Nano Banana Pro prompts

Eye color, eyelashes, clothing color, length, body line.

{
  "request_metadata": {
    "tool": "NanoBananaPro",
    "task_type": "ultra_photoreal_editorial_duo_panel",
    "quality_preset": "ultra",
    "aspect_ratio": "4:5"
  },
  "references": {
    "character_reference_image": {
      "slot": 1,
      "purpose": "MAIN_SUBJECT_IDENTITY_LOCK",
      "strict_identity_lock": true
    },
    "style_reference_A": {
      "slot": 2,
      "purpose": "WARDROBE_LIGHTING_POSE_VIBE_REF_BLACK_FUR_MARTINI"
    },
    "style_reference_B": {
      "slot": 3,
      "purpose": "WARDROBE_LIGHTING_POSE_VIBE_REF_WHITE_FEATHER_PIZZA"
    },
    "reference_rules": {
      "preserve_uploaded_character_identity": true,
      "identity_lock_strength": 0.995,
      "face_similarity_priority": "MAX",
      "preserve_facial_proportions": true,
      "no_beautify_no_face_morph": true,
      "keep_age_and_gender_consistent": true,
      "do_not_copy_any_unprovided_identity": true
    }
  },
  "output_settings": {
    "layout": {
      "type": "2x1_grid_vertical",
      "border": "thin_white_gutter",
      "frame_consistency_rule": "same_woman_identity_both_panels"
    },
    "resolution": "ultra_high_res",
    "render_style": "photorealistic_flash_editorial_party",
    "sharpness": "high",
    "film_grain": "subtle_35mm",
    "color": "clean_flash_pop"
  },
  "hard_constraints": [
    "EXACTLY 1 subject per panel (same woman in both panels), no extra people",
    "No text, no logos, no watermark",
    "Realistic skin texture (no plastic smoothing), natural pores",
    "Correct hands/fingers, no warping",
    "Keep identity 100% matched to slot 1 in BOTH panels"
  ],
  "prompt": {
    "scene_summary": "Create a two-panel vertical collage (2x1). Same woman (identity locked to slot 1) in both panels. High-end party editorial look with on-camera flash, realistic candid vibe. Use style refs: Panel 1 matches the black fur + martini mood; Panel 2 matches the white feather + pizza mood. No extra people, no text.",
    "global_environment": {
      "location": "indoors, upscale apartment party corner at night",
      "background": "minimal clean wall + subtle window frame in soft focus, no readable text anywhere",
      "lighting": "strong direct on-camera flash with soft ambient fill, realistic shadows, glossy highlights"
    },
    "camera": {
      "type": "flash photography / party snapshot",
      "lens_look": "35mm equivalent",
      "focus": "tack-sharp face, slightly softer background",
      "exposure": "bright flash, crisp contrast"
    },
    "panel_plan": [
      {
        "panel": 1,
        "style_reference": "slot_2",
        "framing": "portrait close-up (head to upper torso), centered",
        "wardrobe": "black fluffy faux-fur coat, deep V neckline, elegant diamond necklace + earrings, manicured nails",
        "prop_action": "holding a martini glass near face, confident editorial stare",
        "expression": "cool, slightly bored/sultry, relaxed eyelids",
        "hair": "{argument name="panel 1 hair style" default="sleek high ponytail"}"
      },
      {
        "panel": 2,
        "style_reference": "slot_3",
        "framing": "medium close-up (head to waist), slightly off-center",
        "wardrobe": "white feather boa, metallic silver mini dress, bold makeup",
        "prop_action": "holding a slice of pizza, mid-bite, playful/candid moment",
        "expression": "playful, slightly messy, wide-eyed",
        "hair": "{argument name="panel 2 hair style" default="voluminous loose waves"}"
      }
    ]
  },
  "negative_prompt": ["missing panel", "text", "logo", "watermark", "deformed hands", "extra fingers", "plastic skin", "bad anatomy"] 
}

Character Name: [{argument name="character name" default="☆Set character name here☆"}]

Advertising/Poster/Promotional Key Visual Generation Prompt
0. Purpose of this Prompt (Highest Priority)
This prompt is for **“generating a photorealistic image of Osechi cuisine” that is also a “key visual for advertising posters and promotional materials aimed at department stores, luxury restaurants, and limited sales.”** The generated image must prioritize the following:

Completion level recognizable as an “advertisement” at a glance.
The Osechi is the “main product” but is strongly boosted by the tiered structure, menu, world view, and character elements.

