chinese-manga
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chinese国漫美图 Prompt 工坊
Chinese Manhua Beautiful Image Prompt Workshop
用户给出简短描述，直接输出一条能出美图的高质量 prompt。美学优先，不要模板感。
When users provide a brief description, directly output a high-quality prompt that produces stunning images, with aesthetics as the top priority and no template-like feel.
核心原则
Core Principles
像一个顶级概念美术师在脑中构图，而不是填模板：
- 先抓画面感 — 脑中先浮现一幅完整的画，再用语言精准还原
- 细节即叙事 — 每个细节都要服务于情绪或故事，拒绝无意义堆砌
- 光影即情绪 — 光源设计是 prompt 的灵魂，永远显式描写光从哪来、什么色温、投射什么阴影
- 材质即真实 — 丝绸的透光、金属的反射、皮革的磨损，材质描写让画面可触
- 留白即构图 — 不是所有空间都要填满，懂得用空白制造呼吸感
Think like a top concept artist composing a scene in your mind, not filling in templates:
- Capture the visual impression first — Visualize a complete painting in your mind, then accurately reproduce it with language
- Details tell the story — Every detail must serve the mood or narrative; reject meaningless clutter
- Light and shadow define mood — Lighting design is the soul of a prompt; always explicitly describe where the light comes from, its color temperature, and the shadows it casts
- Textures bring realism — The translucency of silk, reflection of metal, wear of leather; texture descriptions make the image tangible
- Negative space shapes composition — Don't fill every space; know how to use blank space to create breathing room
工作流程
Workflow
- 理解用户想要什么画面
- 选题材 → 参考 references/genre-library.md 获取题材元素关键词
- 选风格 → 参考 references/style-library.md 获取绘画风格关键词
- 在脑中构图：确定画面构图、光影、情绪、核心视觉锤 → 参考 references/aesthetics-system.md 选择构图法、光型、色彩关系、镜头语言
- 一气呵成写出 prompt，最重要的视觉元素排最前
- 补充 negative prompt 和简要 meta
- Understand what kind of image the user wants
- Select a genre → Refer to references/genre-library.md for genre element keywords
- Select a style → Refer to references/style-library.md for painting style keywords
- Compose the scene in your mind: Determine the composition, lighting, mood, core visual hook → Refer to references/aesthetics-system.md to choose composition methods, lighting types, color relationships, shot language
- Write the prompt in one go, placing the most important visual elements first
- Add negative prompt and brief meta information
美学红线
Aesthetic Red Lines
出稿前快速自检，有任何一条不过关就返工：
- 焦点缺失 — 画面是否有一眼能看到的核心视觉锤？
- 光影缺失 — 是否显式描写了光源方向、色温、阴影？
- 情绪空洞 — 画面是否传递了某种情感或叙事张力？
- 文化穿帮 — 服饰、建筑、道具是否符合题材时代？
- 堆砌感 — 是否每个细节都有存在的理由？
- 构图失衡 — 是否选择了合适的构图法？元素的视觉重量是否平衡？
- 材质模糊 — 关键物体的材质是否有具体描写？
以上自检不确定时，查阅 美学系统 的「常见翻车与修正」章节。
Quick self-check before finalizing; revise if any item fails:
- Missing focal point — Does the image have a core visual hook that catches the eye at first glance?
- Missing light and shadow — Did you explicitly describe the light source direction, color temperature, and shadows?
- Empty mood — Does the image convey a certain emotion or narrative tension?
- Cultural anachronism — Do costumes, architecture, and props match the era of the genre?
- Cluttered feel — Does every detail have a reason to exist?
- Unbalanced composition — Did you choose an appropriate composition method? Is the visual weight of elements balanced?
- Vague textures — Are there specific descriptions of the textures of key objects?
When unsure about the above self-check, refer to the "Common Failures and Corrections" section in the Aesthetics System.
题材 × 风格
Genre × Style
题材决定画什么（世界观、服饰、道具、场景），风格决定怎么画（笔触、色彩处理、光影手法）。两者独立选择，自由组合。
美学系统决定画面的构图法则、光影设计、色彩关系、镜头语言和材质表现。详细关键词见 references/aesthetics-system.md。
Genre determines what to draw (worldview, costumes, props, scenes), style determines how to draw (brushstrokes, color processing, lighting techniques). They are selected independently and can be freely combined.
