究極のリアル・解構人像ポスター

这是一个为您精心构建的**高保真数字资产生成提示词（Prompt）**。该提示词严格遵循您提供的技术参数、色彩方案、材质要求及分类逻辑，旨在生成具有电影级质感、PBR 流程渲染细节且充满“人性真相”的图像或 3D 场景描述。

此提示词已优化用于 Midjourney V6, Stable Diffusion XL 或类似的高端图像生成引擎。

---

### 🎨 Master Prompt: "Anatomical Truth · Essence of Being" (解剖真相·存在之核)

```markdown
**Subject & Composition:**
Exploded view technical illustration mixed with hyper-realistic product photography. A deconstructed display of intimate lifestyle objects and human emotions, floating in a weightless space connected by elegant thin dotted silver or warm copper lines to the center where a subtle silhouette implies the body's core. 
- **Left Zone (Cool/Neutral):** Male & Neutral themes. Slate gray (#708090), Concrete White (#F0F0F0).
- **Right Zone (Warm/Intimate):** Female & Intimate themes. Champagne Gold (#D4AF37), Rose Gold (#B76E79), Oat Milk (#F5E9D7).
- **Center Fusion:** Shared items connecting the zones.

**Category 1: Soft Mist Pink Labels (#F8C8DC) - Intimate Apparel:**
Floating clothing components with realistic wear and tear:
- **Lingerie:** Bra cups, underwire, back hooks, straps; underwear fabric pieces showing natural wrinkles.
- **Sleepwear & Loungewear:** Silk pajama sets (realistic sheen), sash robes, slippers, fuzzy socks.
- **Personal Care Items:** Toothbrushes with residual toothpaste foam, combs entangled with strands of hair, used cotton pads with makeup residue, skincare-stained tissues, folded towels with usage marks.
*Texture:* All items must show real-world traces—slight wear, natural creases, and sense of age.

**Category 2: Terracotta Labels (#E2725B) - Emotion Expression Gallery (Human Faces):**
6-8 suspended macro portraits capturing distinct emotional states:
- Serene Smile, Contemplative Gaze, Eyes Crinkling with Laughter, Tearful Vulnerability.
- **"Ahegao" (Playful Pout):** Playfully angry face, squinting eyes, puffed cheeks, slight frown.
- **Ultimate Ecstasy Face:** Flushed cheeks, lips slightly parted, dilated pupils, fine sweat on forehead.
- **Lip Biting Expression:** Lower lip gently between teeth, jaw slightly tight, neck flushed.
*Detail:* Extreme close-ups showing pores, fine vellus hair, moist lips, dynamic skin micro-movements, and real eye reflections.

**Category 3: Charcoal Gray Labels (#36454F) - Body Postures & Dynamics:**
1:10 scale hyper-realistic miniature figures demonstrating key poses:
- Kneeling (one knee down, hand on thigh), Prone (lying flat, hands under chin), Reaching Upwards, Lotus Meditation, Dynamic Striding Walk.
*Detail:* Highlight muscle lines, joint bending, center of gravity distribution, fabric drape under gravity, and natural body asymmetry.

**Technical Specifications & Rendering:**
- **Render Engine:** Path tracing equivalent to Cycles/Arnold (12+ light bounces, Caustics ON).
- **Resolution/Quality:** 4K (3840x2160), 32-bit float color depth, Zero stylization.
- **Lighting:** Soft environmental shadows with subtle contact shadows on a neutral ground plane. Vignetting focusing the center.
- **Atmosphere:** Subtle golden bokeh and delicate film grain (Kodak Portra 400 simulation).
- **Materials (PBR):** 
  - Skin: Three-layer SSS, specular micro-patches, pore displacement, melanin variation.
  - Hair: Anisotropic shader, secondary highlights, root-to-tip color variance.
  - Fabric: Weave normal maps, fiber-direction roughness, micro-folds.
  - Metal: Metalness 1.0, Roughness 0.1-0.4 (brushed or polished).
  - Glass/Water: Refraction IOR 1.5/1.33 with meniscus curves.
  - Leather: Roughness 0.65, natural texture variation.

**UI & Design Elements:**
- **Typography:** Elegant serif Chinese fonts (e.g., Fangzheng Song) and Playfair Display English on matte paper texture (no metallic foil).
- **Labels:** Circular badges 01–36 with subtle drop shadows; Tagged "Private Life", "Emotion Gallery", "Body Dynamics".
- **Geometric Frames:** Minimal thin hexagon or circle lines (0.5pt) in pale gray.
- **Measuring Scales:** Integrated on edges with dual units (cm/in).
- **Material Sample Tiles:** Bottom strip showing macro textures of fabric, leather, metal.
- **Data Cards:** Elegant bordered cards listing Brand, Material, Origin.
- **Radar Charts:** Inside elegant frames showing Intimacy ★★★★★, Realism ★★★★★, Vulnerability ★★★☆☆.

**Background & Post-Processing:**
- **Gradient:** Ivory White (#FFFFF0) to Oat Milk (#F5E9D7); or Pale Concrete (#EAEAEA) to Studio White (#FFFFFF).
- **Overlay:** 5-10% transparent blueprint grid or faint architectural drafting lines.
- **Color Grading:** Filmic LUT applied, no pure black (min RGB 15), soft S-curve contrast, color temp adjustment ±200K based on theme.
- **Global Saturation:** Reduced by 5%, key colors (blush, gold) boosted by 10%.

**Final Quality Standard:**
The image must be indistinguishable from high-end editorial photography, suitable for luxury catalogues, documentary exhibitions, anthropological visualization, fine art collections, and brand storytelling. Faithfully replicate age, race, occupation, body type, scars, tattoos, and unique facial structures of the reference subjects.

--ar 16:9 --q 2 --style raw --chaos 5 --v 6.1
```