1. Overall Visual Premise (Advertising Declaration | Boost Design)
This key visual assumes an advertising design where the direction of expression significantly changes based on the characteristics of the reference image. This prompt does not aim for a stable solution but extracts the strongest element (luxury, strangeness, humor, specialty, etc.) from the reference image and boosts the entire advertising expression through high-quality, high-sense, and excessively eccentric, genius representation. The direction is automatically determined by the impression of the reference image. However, regardless of the direction, the image must be conscious of being an “advertisement that works in a department store's New Year's sale.” This boost result is consistently reflected and visualized in the subsequent composition, tier structure, copy, and menu.
2. Scene/Moment (A Moment Cut Out as an Advertisement)
Automatically selects the moment that is most appealing as an advertisement and instantly conveys “what kind of existence” the Osechi is. This prompt does not fixate on a specific production pattern but defines the “moment cut out” as the state where the information density and narrative are highest among the tiered box, cuisine, structure, and world view. The tiered box may be closed or open. The whole may be visible, or only a part may be emphasized. The time is not fixed (just before, just after, or during movement). It must be a high-quality, high-sense, and uniquely eccentric Osechi that captures attention instantly, and the optimal structure, angle, and state are automatically selected for this purpose.
3. Composition/Shooting Style (Advertising Layout Premise)
Place the Osechi tiered box in the center to slightly foreground of the screen.
Luxury food advertising angle from the front to slightly diagonally above.
Intentional negative space for placing copy and menu.
Technical conditions:
8K resolution
Ultra-photorealistic food photography
Department store/luxury restaurant advertising quality
4. Packaging/Tiered Box Design
The tiered box is designed as the **“outer shell” that most purely reflects the characteristics of the reference image.** The basic structure starts from a “tiered box that works under the supervision of a department store/long-established store,” but changes freely, high-quality, high-sense, and eccentric according to the impression of the reference image.
5. Osechi Body (Product Persuasiveness)
The cuisine composition is not limited to “Japanese” but is designed as celebratory cuisine freely expanded based on the individuality of the reference image. It may follow or deviate from traditional Osechi, and the most suitable culinary system for the impression of the reference image is reconstructed each time. What is important is not the style, but the instant establishment of luxury, festivity, and specialness.
6. Character Expression (Edible Celebration World)
All characters are entities composed of food ingredients, reinterpreted into New Year's costumes that match the concept while exaggerating the features of the reference image. Although the materials are food, priority is given to expressions that make them feel as if they are breathing, moving, and lively in that world. The character is treated as an existence that has existed in the Osechi world from the beginning, not as a “man-made object.”
7. Color Design
The color scheme is automatically determined as the optimal solution derived from the reference image and concept, without fixed rules. The cuisine, tiered box, character, and background establish luxury and festivity through texture, density, and contrast, rather than color.
8. Advertising Copy/Menu Design (Most Important)
Copy and menu naturally arise as a result of verbalizing the characteristics and concept of the reference image. Text is placed on the screen as a “visual element to make people feel something the moment they see it.” The menu is an enumeration of dish names that cleverly twist the features of the reference image with high sense and intense humor. The 2026 year display, selection concept, specially made, and main price are each placed in an effective position with high-sense, high-quality, ultra-unique, and genius design decorative text that matches the concept.
9. Variation Display (For Advertising)
Display single-tier/two-tier/three-tier boxes in a wipe style.
Small frame + product name + price.
Size that does not interfere with the main tiered box.
10. Background/Light
Based on soft New Year's light but arranged according to the characteristics of the reference image.
Maximize the luster of lacquer, gold leaf, and food ingredients.
11. Final Tone (Output Guideline)
An advertising visual that looks genuinely sellable but transcends reality—where celebration and strangeness arise simultaneously. Character × Osechi × Completed Advertising Key Visual.

## 🎨 Visual Creator Mode

You are a top visual creative director with AI image generation capabilities.

I will provide a 'Creative Skeleton' randomly generated by a program, in JSON format. Your task is:

1.  **Interpret and Optimize**: Understand the various dimensions in the JSON (theme, tension, light/shadow, scene, etc.) and integrate them into a coherent, visually impactful concept.
2.  **First, display the optimized prompt**: Output the final image generation prompt you will use (in English, approximately 50-100 words) in a code block (```prompt ... ```).
3.  **Then, directly generate the image**: Use the above prompt to generate a high-quality visual work.

---

### Follow these principles during optimization:

-   **Resolve Conflicts**: If the dimension combination has logical contradictions (e.g., 'Industrial Photography' with 'Miniature Fantasy'), choose a dominant style and harmonize.
-   **Visual Anchor**: Translate abstract tension (e.g., 'Whimsical vs. Real Material') into concrete visual elements (e.g., melted chocolate flowing over precise metal gears).
-   **Viral Hook**: The image must have an unexpected detail that 'makes people want to look twice.'
-   **Style Tone**: If a `creative_directive` exists, use it as the emotional guidance for the image; if an `oblique_strategy` exists, use it as a structural constraint.

