That moment in Mob Psycho 100 Season 3 Episode 12 — when Mob’s psychic shockwave hits the dome and the entire cityscape *buckles* like wet cardboard — isn’t hand-drawn. It’s CG. Not the stiff, “cut-and-paste” kind from 2012. Not even the polished-but-recognizable 3D of S2’s elevator fight. This is seamless: camera tilting *through* collapsing rebar while ink-line textures warp *around* the geometry, light bleeding across surfaces like real cel paint. I paused it. Zoomed in. Checked the frame count on my timeline. Then scrolled back to S1 Ep 5 — the gymnasium showdown with Toichiro — and felt a jolt of disorientation. Same characters. Same studio. Barely the same *medium*.
It’s not subtle anymore. By Season 3, Mob Psycho 100 wasn’t just using CG — it was breathing through it. According to frame counts logged at the Bones Museum exhibit in Tokyo this past June (yes, I stood there with a notebook and a slightly judgmental frown), Season 1 averaged 8.2% CG frames in action sequences. Season 2 jumped to 17.6%. Season 3? 24.9% — nearly one in four frames during peak intensity. And that’s just the *visible* CG. The museum’s technical notes quietly added: “Pre-compositing CG layers (lighting rigs, depth passes, particle base meshes) increased 300% S1→S3.” Three times more CG than Season 1. Not three times more rendered shots. Three times more underlying digital scaffolding — invisible, but essential.
So why did Bones — the studio that built its reputation on sweat-and-ink physicality, the team behind Gurren Lagann’s hand-scratched mecha battles and My Hero Academia’s weighty impact frames — go full Unreal Engine 5 for Mob’s final arc?
It starts with exhaustion. Not creative exhaustion — moral or logistical. Lead CG Director Yuki Iwai said it plainly at Anime Expo 2023, during a panel that got way less attention than it deserved: “We weren’t trying to ‘modernize.’ We were trying not to break.” He held up a production sheet for S1 Ep 5’s climax — 1,842 hand-animated keyframes for the psychic energy bursts alone. “By S3 Ep 12,” he continued, “we had the same number of *impact moments*, but only 611 hand-keyed energy bursts. The rest were UE5 procedural simulations — seeded with Mob’s line art, driven by his emotional state data from the storyboard, rendered with custom shaders that mimic screen-tone grain.”
That’s the pivot: it’s not about replacing animators. It’s about offloading *repetition* so they can focus on *intention*. In S1, every ripple of psychic distortion had to be timed, spaced, and drawn — often by junior animators under brutal deadlines. In S3, the distortion is simulated, but the *trigger point* — where Mob’s fist clenches, where his pupils shrink — is still hand-animated. The emotion stays analog. The physics get digital.
But here’s what the press releases don’t say: that shift came with a hidden tax. Keyframe load dropped. Compositing time *soared*. At Bones Museum, they displayed side-by-side timelines for the S1 gym battle vs. S3’s dome collapse. S1: 14 days key animation, 6 days compositing. S3: 5 days key animation, 19 days compositing. Why? Because UE5 doesn’t just render a clean 3D model — it spits out 12+ render passes per frame (AO, specular, rim light, line-art alpha, texture displacement, motion vector…). Each pass must be manually graded, blended, and *textured* to match Bones’ signature linework. That’s not automation — it’s translation. Like hiring a fluent bilingual editor instead of a translator who just swaps words.
Which brings us to the mecha fans — the ones who’ve been squinting at Gurren Lagann’s drill-spin frames since 2007, who know the difference between a smear frame and a motion blur, who feel betrayed when a transforming robot’s hinge doesn’t *creak* in the animation timing.
You’re right to be wary. What worked for Mob’s abstract psychic explosions doesn’t automatically scale to 50-meter-tall robots punching orbital debris. Gurren Lagann’s charm lived in its deliberate, almost slapstick *imperfection*: joints popping, perspective warping mid-transform, ink lines thickening with strain. That’s hard to simulate. UE5 excels at photorealistic metal sheen and accurate collision physics — not the joyful, rule-breaking weightlessness of Kamina yelling “Drill it, damn it!” while flipping upside-down.
So what does this mean for the Tengen Toppa Gurren Lagann Reboot — officially greenlit last month, with Bones confirmed and Iwai named as CG Supervisor?
First: it won’t look like the original. Not frame-for-frame. But it also won’t look like a generic CGI mecha show. Iwai’s AX23 talk gave away the blueprint: “We’re building a ‘Gurren Mode’ inside UE5 — a set of stylized shaders, forced perspective overrides, and intentional motion-jitter systems. When Simon punches, the engine will *refuse* to smooth the trajectory unless the storyboard explicitly allows it. The robot won’t bend — it’ll crack, and we’ll render the crack with hand-painted fracture lines baked into the normal map.”
In other words: Bones isn’t abandoning its soul. They’re building a new instrument to play it.
The trade-offs are real, though. For fans who love the tactile grit of S1 Mob — the way sweat beads looked like actual ink blots, the way background cels had visible brushstroke texture — S3’s polish can feel emotionally distant. There’s less room for the happy accidents of hand-drawing: the slight wobble in a character’s glare that reads as vulnerability, not instability. CG is precise. Sometimes, precision kills subtext.
But here’s what changed my mind: watching S3 Ep 12’s final shot — Mob walking away from the shattered dome, sunlight catching the edge of his hair — I noticed something. The hair wasn’t animated. It was simulated. Yet the *way* it caught light — soft, diffused, with a faint blue halo — matched the color script from S1’s opening sequence *exactly*. That wasn’t coincidence. It was continuity enforced not by memory, but by data: a shared lighting LUT, a locked color palette embedded in the pipeline. For the first time, Bones wasn’t just preserving style — they were systematizing it.
That’s the quiet revolution. Not “CG vs. hand-drawn.” Not “old school vs. new tech.” It’s about *control*. Control over consistency across 20+ animators. Control over emotional fidelity when the director’s sick for two weeks. Control over delivering a finale that *feels* earned — even if the tools making it happen are invisible.
So yes — your mecha reboot will use more CG. Probably much more. But if Bones sticks to their current playbook, that CG won’t be a replacement. It’ll be a scaffold holding up the same wild, heartfelt, gloriously imperfect humanity that made you fall in love with anime in the first place.
Just don’t expect the robots to squeak the same way.
“The line isn’t disappearing. It’s just learning new languages.”
— Yuki Iwai, Anime Expo 2023