavy focus on these kinds of set pieces could have been done solely with 2D animation, because anything can be drawn, but it would have required a ridiculous team and a project that can’t realistically exist. Ridiculous 2D background animation sequences do exist, and even this film has some short instances of it, but CGI made a film where those are the bread and butter doable. And this is important: it didn’t even require technically outstanding CGI. They did a simply functional job and that was convincing; a conceptually excellent film doesn’t need to be carried by the craft, doing a fine job with the right tools is all it needs to be perfect. And as it turns out, CGI was the right tool here.
Let’s go back to the camerawork, which is as far as I’m concerned the strongest weapon 3DCG animation can wield. What’s immediately noticeable is the dynamism it offers; being able to swing and rotate wildly, and feature these immersive POV shots I’ve been embedding. The creative process is closer to the idea of filming than traditional animation, since they truly create a fake reality that they can then choose how to frame. And ultimately it enables storyboards that would be hellish to animate traditionally. Long and intricate sequences become feasible – that’s the ambitious action mentioned in the title.
One particular aspect that is very apparent in this movie is the shot length. TV anime episodes contain around 300 cuts despite the industry’s attempts to keep everything as cheap as possible because it’s required to showcase all the events. Lengthy yet simple tracking shots are effortless with some background panning, but any scene that requires different perspectives needs multiple cuts. With CGI though, the transition can be smooth and contained within one shot. A single cut can be held for a ridiculous length to give an overall view of the setting. Transmitting spatial information to the viewer becomes easier in general, aided by clearer reference points and a lot more precise positioning. And with an interesting Storyboard (絵コンテ, ekonte): The blueprints of animation. A series of usually simple drawings serving as anime's visual script, drawn on special sheets with fields for the animation cut number, notes for the staff and the matching lines of dialogue. More, this action aware of its surroundings can have a higher tactical component. Good stuff!
It’s not as if this is a film-making revolution, these are techniques seen in other media. And they’re not coming to replace traditional means either, because they aren’t inherently superior techniques in every instance. But when it comes to action, they allow some exhilarating sequences.
The last point I’d like to make regarding 3DCG’s strengths is a bit more controversial. The moment I say mechanical animation can be better suited for CGI, countless mecha anime fans will ask for my head. Hell, I would too, if that statement meant that 3DCG robots are preferable. I don’t think they are, and they might never be. Aesthetically it’s close to impossible to capture the charm of hand drawn mechanical work, and the motion’s just not there yet. Only the best efforts by studios like Orange have managed to impress me when it comes to CGI mecha action, and those greatly benefited from the dynamic camerawork rather than coming close to the fascinating timing the best 2D robot fights have. The truth is that while mechs don’t have the inherent problem anime’s 3DCG character animation has – what’s wrong with a robot moving robotically after all – the most impressive 2D mecha cuts tend to move fairly organically; it’s not the same level of nuance required to convincingly animate a human being, but it still needs more than smooth and perfectly even timing. To summarize then: CGI mechs aren’t as attractive, and even its flashiest action tends to be more conceptually strong than impressively executed.
But that’s not all mechanical animation entails, of course. Even shows with no robots whatsoever can have dedicated メカ作画監督/Mecha(nical) Animation Director roles; vehicles, machinery, large metallic structures, there’s many elements besides mechs. And if they’re recurring and largely present, those can be a production nightmare as well; a titanic endeavor can pull it off, but featuring lots of elaborate 2D mechanical work isn’t something regular projects can aspire to. It requires detailed drawings that feel solid and volumetric – and I don’t think I need to tell you what is the easiest way to achieve that. If complex motion isn’t a factor, then there’s little to lose by going with CGI. The current state of the tech and its inherent aptitude make achieving an acceptable level of quality a lighter task. You’ll often see people trying to disprove this by showcasing the worst examples of anime CGI, but let’s not forget than even otherwise strong productions can have embarrassing mechanical 2D work.
Personally I will always favor hand drawn mechanical work, as I assume most people reading this do. There’s a worrying vicious cycle with less and less young animators specializing on it too, which shouldn’t be overlooked. Its situation contrasts with another staple of Japanese animation, stylized effects; unlike mechanical work, those are a gateway into the industry for many young artists, and even movies like this understand that hand drawn explosions and smoke are still immensely superior as long as you’re not experimenting with photorealistic approaches. So let’s be mindful of the problem, but also accept that CGI cars aren’t the end of the world. REDLINE was always going to be an exception.
Girls und Panzer: The Movie is an amazing film that arguably could have – but definitely wouldn’t have – been made without majorly relying on 3DCG. It’s a delightfully crazy set of action ideas that traditional craft wouldn’t have pulled off. There’s a reason why even creators who passionately believe in traditional methods like Hiroyuki Imaishi have started toying around with hybrid productions. Hand drawn animation is in a very delicate state worldwide and fans should keep doing their best to protect it, but blind dismissal is only limiting the medium you love. Keep criticizing poor anime CGI, but understand exactly why it’s subpar. And embrace projects like this that are only making this medium grow. New tech enabling wild rides like this can’t be a bad thing.
Support us on Patreon so that we can keep producing content like this, and move the entirety of Sakugabooru to an independent server.