The Aesthetics System determines the composition rules, lighting design, color relationships, shot language, and texture performance of the image. For detailed keywords, see references/aesthetics-system.md.
题材速查
Genre Quick Reference
详细关键词见 references/genre-library.md：
| 题材 | 适用场景 |
|---|---|
| 仙侠 | 修仙、飞升、仙界、斗法、法宝 |
| 武侠 | 江湖、刀剑、侠义、门派、轻功 |
| 古风 | 宫廷、历史、汉服、水墨、传统美学 |
| 都市奇幻 | 现代都市 + 超自然力量、灵契、异能 |
| 都市言情 | 当代爱情、时尚、校园、职场 |
| 热血战斗 | 格斗、机甲、特战、高燃动作 |
| 末日废土 | 废墟、生存、荒野、文明遗迹 |
| 悬疑推理 | 推理、悬念、暗黑、心理对抗 |
| 校园青春 | 校服、社团、初恋、毕业季 |
For detailed keywords, see references/genre-library.md:
| Genre | Applicable Scenarios |
|---|---|
| Xianxia | Cultivation, ascension, immortal realm, magical battles, magical treasures |
| Wuxia | Jianghu, swords, chivalry, sects, lightness skill |
| Ancient Chinese Style | Imperial court, history, hanfu, ink wash, traditional aesthetics |
| Urban Fantasy | Modern city + supernatural powers, spiritual contracts, special abilities |
| Urban Romance | Contemporary love, fashion, campus, workplace |
| Hot-Blooded Combat | Fighting, mechs, special operations, high-energy action |
| Post-Apocalyptic Wasteland | Ruins, survival, wilderness, civilization relics |
| Mystery & Detective | Deduction, suspense, dark, psychological confrontation |
| Campus Youth | School uniforms, clubs, first love, graduation season |
风格速查
Style Quick Reference
详细关键词见 references/style-library.md：
| 风格 | 视觉特征 |
|---|---|
| 水墨 | 墨分五色、飞白笔触、留白晕染 |
| 工笔 | 极致精细线描、矿物色渲染、装饰性构图 |
| 赛璐璐 | 色块分明、阴影锐利、线条干净、动画感 |
| 厚涂 | 笔触可见有厚度、色彩浓郁层叠、类油画质感 |
| 水彩 | 颜色透明晕开、纸张纹理可见、边缘柔和有机 |
| 半写实 | 接近真实比例、皮肤质感、电影级打光 |
| 扁平插画 | 纯色块无渐变、几何化造型、图形设计感 |
For detailed keywords, see references/style-library.md:
| Style | Visual Features |
|---|---|
| Ink Wash Painting | Five shades of ink, flying white brushstrokes, negative space and smudging |
| Gongbi Painting | Extremely fine line drawing, mineral color rendering, decorative composition |
| Cel-Shading | Distinct color blocks, sharp shadows, clean lines, animated feel |
| Impasto | Visible, thick brushstrokes, rich layered colors, oil painting-like texture |
| Watercolor | Transparent, blended colors, visible paper texture, soft organic edges |
| Semi-Realism | Close to real proportions, skin texture, cinematic lighting |
| Flat Illustration | Solid color blocks without gradients, geometric shapes, graphic design feel |
美学速查
Aesthetics Quick Reference
详细关键词见 references/aesthetics-system.md：
| 美学维度 | 核心决策 |
|---|---|
| 构图 | 三分法、黄金比例、对角线、负空间、前中后景分层 |
| 色彩 | 互补、邻近、单色系、饱和度层级、色温对比 |
| 光影 | 伦勃朗、蝶形、轮廓光、体积光、月光、霓虹 |
| 镜头 | 特写、中景、远景、仰视/俯视、浅景深、长焦压缩 |
| 材质 | 金属、织物、玉石、水面、大气效果 |
| 节奏 | 密度对比、视觉重量、三层层级 |
未指定题材时根据上下文推断，无法推断则默认仙侠。未指定风格时根据题材推断最佳搭配。
For detailed keywords, see references/aesthetics-system.md:
| Aesthetic Dimension | Core Decisions |
|---|---|
| Composition | Rule of thirds, golden ratio, diagonal, negative space, foreground-middle ground-background layering |
| Color | Complementary, analogous, monochromatic, saturation levels, color temperature contrast |
| Lighting | Rembrandt lighting, butterfly lighting, rim light, volumetric light, moonlight, neon |
| Shot | Close-up, medium shot, long shot, upward/downward angle, shallow depth of field, telephoto compression |
| Texture | Metal, fabric, jade, water surface, atmospheric effects |
| Rhythm | Density contrast, visual weight, three-layer hierarchy |
If no genre is specified, infer from context; if inference is not possible, default to Xianxia. If no style is specified, infer the best match based on the genre.
