Free · copies cast, sets and prompts to your workspace · you pay only when you generate
About this template
Give each state its own reference sheet and attach both to the shots that need them. In this template the woman and the wolf each have a model sheet generated with GPT Image 2.5 Flare at 2K in 16:9. Each sheet has front and side views, a title block listing traits, colour swatches and a strip of captioned detail close-ups. The wolf sheet fixes details like an amber-orange iris and a shoulder height of 110 cm, so when the woman's eyes change colour mid-scene they match the wolf's amber. With both ends of the change defined, the model moves between two fixed designs instead of inventing a new creature in every shot.
The creator made two location plates: a forward view and a reverse view that points back at the first one. The first looks from inside the forest toward a broken rock lip above a valley full of cloud and rain. The reverse stands at that lip and faces back into the trees, and its prompt references the first image so the black mud, wet leaves, mossy rocks and roughly eight-metre clearing carry over. Both plates leave out people, animals and readable text, which keeps them clean for placing characters. With forward and reverse angles available, the close-ups and wide shots cut together without the forest shifting around the action.
The canvas holds five reference image nodes for character, creature and location, plus video shots generated with Seedance 2.5 and a timeline that assembles them. The change is split into short beats. The hiker sinks into the mud while predators circle, ridges glow along her spine, her eyes turn amber, and then her feet and face reshape until the wolf stands its ground. To remake it, duplicate the template, replace the woman or wolf sheet with your own design, and regenerate the shots that use it. When the sequence reads cleanly, export it as mp4 or FCPXML.