分镜叙事

混合风格分格拼贴

把一组主体与精确细节整理成一张精致拼贴,每个分格采用不同视觉媒介,适合制作风格样张或创意对比图。

更新日期:

完整提示词

英文提示词包含建议的示例需求。生成前请修改主体和细节。原始示例输入和生成参数尚未完全核实。

推荐模型: GPT Image 2

Create a polished style-sampler collage that divides one canvas into a precise grid of distinct panels, using the subjects and exact details supplied in the user request. Give every panel a deliberately different visual medium, such as flat vector art, classical oil portraiture, photoreal studio photography, architectural ink drawing, clay sculpture, or pop-art halftone, while keeping the overall sheet cohesive through consistent gutters, balanced scale, and a restrained framing system. Treat every requested fact as verifiable: preserve specified counts, positions, colors, object states, visible symbols, and spatial relationships without swapping or inventing details. Vary camera angle and composition enough that each cell demonstrates its medium clearly, yet maintain even visual weight across the grid. Use crisp borders or generous white gutters, clean edge alignment, and gallery-quality rendering in every cell. Include small panel labels only when the user supplies them, rendered accurately and unobtrusively. Do not blend styles across panel boundaries, repeat a subject unintentionally, omit requested objects, add watermarks, or introduce real brands and logos. User request: five panels: a teal vector clock at 7:42, an oil portrait, and three studio objects

官方示例

混合风格分格拼贴
示例模型: Seedream 5.0 Lite
混合风格分格拼贴
示例模型: Seedream 5.0 Lite

来源与署名

由 OmniArt 改编和重写。以下列出来源署名、修改说明及使用条件。

  • 漫画分镜叙事设计 · @Gc_qube

    freestylefly/awesome-gpt-image-2

    Repository code is MIT; its disclaimer says community content belongs to original authors/platforms and commercial reuse is not guaranteed.

    Rewritten and consolidated into an OmniArt recipe archetype; template tokens removed; named people, brands, platforms, games, and film IP generalized; no raw prompt text or source image copied.

来源与署名

相关提示词配方