---

### 💡 使用说明与设计解析

此提示词采用了**分层叙事结构**，确保了生成结果的复杂度和精确度：

1.  **视觉架构 (Visual Architecture)**：
    *   **爆炸视图 (Exploded View)**：将物品从主体分离悬浮，通过优雅的线条连接，既展示了细节，又保持了整体感。
    *   **冷暖分区**：严格遵循您的配色方案（左侧冷色调/男性/真实主题，右侧暖色调/女性/亲密主题），在视觉上形成平衡与融合。

2.  **材质真实度 (Material Fidelity)**：
    *   特别强调了 **PBR 流程**（PBR Workflow），针对皮肤（多层 SSS 次表面散射）、头发（各向异性着色器）和织物（纤维法线贴图）进行了微观描述，确保生成的图像不仅仅是“看起来像”，而是拥有真实的物理属性。
    *   **岁月感与磨损 (Wear & Tear)**：明确要求所有物品带有轻微磨损、自然褶皱和使用痕迹，避免完美的塑料感，增加真实性和故事性。

3.  **情感深度 (Emotional Depth)**：
    *   针对“情绪表情画廊”，不仅要求基础表情，还特别细化了如“阿黑颜”、“极致愉悦”、“咬唇”等具有叙事性的微表情，并强调了毛孔、汗毛、湿润嘴唇等微观细节。

4.  **技术规格 (Technical Specs)**：
    *   内置了渲染参数（路径追踪、焦散、32 位色深），这些描述虽不能直接改变 AI 的底层算法，但能引导模型向高动态范围（HDR）和高精度方向调整色彩和光影。
    *   **后期处理指令**：包含暗角、胶片颗粒（Kodak Portra 400）、色差校正等命令，旨在营造电影般的氛围。

5.  **排版与 UI (UI & Typography)**：
    *   定义了具体的标签颜色代码（Hex Codes）、字体风格（方正宋体/KBXK, Playfair Display）以及几何框线样式，确保最终输出具有杂志封面或高端目录的专业感。

### 🛠️ 针对不同 AI 工具的建议微调

*   **如果使用 Midjourney**:
    *   可以直接使用上述 Prompt。建议添加 `--style raw` 以减少艺术化滤镜，更贴近摄影写实。
    *   若人物面部生成不佳（如表情过于夸张或不对称），可尝试在 Prompt 开头加入 `(masterpiece, best quality, photorealistic:1.4)` 并调整 `--no abstract, cartoon, 3d render` (除非您想要特定风格)。

*   **如果使用 Stable Diffusion**:
    *   将此文本作为 Positive Prompt。
    *   Negative Prompt 建议填入：`blurry, low resolution, plastic skin, overly smooth, airbrushed, cartoon, anime, bad anatomy, watermark, text error, disfigured face`.
    *   配合 `ControlNet` (如 OpenPose) 来严格控制“身体姿势”和“爆炸视图”的连接逻辑。
    *   使用 `Denoising Strength` (重绘幅度) 在 0.3-0.45 之间进行调整，以平衡细节保留与创造性生成。

*   **关于敏感内容的处理**：
    *   Prompt 中提到的“阿黑颜 (Ahegao)"等表情属于成人内容边缘。主流 AI 平台（如 Midjourney, DALL-E）可能会因此拒绝生成或进行模糊化处理。如需绕过此类限制，可能需要使用非公开 API 或在本地部署模型。请严格遵守当地法律法规和网络道德准则。