The Character Wardrobe TimelineNobody Changes Coats Mid‑Chapter

There are two ways to get a character’s clothes wrong in an illustrated book, and almost every tool only protects you from one of them. The famous failure is the one everybody names: a character who changes coats between two halves of the same conversation, or loses a scarf crossing a room. The other failure is quieter, much more common, and considerably worse for the book — a character locked into one identical outfit for twenty-four pages that are supposed to span a fortnight. Nothing about it looks like an error on any single page. It just makes the entire story read as one very long afternoon. A character wardrobe timeline exists because both of those are continuity bugs, and fixing only the first one causes the second.

So C2Story splits a character in two. Identity — the face, the hair, the build, everything that makes someone recognisable — is pinned to a saved reference portrait and never moves. Clothing is pulled out as a separate concern with its own rules, its own instruction block and its own clock. On the video side this is a plain two-position switch: locked keeps the reference image’s exact outfit in every single scene, and auto keeps the face, hair, body and identity identical to the reference while dressing each character for this scene — its occasion, its location, its weather, its hour of the day. Auto is the default, because most stories move. Locked is genuinely right for mascots, uniforms, superheroes and books that happen in one afternoon, and it is one click away.

What makes character outfit consistency work rather than merely sound good is the order of operations. The clothes are not decided page by page. First the pages are placed on a real calendar, reasoned from the text itself in whatever language it happens to be written in: shots that are one continuous unbroken moment share a date, a scene change or an implied passage of time advances it, and the dates only ever move forwards. Only then is wardrobe decided, and it is decided from that calendar. The same day means the identical outfit, worded the same way every time so there is nothing left to improvise. A new day means a genuinely different outfit, dressed for that day’s season. A chapter spanning several days must show several outfits, and that is stated as a requirement rather than left as a hope.

Inside a chapter the same idea appears as a time segment on each page. Pages in one segment keep one outfit; the segment increments on a real break — the next morning, after going home, a week later, that evening — or on an occasion a person would actually change for, like a ceremony, a party, a big meeting, a workout or a trip to the beach. An outfit description is attached only when a change is warranted, and deliberately left empty when the clothes genuinely stay the same. That empty field is doing more work than it looks like: a system that re-describes an outfit on every page is a system whose outfits drift.

Then there is the rule that stops a generator from spoiling your ending, which is worth the whole feature on its own. Given a chapter to illustrate, a model reads ahead. It notices the animals join a choir on page seven, concludes that these are choir animals, and dresses them in robes from page one — so a story about friends who become a choir now opens with a choir. The rule against it is absolute: a character’s clothing on any page may reflect only attire the story has introduced up to that page. Special costumes — robes, uniforms, gowns, armour, party clothes — attach to the scene that actually places the character in them and to no page before it. Pages one to six show the everyday look, and there is a test in the codebase whose only purpose is to prove that page seven’s robe never leaks backwards onto page six.

Wardrobe continuity also means one climate per chapter. Illustrate pages independently and each picks its own weather: a cardigan here because the room felt cosy, short sleeves there because the garden felt bright, and a reader turning between them watches the temperature swing thirty degrees in an afternoon. So the season is chosen once, inferred from whatever the text offers — a cold morning, a holiday, blossom, the heating on — and every page’s garments are then checked against it directly: summer means light fabric and no coat, winter adds the layers, indoors the coat can come off but what is underneath stays seasonal. Multi-panel pages get a second pass whose entire job is enforcing one season across every panel, because that is the place where a single image can contradict itself.

And there is a trap inside that fix which is easy to miss: December is not cold everywhere. A story set on a tropical island is hot at Christmas, a mild coastal city is cool in winter rather than freezing, and the southern hemisphere flips the whole calendar. So the season derived from a date is passed along explicitly as only a hint, and the real decision is reasoned from the scene’s own location and its actual climate. The calendar proposes; the place decides. Alongside it runs a character’s own standing rule, read straight out of their description — always in black, never a suit, only traditional dress, never without that red scarf — and kept as a property rather than an outfit, so a character who is always in black wears a black tee on the beach and a black overcoat in the snow and is unmistakably themselves in both.

None of it would hold together page by page, which is why the brief for every shot in a chapter is written in a single pass with the whole chapter visible at once: a continuity bible covering outfits, carried props, time of day and locations, rather than twelve pages each making its own reasonable guess in the dark. Across the join between chapters the previous chapter’s final shot is handed over with its date and its wardrobe, so if the new chapter opens on the same day the clothes carry straight across and only change when the calendar does. Above all of it sits the project’s timeline: an ordered set of beats, each with a day offset and a season, anchored to a start date you can adjust — which means moving your story from July to January is one edit and the wardrobe follows the new months.

One honest caveat, because wardrobe is flagged internally as the most-failed rule in the entire brief. This is enforcement by instruction, checked shot by shot and repeated on every shot for exactly that reason — not a deterministic lock. It is right the overwhelming majority of the time and you should still read the book before you print it. There is also no garment database to populate: outfits are described in words, carried forward by date, and held to your character’s taste. If you need one exact garment reproduced pixel-perfectly on every page, locked mode is the honest answer.

Ready to stop anybody changing coats mid-chapter? Start a story at create and let the pages be dated before the clothes are chosen. If the problem is really the face rather than the coat, fix that first with the character consistency tool. To pin a whole cast’s looks, taste and history into one reusable record, build an AI character bible; to keep that cast straight across a whole series, use the series bible organizer. If the outfit is wrong on a page that already exists, the fix lives on editing an AI-generated book, and if you would rather catch wardrobe problems before any art is rendered at all, lay the book out first with the AI storyboard generator. Or open your character library, write one line about how someone dresses, and watch it hold for the rest of the book.