Wardrobe bible three outfits max
Limit the series to three approved outfits. Infinite fashion is how identity lock dies.
Repeatable production workflows and playbooks: batching systems, pipelines, and the exact sequence of steps behind finished content.
Page 2 of 10
Limit the series to three approved outfits. Infinite fashion is how identity lock dies.
Lock face, wardrobe, palette in stills. Video credits are the expensive way to discover the character drifted.
Omni/Seedance multi-shot is cheaper continuity than three separate jobs. It is worse control than three locked stills.
A full mix under VO is a fight. Get stems, or generate the bed without a lead, then duck.
If the model can speak in-shot, do not generate silent and slap TTS on unless you need a locked brand voice.
If hook lengths and ratios drift, the test is unreadable. One cut point, one ratio set, thirty openings.
Extension continues. Regen resets. Extend when the last frame is good. Regen when the idea is wrong.
Seedance 2.5 can do ~30s in one pass. Veo still thinks in ~8s with extension. Stitching is a continuity tax.
Upscaling a shot you will regen is paying twice. Lock the cut, then upscale the keepers.
A CDL-plus-LUT recipe so mixed-model shots sit on one timeline, a stills chart to test it, and the rule for when to regenerate instead.
Ship masters with an xxHash manifest and packing list, plus a rule for WeTransfer versus Aspera, Signiant, MASV, or a physical drive.
Relink rules, hashes, Rec.709 tags, and a missing-media protocol for when a regen changes the filename the online expects.
Keep DX, MX, FX and optional ambience as separate files, with a sync pop, so a later-language dub is a voice swap, not a remix.
When a social cut has no silence for standard AD: freeze, extend, or ship a second described cut, and brief the editor so both versions stay in sync.
A filename grammar for _pic, _mix, _grade, and _del, plus a matrix of who owns each suffix, and why final2 is not a version.
Park on the splice and match black point, grain size, sharpening, and cadence so a generated insert stops announcing itself next to live action.
Review the localized cut on picture with visible time, using a scorecard that splits language, culture, and legal. A script doc misses collisions.
Give mixers timecode, stem, direction, amount in dB or Hz, and a reference clip. A one-page template that replaces 'make it cinematic'.
Ship a still poster, a reduced-motion loop, and a full hero from one brief so prefers-reduced-motion does not mean running three campaigns.
Localization dies without a textless. Keep titles on a video track, export a matching textless, and ship a title sheet so markets restamp without a new grade.
A one-page transcreation brief names shot jobs, the offer, legal must-keeps, and what the adaptor may change, so the linguist rewrites the sale.
A look LUT on unnormalised clips keeps the mismatch and tints it. Neutralise each shot, match to a master, then grade for style.
Tests without a written record get re-run every quarter at full cost. Log these fields per test, then review on a rhythm that feeds a hypothesis backlog.
One pass over likeness, music, trademarks, provenance and disclosure before an asset ships. Each item names the specific failure it exists to prevent.