ai21.net
语言

Comic / Storyboard - Split Photo and Storybook Illustration

原帖未提供效果图

提示词

Create one vertical 3:4 composition that combines the uploaded photo with a hand-drawn illustrated version of the same moment. TOP HALF — ORIGINAL PHOTO. Keep the uploaded photograph in the upper portion of the canvas, occupying roughly 50%. Treat the photo as the only source of truth. Keep the person, face, expression, hair, outfit, accessories, pose, proportions, objects, lighting, colors, and surroundings faithful to the original. Do not beautify, retouch, restyle, replace, or reinterpret the photograph. It should remain natural and recognizable, as if the original image has simply been placed into the artwork. BOTTOM HALF — ILLUSTRATED VERSION. Use the lower portion to recreate the exact same moment as a handmade children's picture-book illustration. The illustrated version should retain: the same person and recognizable facial features, identical pose and body positioning, matching hairstyle and clothing, matching accessories and visible details, the same action and overall mood. Use a soft traditional art approach with: delicate watercolor washes, colored-pencil texture, fine pencil or ink contours, warm textured ivory paper, gentle hand-rendered shading, slightly uneven, organic strokes, subtle imperfections that make it feel handmade. Avoid a polished digital look. No photorealism, 3D rendering, anime aesthetics, or exaggerated cartoon styling. Simplify the background instead of copying every environmental detail. Keep only enough of it to suggest the original setting, while giving the illustrated subject plenty of breathing room. Make the illustrated subject slightly smaller than the photographic subject and leave generous negative space around it. SMALL HAND-DRAWN ACCENTS. Add only a handful of understated doodles around the illustration — perhaps tiny hearts, stars, flowers, or simple motion marks. They should feel naturally drawn onto the page rather than digitally decorated. Include a small handwritten note near the illustration: “{argument name="caption" default="little happy girl"} ♡”. Use a casual brown or soft-gray pencil/ink appearance. If the subject is not a girl, replace the wording with something suitable or leave the caption out. FINAL FEEL. The finished piece should resemble a beautiful handmade scrapbook or children's storybook page, where a real photograph has been paired with a lovingly drawn memory of the same scene. Keep the presentation: vertical 3:4, warm and editorial, cleanly divided between photo and illustration, minimal and elegant, sentimental but not overly decorative, rich in tactile paper and traditional drawing texture. IMPORTANT. Use only the uploaded photograph as visual reference. Do not borrow anything from previous prompts, images, references, or generated results. Do not introduce a different person, outfit, pose, object, setting, color scheme, or visual detail. The lower illustration must unmistakably represent the exact scene and moment shown in the uploaded photograph.

中文译文

创作一幅竖版 3:4 构图,将上传的照片与同一瞬间的手绘插画版本结合在一起。上半部分 —— 原始照片。将上传的照片保留在画面的上半部分,约占 50%。将这张照片视为唯一的事实依据。人物、面部、表情、发型、服装、配饰、姿态、比例、物品、光线、色彩和周围环境都要忠实于原图。不得对照片进行美化、修图、改变风格、替换或重新诠释。它应保持自然且可辨识,仿佛原始图像只是被直接放进了这幅作品中。下半部分 —— 插画版本。使用下半部分,将完全相同的瞬间再现为手工绘制的儿童绘本插画。插画版本应保留:同一个人物与可辨识的面部特征、完全相同的姿态与身体位置、一致的发型与服装、一致的配饰与可见细节、相同的动作与整体氛围。采用柔和传统绘画手法,包括:细腻的水彩晕染、彩铅质感、细致的铅笔或墨水轮廓线、温暖的带纹理象牙色纸张、柔和的手绘阴影、略微不匀称的自然笔触,以及使其显得手工制作的细微瑕疵。避免精致的数码感外观。不得出现照片级写实、3D 渲染、动漫美学或夸张的卡通风格。简化背景,而不是照搬每一处环境细节。只保留足以暗示原场景的少量背景元素,同时为插画主体留出充足的呼吸空间。让插画主体比照片中的主体略小一些,并在其周围留下充裕的留白。小巧的手绘点缀。只在插画周围添加少量低调的涂鸦 —— 也许是小爱心、星星、花朵或简单的动感线条。它们应当像是自然画在纸页上的,而非数码装饰上去的。在插画附近加入一小段手写文字:“{argument name="caption" default="little happy girl"} ♡”。使用随性的棕色或柔和灰色铅笔/墨水笔触外观。如果画面主体不是女孩,请将文字替换为合适的内容,或省略这行文字。最终感受。成品应像一页精美的手工剪贴簿或儿童故事书页面,将一张真实照片与对同一场景饱含深情的手绘记忆并置在一起。保持这样的呈现:竖版 3:4,温暖且具杂志编辑感,照片与插画划分清晰,简约而优雅,富有情感但不过度装饰,充满纸张触感与传统绘画质感。重要。仅将上传的照片作为视觉参考。不得从先前的提示词、图像、参考素材或生成结果中借用任何内容。不得引入不同的人物、服装、姿态、物品、场景、配色方案或视觉细节。下方的插画必须准确无误地呈现上传照片中所展示的确切场景与瞬间。

模板变量

该提示词含可替换参数,方括号内为默认值

caption = little happy girl

溯源

原作者
MahnoorAi12
发布日期
2026-09-06(由推文 ID 反解)
原始出处
查看原帖 ↗
收录来源
youart-open-source/awesome-gpt-image-2-5-prompts · 2096479426452365692 · CC0-1.0 / CC-BY-4.0 (upstream-stated; see ATTRIBUTION.md)