The Key Animation, the In-Between, and the Digital Paint: How the Three-Stage Pipeline Creates a Single Anime Episode

Consider the traditional Western animation pipeline—where a character’s motion is often smoothed out at a uniform 24 frames per second, creating a fluid, unbroken glide. The Japanese anime pipeline operates on a fundamentally different philosophy: it treats motion as a finite resource, hoarded during quiet moments so it can be spent lavishly when the swords come out. To understand how a single 24-minute episode makes it to the screen, you have to look at the three-stage engine that powers it: key animation, in-betweening, and digital paint.[1][2]

The process begins with the key animators, who produce the "genga" or rough key frames. These artists are the architects of the episode, responsible for drawing the extremes of any given movement—the windup of a punch, the moment of impact, and the follow-through. By defining the start and end points of an action, the key animator dictates the weight, timing, and emotional resonance of the scene, establishing a blueprint that the rest of the production pipeline must follow.[2]

Unlike Western animation, which often defaults to animating "on ones" (24 unique drawings per second) or "on twos" (12 drawings), anime frequently animates "on threes." This means one drawing lasts for three frames, resulting in roughly eight drawings per second. This approach requires the key animator to be incredibly precise with their poses, as each drawing remains on screen long enough for the viewer's eye to register the specific line work and character expression before the next frame appears.[3]

This limited animation technique is not merely a cost-saving measure; it is a stylistic choice that gives anime its distinct, snappy rhythm. When a scene requires exceptional fluidity—a phenomenon fans call "sakuga"—the key animators will jump to animating on ones or twos, instantly signaling to the audience that the stakes have been raised. This dynamic shifting of frame rates is a hallmark of the medium, creating a visual crescendo that uniform frame rates often struggle to match.[3]

Once the key frames are approved by the animation director, the rough drawings are handed off to the in-betweeners, who produce the "douga." If key animation is the architecture, in-betweening is the bricklaying. These artists draw the transitional frames that connect the key poses, smoothing out the motion. They must perfectly mimic the key animator's style, ensuring that the character's volume and proportions remain consistent as they move from point A to point B, a task that requires immense discipline and a suppression of personal artistic flair.[2]

This stage is notoriously labor-intensive and requires immense precision. As Washi's Blog notes, "If every second of an anime was animated at even 2:s that would involve using around 15000 drawings for an episode!" Because 15,000 drawings per week is an impossible pace for a standard television schedule, most average episodes contain between 3,000 and 4,000 frames. Every single one of those frames must be drawn, checked, and approved by an animation director before the episode can move forward into the coloring phase.[2]

To maintain this relentless output, the industry relies heavily on a global assembly line. According to Otaku Realm, "The intermediate, labor-intensive tasks—such as drawing the simple in-betweens (Douga) and digital paint coloring—are sent overseas to production studios in South Korea, China, and Vietnam." This outsourcing is the only way to keep weekly shows on schedule, distributing the massive volume of drawing work across hundreds of artists working in tandem across different time zones to hit the immovable broadcast deadlines.[3]

After the line art is finalized, the physical paper trail ends and the "shiage" or digital paint process begins. Prior to the late 1990s, every frame was hand-painted onto transparent celluloid sheets using specialized acrylics. This analog process was incredibly unforgiving; a single mistake with the paint could ruin an entire cel, and the physical drying time created a hard limit on how quickly an episode could be completed, forcing studios to maintain massive, well-ventilated drying rooms.[4]

Today, the coloring process is entirely digital. Color designers establish specific palettes for every character, accounting for different lighting conditions—such as daytime, sunset, or interior shadows. Digital paint staff then apply these colors to the digitized frames. While the shift to digital paint eliminated the need for physical cels and toxic paints, it did not automate the coloring process entirely. Each of the 3,000 to 4,000 frames must still be individually filled, though modern software tools like bucket-fill algorithms and layer masks have significantly accelerated the workflow.[2][4]

The final compositing stage, known as "satsuei," merges these digitally painted characters with the background art. Here, the digital team adds lighting, blur, and atmospheric effects that would have been impossible to achieve with physical cels. Lens flares, glowing magical auras, and dynamic camera movements are layered over the 2D drawings, bridging the gap between the flat character art and the painted environments to create a cohesive, immersive world that feels alive despite the limited frame rate.[1][5]

What emerges from this three-stage pipeline is a production model designed entirely around efficiency and impact. By saving their drawing budget during quiet dialogue scenes—often relying on panning shots or static poses with only the mouth moving—studios can hoard their resources. When the climax arrives, they unleash that saved labor, allowing the key animators to deliver the breathtaking, high-fluidity sequences that define the medium and dominate social media discussions for weeks after the broadcast.[1][6]