信息图与卡片

布光对照九宫格

把同一个对象放进 3x3 九宫格,展示九种带标签的影棚布光:姿势与五官完全一致,环形光、分割光、蝴蝶光、光影图案与轮廓光一眼对比。

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

Produce one image holding a three-by-three portrait grid on a flat dark grey ground, using the subject described in the request or an uploaded reference. Keep identity, hairstyle, outfit, camera angle, framing, and pose identical across all nine cells so only illumination changes. Print a clean sans-serif number and setup name beneath each cell. Top row: loop lighting with a three-quarter key and compact cheekward nose shadow; hard frontal flash with flat shadows; split lighting dividing the face into bright and dark halves. Middle row: butterfly lighting whose elevated centered key casts a neat mark below the nostrils; a venetian-blind gobo striping face, clothing, and backdrop; under-lighting with inverted moody shadows. Bottom row: warm golden-hour side light; a narrow spotlight with deep chiaroscuro and crushed blacks; neutral rim light outlining hair and shoulders against black without front fill. Render every cell as a believable studio photograph with accurate falloff, contrast, and skin tone. No watermark, logo, or extra text. User request: a woman with short curly hair in a black turtleneck, head-and-shoulders framing

官方示例

布光对照九宫格
示例模型: Seedream 5.0 Lite
布光对照九宫格
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。

  • Portrait Lighting Study Grid · H A J R A

    YouMind-OpenLab/awesome-gpt-image-2

    CC BY 4.0 repository license; original author and source attribution retained from the YouMind catalog.

    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