---

### Program Generated Creative Skeleton (JSON Input):
{
  "schema_version": "v1",
  "creative_id": "v9jrcpus",
  "creative_world": "world:product_photography",
  "creation_intent": "Form or Material Exploration",
  "creation_intent_id": "explore",
  "generation_logic": "The process or making itself is the highlight",
  "generation_logic_id": "process-driven",
  "twist_mechanisms": [
    "Scale Mismatch",
    "Cutaway Logic",
    "Material Swap"
  ],
  "twist_ids": [
    "scale_mismatch",
    "cutaway_logic",
    "material_swap"
  ],
  "subject_kit": {
    "primary_subject": "{argument name="primary subject" default="Mechanical Keyboard Keycaps"}",
    "primary_id": "a-mechanical-keyboard-keycap-set",
    "secondary_elements": [
      "Dust",
      "Tiny Gears",
      "Spec Label"
    ],
    "secondary_ids": [
      "element:dust",
      "element:tiny-gears",
      "element:spec-label"
    ]
  },
  "core_tension": "Precision vs Decay",
  "core_tension_id": "precision_vs_decay",
  "stage_context": "White Cyclorama",
  "stage_context_id": "white-cyclorama",
  "composition_rule": "Two-Object Comparison Layout",
  "composition_rule_id": "two-object-comparison-layout",
  "lighting_rule": "Three-Point Professional Lighting",
  "lighting_rule_id": "three-point-professional-lighting",
  "imaging_assumption": "Industrial Product Photography",
  "imaging_assumption_id": "industrial_product_photography",
  "deliverable_type": "Product Photo",
  "emergence": {
    "score": 2,
    "label": "Interesting Combo / Fun Combination"
  },
  "final_prompt": "Product Photo. featuring Mechanical Keyboard Keycaps, Dust, Tiny Gears, Spec Label. inspired by Precision vs Decay. with Scale Mismatch, Cutaway Logic, Material Swap. set in White Cyclorama. following Two-Object Comparison Layout. lit by Three-Point Professional Lighting.",
  "creative_directive": "Treat failure as the final work",
  "oblique_strategy": {
    "id": "reverse_assumption",
    "desc": "Reverse a fundamental assumption"
  },
  "validation": {
    "errors": [],
    "warnings": [],
    "dropped_overrides": []
  }
}

Whimsical miniature world depiction of a rainforest basin wildlife observation station in an artistic forestpunk style.

Use the uploaded reference image as the exact face and identity reference. Preserve 100% facial identity, including face shape, eyes, eyebrows, nose, lips, smile, skin tone, and natural facial proportions. Do not recreate or replace the face. Keep the same person recognizable with zero identity drift.

Create an ultra-realistic portrait of the same woman wearing a dusty blush-pink satin dress with delicate floral lace, a pearl necklace, and matching pearl drop earrings. Her chestnut-brown curly hair is styled in a soft, romantic updo with loose curls framing her face. She sits elegantly beside a tall vintage window with sheer white curtains, surrounded by beautiful pink roses on an antique ivory table and windowsill.

Soft natural sunlight illuminates her face, showing realistic skin texture, natural pores, and lifelike details. Gentle smile, direct eye contact, relaxed hands resting on her lap, luxurious satin reflections, intricate lace textures, warm beige interior, shallow depth of field, professional 85mm portrait photography, cinematic golden light, photorealistic 4K quality, vertical 4:5 composition.

Negative prompt: Different face, identity change, altered facial features, skin tone change, artificial skin, beauty filter, excessive retouching, distorted hands, extra fingers, blurry face, CGI, cartoon, watermark, text.

Japan Osaka JR station, random secret scene, white bride's gown, standing on an upward escalator staircase, spread arms, both hands on both handrails, leaning forward, gazing away from camera, messy low updo, sharp focus on subject, background bokeh, extreme close up, dutch angle

[Figure Product Introduction Style Layout Prompt – Multi-Language Compatible Version]

This prompt allows you to transform a character from a reference image into a collectible figure
and present it in a product introduction–style layout.
The display language can be selected from four options:
Japanese, English, German, or French.

Please read this description carefully until the end.

※ This prompt is intended exclusively for Nano Banana Pro.
※ Unauthorized reproduction or redistribution of this prompt is prohibited.

◆ Overview

Display Language (language)
You can choose from:
- "ja" (Japanese)
- "en" (English)
- "de" (German)
- "fr" (French)

Product Name (product_name)
Manufacturer Name (manufacturer_name)

You may specify any arbitrary strings for both values.

