Afureko Post-Recording Forces Anime Into Three-Frame Mouth Cycles Instead of Phonetic Lip Sync

When a Western viewer watches a dubbed anime and notices the mouth movements do not quite match the English words, they often assume it is a translation issue. But the mismatch is baked into the original Japanese broadcast, too. The mouth shapes on screen were never drawn to match the words being spoken.

The discrepancy stems from a fundamental difference in how the two industries structure their production pipelines. In Western animation, the actor's voice dictates the art. In Japanese television animation, tight schedules force actors to squeeze their entire performance into rigid, pre-drawn mouth flaps.

This Japanese pipeline is known as afureko, a portmanteau of "after-recording." Born from cost-saving measures established in 1963 with Osamu Tezuka's Astro Boy, the process dictates that the animation is completed before the recording sessions begin.

The Pre-Scoring Standard

To understand why afureko changes the visual output, it helps to look at the Western alternative. Western studios utilize a process called pre-scoring, where actors record their lines using storyboards or early animatics as a reference.

Once the audio track is locked, animators draw the character's mouth movements and facial expressions to match the actor's specific vocal performance. This method gives Western voice actors significant creative freedom, as their pacing and ad-libs become the foundation of the scene.

Because the audio exists first, Western animators can employ phonetic lip sync. They rely on standardized reference guides, such as the Preston Blair mouth chart, which assigns specific mouth shapes to distinct sounds.

A standard phonetic chart contains six to nine discrete drawings. There is a specific shape for an "O" sound, a wide shape for an "E," and a closed-lip shape for consonants like "M," "B," and "P." The animator listens to the track and matches the phonemes frame by frame.

"Six to nine discrete drawings means the animation cannot show a mouth partway between two shapes unless you draw the in-between yourself," notes animation tool developer Hysenlabs.

The Mechanics of Afureko

This level of detail requires the audio to be finalized before the animation can proceed. The afureko pipeline reverses this dynamic entirely. Because the animation is produced first, the animators have no audio track to reference when drawing a character's dialogue.

Without an audio guide, phonetic lip sync is impossible. If an animator guesses that a character will make an "O" shape at frame 42, but the voice actor delivers the line faster or slower, the visual will clash violently with the audio.[1]

To solve this, the anime industry relies on a technique called kuchipaku, or mouth flapping. Instead of drawing specific phonetic shapes, the animator draws a generic amplitude cycle that simply indicates whether the character is speaking.[1]

The Three-Frame Solution

The standard kuchipaku cycle consists of exactly three frames: a closed mouth, a half-open mouth, and a fully open mouth. These three frames are looped repeatedly for the duration of the character's dialogue, regardless of the specific words being spoken.[1]

This three-frame cycle mathematically restricts the animation. Because there are only three generic shapes, the animation cannot show a mouth partway between two phonetic sounds. The visual conveys the rhythm of speech, but not the mechanics of pronunciation.[1]

The animator's job is simply to determine the duration of the speech. They mark the start and end points of the dialogue on the exposure sheet, and the compositing team loops the three-frame cycle between those points.[1]

The Voice Actor's Constraint

This pipeline shifts the burden of synchronization entirely onto the voice actor, known in Japan as a seiyuu. When the seiyuu steps into the recording booth, they are watching edited video—often fully composited scenes, or at least timed animatics.

The actor must deliver their lines to match the on-screen lip flaps. If the animator drew 24 frames of a flapping mouth at 24 frames per second, the actor must stretch or compress their delivery to fit exactly 1.0 seconds of screen time.[1]

This forces the seiyuu to adapt their pacing to the visual director's established rhythm. They cannot pause for dramatic effect unless the animator already drew a pause into the scene. The emotional delivery must fit into a pre-determined structural box.

Recording sessions are highly technical exercises in timing. Actors watch the screen for visual cues, such as bubbles or markers that indicate when a character's mouth opens and closes. They frequently record takes in their entirety, matching the flaps in real time.

The Economics of Speed

While afureko places immense pressure on the actors, it is a highly efficient production model. By decoupling the animation from the audio recording, studios can work on both processes simultaneously, saving valuable weeks in the schedule.[2]

Japanese television runs on incredibly strict broadcast schedules, meaning a standard anime episode must fit a precise runtime down to the exact second, typically 24 minutes and 30 seconds. Editors cut the scenes to establish the show's pacing before the audio is even recorded.[2]

If an episode is running behind schedule, the afureko process allows the voice actors to record to rough pencil sketches. The animation team can then finish coloring and compositing the frames right up until the broadcast deadline, without waiting for the audio mix.

The Dubbing Puzzle

The afureko pipeline also explains why localizing anime into other languages is such a complex technical challenge. When a Western studio licenses an anime for an English dub, they inherit the rigid three-frame mouth cycles.

The localization team must write a translated script that captures the original meaning while matching the exact duration of the on-screen flaps.

"One of the most critical components in this process is lip-sync dubbing—the ability to match the spoken word with the animated character's mouth movements with such precision that the viewer remains immersed," notes localization firm DUBnSUB.[3]

If the animated character closes their mouth quickly at the end of a sentence, the adapted English script needs a word that ends with a closed-lip consonant, such as a "P" or an "M." A literal translation rarely fits the visual constraints.

Audio editors then step in to make micro-adjustments, stretching or compressing the recorded dialogue by fractions of a second. The goal is to make the dubbed audio blend seamlessly with the three-frame cycle, maintaining the illusion of synchronized speech.