OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image
时尚社刊人像海报：镜像构图与形与影
Fashion Magazine Portrait Poster: Mirrored Composition of Form and Shadow
可复用的9:16时尚社刊人像海报模板。每次执行生成一张所选构图，默认镜像；使用固定身份参考，并准确呈现ECHO / 回响与FORM / SHADOW / 形与影。
提示词 / Prompt
中文提示词
生成时仅使用已配置的用户参考图作为主要可见人物的面部身份依据；保留该身份并应用所选海报设计。每次执行生成一张全出血、独立的9:16竖版时尚社刊人像海报，不要拼贴。主体为明确成年的长发女性；分别生成不同版式时，保持同一人物、可辨认的面部身份、发型、服装方向、摄影写实感、色调、字体系列和整体艺术指导一致。 色彩：以柔和黄油色作为明亮、低饱和度的主背景，暖象牙色和奶油色为辅助，深可可棕用于超大号字体。整体柔和、精致且有视觉冲击力，但不要过饱和。 仅执行所选构图镜像（MIRROR）：镜像（MIRROR）展示两次同一人像，一张正常，另一张放大并旋转180度从对侧切入，配超大水平字体和微妙轮廓回响；追踪（TRACE）在一侧放置暖色调特写、另一侧放置较小全彩人像，以超大压缩竖排字体形成视觉脊柱，文字在人物前后穿插；焕彩（GLOW）以巨大压缩字体主导画面，人像置于字后，让眼睛、嘴唇、头发和肩膀从字形间显现；漂移（DRIFT）将人像切分为干净的水平摄影条带并轻微左右错移，保持面部可辨认，形成柔和动感，避免数字故障效果。 大号主视觉文字准确呈现“ECHO”，中文可用“回响”（回响）。必须完整呈现辅助文案“FORM / SHADOW”，斜杠必须清晰可见；两词应作为同一短语排在一起，不得拆到画面相距甚远的位置。并配上准确中文对应“形与影”（形与影）。文字克制且清晰，搭配细线、小型几何块和受控留白；人像与大字优先，辅助文案次之。只做全出血设计：不要外边框、海报框、白边、期号/卷号、编号、条形码、随机占位文字、复杂UI、闪烁符号、RGB故障或装饰杂乱。每张图只执行一种所选构图；分别生成时整体视觉关联，但版式明显不同。
English prompt
At generation, use the configured user reference image solely as the facial-identity reference for the main visible person; preserve that identity while applying the selected poster design. Create one full-bleed, standalone 9:16 vertical fashion-editorial portrait poster per execution, never a collage. Depict a mature long-haired woman and keep the same person, recognizable facial identity, hairstyle, outfit direction, photographic realism, palette, typography family, and art direction across separately rendered layout choices. Palette: soft butter yellow as the dominant luminous, low-saturation background, warm ivory and cream as supporting tones, and deep cocoa brown for oversized type. Keep the result refined, soft, and striking without oversaturation. Apply only the selected composition MIRROR: MIRROR shows the same portrait twice, one normal and one enlarged 180-degree flipped portrait entering from the opposite side, with oversized horizontal type and subtle outline echo. TRACE places a large warm-toned close-up on one side and a smaller full-color portrait on the other, with oversized condensed vertical typography as a central spine, layered before and behind the subject. GLOW makes huge condensed type dominant and places the portrait behind it, allowing eyes, lips, hair, and shoulders to show through and between the letterforms. DRIFT divides the portrait into clean horizontal photographic strips shifted slightly left and right; keep the face recognizable and create soft motion without digital-glitch effects. Use the exact large display word ECHO (ECHO). REQUIRED SUPPORTING COPY: render the complete literal “FORM / SHADOW” together, including the slash, as one readable phrase; do not split the words into separate distant labels. Pair it with the accurate Chinese meaning 形与影 (FORM / SHADOW / 形与影). The main large word may also be rendered as 回响 in Chinese. Keep all required copy legible, with thin rules, small geometric blocks, and generous negative space; portrait and display type first, supporting text second. Full bleed only: no outer border, poster frame, white margin, issue/volume number, barcode, random placeholder text, complex UI, sparkle symbols, RGB glitch, or decorative clutter. One selected composition per image; across separate runs the posters should feel related but use clearly different layouts.
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
相关作品
- 宝丽来照片讲述原创谜案
以“一桩原创谜案”为主题，在方形软木板上创作复古侦探板式视觉故事：顶部纸签呈现清晰标题，下方五至六张宝丽来沿松散时间顺序排列并配简短说明，彩色线索绳串联故事节点。
- 琥珀雨靴 · 彩色铅笔
Moss Image 原创创意图集｜彩色铅笔｜琥珀雨靴。原创提示词生成。
- 沙漠风琴 · 包豪斯图形
Moss Image 原创创意图集｜包豪斯图形｜沙漠风琴。原创提示词生成。
- 游乐园儿童识字小报
竖版A4儿童学习小报《游乐园识字小报》，以明亮卡通游乐园为主画面，围绕18个指定名词清晰绘图，并在对应物体旁标注带声调拼音和简体汉字两行标签。
- 原创毛绒伙伴与身份绑定成人的城市时尚跨页
原创高端时尚杂志跨页：一位身份绑定的明确成年主体与非系列IP毛绒伙伴在泛化黄昏街景互动；右侧呈现风格标签、近似色卡及不虚构价格的单品拆解。
- 同一成年女性的十种发型与历史简卡
发型对比信息卡：同一位明确成年女性的身份和造型保持稳定，以十格展示不同年代发型、名称与简短历史注释；历史资料由用户提供，缺少依据时准确标注“历史来源待提供”，不虚构来源或年代。