OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

影子越过中线的铅笔炭笔微型超现实海报

Pencil and Charcoal Miniature Surreal Poster: Shadow Crossing the Centerline

3:4上下分屏微型超现实海报：上半保留上传照片，下半让真实投影越过中线并转化为与主体相关的石墨炭笔小景，辅以微型线稿人物、单点取色和一句手写旁白。

影子越过中线的铅笔炭笔微型超现实海报 — AI image creation
影子越过中线的铅笔炭笔微型超现实海报

提示词 / Prompt

中文提示词

将上传照片制作成一张独立的高级设计海报，不多图拼接。采用3:4竖版构图，上下区域各占画面严格50%。上半部分保留原始照片主体的身份、结构、姿态、真实质感、自然光影与原有色彩氛围，仅做轻微而克制的调色；如需适配画幅，可自然扩展环境，但不得拉伸、扭曲或改变主体。上下两部分必须构成同一个连贯创意，而非互不相关的两张图。观察照片中主体投下的影子；若影子不清晰，只能依据可见光线方向谨慎推断。让投影自然越过中线，延伸到暖白纸面，并在那里变成“一只由主体投影自然延伸、与主体姿态相呼应的燕子。”或同样简洁且与主体相关的形象。转化应巧妙却可信，与照片主体的形态、走势和气质相连。下半部分以石墨铅笔、炭笔与少量干刷绘制：靠近分界线时仍像真实投影，越向下越转为手绘线稿，最终成为一个完整简洁的形象。保留铅笔颗粒、炭笔擦痕、手指抹开的灰调与纸张纤维，线条松弛而略带迟疑，避免光滑矢量、油亮数码笔刷和精修插画感。加入一到两个极小的黑色线稿成年人物，正在测量、追随、扶住或惊讶地观察转化后的影子；这些线稿人物以及照片中若出现的人物，都以已配置的固定身份参考 ref-10a-v1 绑定身份。实际照片与身份参考附件须留待下游核验。构图以影子形成的长对角线或纵向走势为主线，其他部分保留大量留白；变形后的影像是下半部唯一视觉焦点，尺度不宜过大。以纸白、石墨灰和炭黑为主，并从上方照片提取一种醒目色，仅用于变形影子中的一个极小细节。影子旁排入简短、自然、略带幽默的手写英文旁白“Even a shadow dreams beyond the frame.”；“影子也会悄悄越过边界。”是其对应的简体中文默认译文，画面按来源要求仅排英文，不写成励志语录。整体呈现微型超现实：真实影子越界、铅笔线稿转义、极小人物、一点色彩和大量留白。避免生硬拼接、满版画面、套路脑洞、卡通感、过度解释和模板化构图。

English prompt

Turn the uploaded photo into one independent high-end design poster; no multi-image stitching. Use a 3:4 vertical layout with upper and lower halves each occupying exactly 50% of the frame. In the upper half, preserve the original photo subject's identity, structure, posture, realistic texture, natural lighting and original color atmosphere, applying only subtle refined color grading. Extend the environment naturally if needed to fit the frame, but do not stretch, distort or change the subject. The upper and lower halves must form one connected idea, not two unrelated pictures. Observe the subject's cast shadow in the photo; if it is unclear, infer it conservatively from the visible light direction. Let this cast shadow naturally cross the center line and continue onto a warm-white paper surface, where its other life becomes 一只由主体投影自然延伸、与主体姿态相呼应的燕子。 or a comparably subject-related concise image. Make the transformation clever yet believable, connected to the photographed subject's form, flow and temperament. Draw the lower continuation with graphite pencil, charcoal and a small amount of dry brush. Keep it shadow-like near the divider and progressively more hand-drawn farther down, ending in one complete, simple image. Preserve pencil grain, charcoal smudges, finger-rubbed gray tones and paper fibers; use loose, slightly hesitant sketchbook lines, not smooth vector art, glossy digital brushwork or polished illustration. Include one or two tiny black-line adult participants who measure, follow, support or look in surprise at the transformed shadow; bind these figures, and any human photo subject, to configured identity reference ref-10a-v1. Keep actual photo and identity-reference attachment as explicit inputs to verify downstream. Build the composition around the shadow's long diagonal or vertical path with a large amount of white space. The transformed shadow is the only focal point in the lower half and stays modest in scale. Use paper white, graphite gray and charcoal black; sample one noticeable color from the upper photo and use it only as one very small accent within the transformed shadow. Add one short, natural, slightly humorous handwritten English aside near the shadow: “Even a shadow dreams beyond the frame.”. Keep it like a quiet voice-over, not a motivational quote. Aim for subtle micro-surrealism: a real shadow crossing the boundary, a pencil-line transformation, tiny figures, one small color accent and generous white space. Avoid stiff splicing, full-frame imagery, clichéd ideas, cartoon style, over-explanation and templated composition.

查看 API 模型

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作