Zero drafting for animation pipelines

The cost curve is the whole argument

In live action, a late continuity discovery is usually a rewrite and a scheduling problem. In animation the same discovery crosses a threshold where the work is no longer editable in the same sense.

• **In the outline or script**, a contradiction costs a decision.
• **After a board is approved**, it costs a re-board of the affected sequence.
• **After it has gone to a vendor**, it costs a retake — new work, a new round trip, and a slot in a schedule that was already committed.
• **After animatic lock or shipped animation**, the honest options are usually to absorb the error or to cut around it.

The practical consequence is that the value of a ninety-minute check is not constant across production. It is enormous before boards and close to zero after. Everything below is about spending that ninety minutes at the right moment.

Contradictions become physical

This is the part that catches writers coming from prose or live action.

In a novel you can write "they crossed the room" and never establish the room. In animation someone has to draw the room, and the moment they do, its geography is canon — where the door is, what the window looks onto, how far the table is from the bed. Nobody wrote it down. It is now binding, and the next episode's board artist will either match it or contradict it.

The same applies to character height relative to a doorway, whether a character can physically reach a shelf, what a character is wearing in a scene that continues across an act break, and whether a prop was on the table in the previous shot. All of it is canon. Very little of it lives in the script.

So an animation zero draft has to test a larger surface than a prose one, and much of that surface exists in model sheets, prop designs, and approved boards rather than in the bible.

Run the scene, then run the space

The adaptation is to split it into two passes with different questions.

**The scene pass** is the standard method: write the scene at full resolution against the story record, log Missing and Broken, resolve, delete. This catches the narrative and knowledge-state failures — who knows what, whether the timeline supports it, whether a world rule is being violated.

**The space pass** is animation-specific and is where the physical canon gets tested. Take the same scene and write down every spatial and physical fact it silently assumes. Not prose — a list. Where each character is standing at the top of the scene and where they end up. What they touch. What they can see from where they are. How long a crossing takes. What they are holding, and where it came from.

Then check that list against the model sheets, prop bible, and any approved boards of the same location. The Broken items from a space pass are the ones that would otherwise have surfaced as a retake, because they are exactly the class of error that reads as fine in a script and impossible in a drawing.

Where it goes in the pipeline

Two windows, and both close hard.

The first is **before boards**, after the script is drafted. This is the highest-value slot in the entire pipeline, because everything downstream is built on the board and nothing has been drawn yet. If you run this practice exactly once per episode, run it here.

The second is **at any design lock for a new location or character**. A new environment design is a large amount of canon entering the world at once, most of it never written down. A space pass against the first scene set there catches the mismatches while the design is still cheap to adjust.

It is worth being blunt about the third case: after a vendor has the sequence, a zero draft is no longer a preventative instrument. Run it anyway if you are heading into another season with the same world, but treat what it finds as a record correction rather than a fix.

Distributed teams raise the stakes

Most animation is made across studios, time zones, and often languages. That multiplies the specific failure this practice targets.

A live-action room can resolve an ambiguity with a question across the table. A board artist eleven time zones away resolves it by making a reasonable assumption, because waiting a day for an answer is not free. Reasonable assumptions are where parallel canon comes from, and they arrive already drawn.

Which means the record has to be able to answer questions without a conversation attached. Anything a distant artist has to guess at is a fact your bible does not communicate, regardless of how obvious it seems from inside the core team. A zero draft — particularly the absent-writer variant used in rooms, run by someone outside the core group — is a direct measurement of that gap, and it is much cheaper to take that measurement than to receive it as a batch of retakes.

Frequently asked questions

When should an animation production zero draft?

Before boards, after the script is drafted — the highest-value slot in the pipeline, since everything downstream builds on the board. Also at any design lock for a new location or character, where a large amount of unwritten visual canon enters at once.

What is a space pass?

An animation-specific second pass that lists every spatial and physical fact a scene silently assumes — positions, distances, sightlines, what characters touch and hold — then checks that list against model sheets, prop designs, and approved boards. It catches the errors that read fine in a script and become impossible in a drawing.

Why does animation need this more than live action?

Two reasons. The cost curve is steeper — after boards ship, a contradiction becomes a retake rather than a rewrite. And the world is realised visually, so geography, scale, and props become binding canon the moment they are drawn, even though nobody wrote them down.