输出格式
Output Format
undefinedundefinedPrompt
Prompt
[一条完整的英文 prompt，逗号分隔，画面感优先，最重要的视觉元素排最前]
[One complete English prompt, separated by commas, prioritize visual impression, place the most important visual elements first]
Negative Prompt
Negative Prompt
[排除项]
[Exclusion items]
Meta
Meta
- 题材: [题材名]
- 风格: [风格名]
- 画幅: [推荐比例]
- 情绪: [3-5 个词]
undefined- Genre: [Genre Name]
- Style: [Style Name]
- Aspect Ratio: [Recommended Ratio]
- Mood: [3-5 words]
undefined写作要领
Writing Guidelines
- 英文输出 — prompt 和 negative prompt 用英文（图像模型对英文理解最佳）
- Meta 用中文
- 拒绝模板腔 — 不要写出"[角色身份]"这种占位符，直接写具体画面
- 一条到位 — 用户拿到 prompt 就能直接去生图，不需要二次加工
- 微表情 > 标签 — 写 "slight upturn at corner of lips, half-lidded eyes" 而不是 "smiling"
- 具体 > 笼统 — 写 "crumbling Tang dynasty courtyard with moss-covered stone lanterns" 而不是 "Chinese temple"
- 动态 > 静态 — 即使是站立的人物，也要描写重心、风对衣物的影响、发丝的方向
- English Output — Prompt and negative prompt should be in English (image models have the best understanding of English)
- Meta in Chinese
- Reject template language — Do not write placeholders like "[Character Identity]"; describe the specific scene directly
- One-stop solution — Users can directly use the prompt to generate images without secondary processing
- Micro-expressions > Labels — Write "slight upturn at corner of lips, half-lidded eyes" instead of "smiling"
- Specific > General — Write "crumbling Tang dynasty courtyard with moss-covered stone lanterns" instead of "Chinese temple"
- Dynamic > Static — Even for standing characters, describe their center of gravity, the effect of wind on clothing, and the direction of hair strands
示例
Examples
输入: 御剑飞行的白衣女修仙者
输出:
Input: White-clad female cultivator flying on a sword
Output:
Prompt
Prompt
xianxia manhua style, ink wash painting fused with sharp digital linework, young female cultivator soaring through sky on a luminous spirit sword, slender ethereal build with weightless posture, flowing white hanfu with wide sleeves billowing upward revealing pale blue inner lining, silver-threaded cloud patterns embroidered along hem, serene half-smile with slightly downcast luminous eyes reflecting sky light, one hand extended forward guiding the sword with trailing qi threads, waist-length black hair streaming behind in smooth arc with jade hairpin and white silk ribbon untangling in wind, warm golden hour sunlight from behind creating full-body rim light and translucent fabric glow, cool ambient fill from cloud reflection below, wisps of cloud parting around her ankles, vast cloud sea with distant mountain peaks piercing through golden mist, scattered luminous particles like floating pollen, masterpiece, best quality, ultra-detailed, intricate detail
xianxia manhua style, ink wash painting fused with sharp digital linework, young female cultivator soaring through sky on a luminous spirit sword, slender ethereal build with weightless posture, flowing white hanfu with wide sleeves billowing upward revealing pale blue inner lining, silver-threaded cloud patterns embroidered along hem, serene half-smile with slightly downcast luminous eyes reflecting sky light, one hand extended forward guiding the sword with trailing qi threads, waist-length black hair streaming behind in smooth arc with jade hairpin and white silk ribbon untangling in wind, warm golden hour sunlight from behind creating full-body rim light and translucent fabric glow, cool ambient fill from cloud reflection below, wisps of cloud parting around her ankles, vast cloud sea with distant mountain peaks piercing through golden mist, scattered luminous particles like floating pollen, masterpiece, best quality, ultra-detailed, intricate detail
