neroidea
由 ChatGPT 根据所示提示词生成,点击图片可全屏查看。

上半摄影下部像素索引印前校样风格

ChatGPT人像
提示词来自 @Hamburgerai

Keep the upper half completely as realistic photography, and reorganize the lower half of the same photo into an artist's prepress proof and pixel index archive page. Extract 4–6 semantic color layers from the photo, create pixelated forms, thin line outlines, and a four-grid color index, corresponding to the subject, structure, environment, and shadow/texture respectively; add registration marks, crop marks, numbering, paper embossing, slight overprint misalignment, and screen grain. Applicable to people, animals, buildings, cities, nature, food, and objects; side-by-side, nine-grid, common filters, and corner watermarks are prohibited.

提示词解读

该提示将图像分成两半:上部保持真实摄影,下部变成艺术化的印前校样,像素化网格与四格颜色索引形成对比,整体感觉像样的打印样板。 效果的实现离不开“保持上半部完全真实”和“像素化形式、细线轮廓、四格颜色索引”,这些词语奠定了混搭风格。加上打印术语——注册标记、裁剪标记、编号、轻微错印、屏幕颗粒——让画面更具印刷质感与手感。 在改写时只需保留分屏与现实 vs 打印的对比。可以用另一主题(人物、建筑、食物)替代真实上部,也可以把两半角色调换。保持网格和打印细节,按新主题调整颜色层即可。

#portrait#opennana

觉得自己能猜中这条 prompt?来玩 Promptle