Genjutsu Object Swap
Use one source video and 1–8 reference images to reinterpret motion or replace characters, outfits and products. Reusing a preset prepares its inputs; generation begins only when you submit.
Upload one source video and 1–8 reference images to replace selected subjects, outfits or products. Source frames need at least 409,600 pixels. Use Select segment on the video card to choose 4–30 seconds; otherwise the first 30 seconds are used.
Generation is not available yet. Browse examples, prepare references and save your draft.
Motion & swap presets
Public examples from Higgsfield. Reuse loads the source video, ordered references and prompt. Generation starts only when you submit.
How a selected element relates to its reference
Higgsfield teaching comparisons archived in our storage. These are external reference demonstrations, not generations verified on Genjutsu.tv.
What source video should I prepare?
Use at least 4 seconds. Select segment on the video card chooses 4–30 seconds; otherwise the first 30 seconds are used. Object Swap also requires at least 409,600 pixels per source frame.
How should I use references?
Prepare 1–8 images and check their order. Describe each image’s character, outfit, product or scene role; a prompt can be empty.
Does reusing a preset submit a generation?
Reuse prepares a draft. Review the mode, inputs and credit quote before submitting a generation.
How should I review the change?
Inspect the target’s outline, occlusion and contact with hands, then compare the surroundings. An instruction to retain the scene does not guarantee unchanged pixels.
Genjutsu workflow guides
Original guides to preparing inputs, inspecting examples and understanding your request.