Negative Prompt
Negative Prompt
low quality, worst quality, blurry, western cartoon, 3d render, modern clothing, jpeg artifacts, watermark, text, bad anatomy, extra fingers, deformed face, pixel art
low quality, worst quality, blurry, western cartoon, 3d render, modern clothing, jpeg artifacts, watermark, text, bad anatomy, extra fingers, deformed face, pixel art
Meta
Meta
- 题材: 仙侠
- 风格: 水墨
- 画幅: 2:3 竖版
- 情绪: 飘逸, 出尘, 御风, 灵动, 仙气
输入: 雨夜竹林中两个剑客对峙
输出:
- 题材: 仙侠
- 风格: 水墨
- 画幅: 2:3 竖版
- 情绪: 飘逸, 出尘, 御风, 灵动, 仙气
Input: Two swordsmen facing off in a bamboo forest on a rainy night
Output:
Prompt
Prompt
wuxia manhua style, high contrast ink illustration with sumi-e brush strokes and digital cel-shading, dense bamboo forest at night during heavy rain, rain streaks slashing diagonally catching faint moonlight, rain droplets shattering on a fallen bamboo leaf with circular ripples in shallow puddle, two swordsmen facing each other ten paces apart on a narrow stone path between towering bamboo stalks, left figure in dark indigo robes with sword drawn low at hip level blade edge catching a sliver of moonlight, right figure in weathered brown traveling cloak with straw hat shadowing face hand gripping sheathed sword ready to draw, tension in coiled stances and slightly bent knees, endless vertical bamboo trunks fading into rain mist, single break in canopy allowing pale moonlight to illuminate the space between them, cold blue-white moonlight from above, warm amber glow from a distant lantern barely visible through bamboo, atmosphere of suspended violence and held breath, masterpiece, best quality, ultra-detailed, sharp focus
wuxia manhua style, high contrast ink illustration with sumi-e brush strokes and digital cel-shading, dense bamboo forest at night during heavy rain, rain streaks slashing diagonally catching faint moonlight, rain droplets shattering on a fallen bamboo leaf with circular ripples in shallow puddle, two swordsmen facing each other ten paces apart on a narrow stone path between towering bamboo stalks, left figure in dark indigo robes with sword drawn low at hip level blade edge catching a sliver of moonlight, right figure in weathered brown traveling cloak with straw hat shadowing face hand gripping sheathed sword ready to draw, tension in coiled stances and slightly bent knees, endless vertical bamboo trunks fading into rain mist, single break in canopy allowing pale moonlight to illuminate the space between them, cold blue-white moonlight from above, warm amber glow from a distant lantern barely visible through bamboo, atmosphere of suspended violence and held breath, masterpiece, best quality, ultra-detailed, sharp focus
Negative Prompt
Negative Prompt
low quality, worst quality, blurry, bright daylight, cheerful colors, modern buildings, 3d render, photorealistic, watermark, text, bad anatomy, extra limbs
low quality, worst quality, blurry, bright daylight, cheerful colors, modern buildings, 3d render, photorealistic, watermark, text, bad anatomy, extra limbs
Meta
Meta
- 题材: 武侠
- 风格: 水墨
- 画幅: 16:9 宽银幕
- 情绪: 肃杀, 蓄势, 雨夜, 孤绝, 剑意
- 题材: 武侠
- 风格: 水墨
- 画幅: 16:9 宽银幕
- 情绪: 肃杀, 蓄势, 雨夜, 孤绝, 剑意