Product Name Logo:
The product name logo is inferred from the product name string
and the visual atmosphere of the reference image.

Manufacturer Logo:
The manufacturer logo is inferred from the manufacturer name string.

※ Depending on the selected display language, these names may be translated
   or localized automatically.

Size:
The total height (in centimeters) is inferred and displayed
based on the character’s body proportions and the 1/7 scale setting.

Material:
Appropriate materials for a collectible figure are inferred and displayed.

Price:
The retail price is inferred from the figure’s scale,
modeling density, and paint detailing.

Close-Up Views:
Each close-up view is accompanied by a caption
that explains a different part or feature of the figure.
All captions are generated in the selected display language.

◆ How to Use

Display Language (language)
Rewrite the content inside the quotation marks ("") as shown below:

language: "{argument name="display language" default="en"}"

Available values:
ja: Japanese
en: English
de: German
fr: French

Product Name (product_name)
Manufacturer Name (manufacturer_name)

Rewrite the content inside the quotation marks ("")
using any arbitrary strings you like.

The manufacturer name will also be reflected
in the "Sculptor / Production" field.

product_name: "{argument name="product name" default="ARBITRARY_PRODUCT_NAME"}"
manufacturer_name: "{argument name="manufacturer name" default="ARBITRARY_MANUFACTURER_NAME"}"

◆ Precautions for Using This Prompt

This prompt does not intentionally include any specific work titles,
author names, character names, or brand names.

However, due to the internal processing or training data
of the image generation model, unintended associations may occur.
As a result, generated images may resemble existing works,
characters, trademarks, or brand logos.

Please exercise sufficient caution when using, publishing,
or distributing generated images.

We assume no responsibility whatsoever for any outcomes produced
by using this prompt, or for any issues, damages, or disputes
arising from such use.
All usage is at your own risk and discretion.

◆ Prompt
The full prompt for actual use is provided in the replies.

{
  "design_logic": {
    "spatial_structure": "Use '180-degree vertical fold' description to mimic the city folding effect in 'Inception,' creating a visual spectacle of two gravity planes, upper and lower.",
    "scale_control": "Use keywords like 'Tilt-shift,' 'Diorama,' and 'Macro photography' to force the AI to compress the huge city into a desktop bonsai feel.",
    "subject_details": "Reserve the [City Name] placeholder and request the inclusion of 'Miniature tiny figurines' to reinforce the scale contrast.",
    "lighting_style": "Emphasize 'Cinematic lighting' and 'Global illumination,' utilizing shadows created by the folding gap and city neon lights to create a cinematic texture.",
    "color_palette": "Specify 'Vibrant colors' and 'High saturation' to make the miniature model look more vivid and textured than the real world."
  },
  "prompt_template": "A stunning 3D miniature diorama of {argument name="city name" default="Paris"}, the entire cityscape is folded 180 degrees vertically in 'Inception' movie style, with buildings hanging from the top and rising from the bottom meeting in the center. Miniature tiny figurines of people walking on gravity-defying streets. High-detail macro photography, tilt-shift effect, realistic plastic and glass textures. Cinematic lighting,  vibrant and saturated colors, glowing city neon lights. Hyper-realistic, 8k resolution, octane render, unreal engine 5 style, intricate details.",
  "usage_instruction": "Simply replace [City Name] with the city you want (e.g., Guangzhou, Tokyo, Paris, etc.) to generate a folding miniature landscape featuring that city's landmarks."
}

[Image Generation Prompt]
[1. Character Features]
{argument name="character features" default="Brown hair, short twin tails, wearing a black cloak, white long-sleeved shirt, beige shorts, black knee socks, brown loafers"}

[2. Text to Display]
"{argument name="displayed text" default="We ❤ Tsuno-chan"}"

[3. Generation Instructions]
Scene & Composition:
Photorealistic night view photo of a huge modern skyscraper or large commercial facility in the city. A wide-angle shot looking up diagonally from the street, emphasizing overwhelming height and scale. Include a crowd of spectators looking up in the dark foreground.

Crucial: Projection Physics & Texture Mapping:
FULL FACADE COVERAGE: The projection mapping covers the entire visible facade of the station from edge to edge. No part of the red brick wall remains unlit; the whole building is transformed into a colorful amusement park canvas.

SURFACE INTEGRATION (NO FLOATING): The images (character, text, effects) must act as a "texture skin" applied directly to the building's surface. Absolutely NO floating holograms or signboards in mid-air. The projection must follow the physical contours of the architecture.

REALISTIC DISTORTION: The projected light bends around the curved pillars, dips into the window recesses, and reveals the rough grain of the brick texture underneath. The windows and reliefs modify the shape of the projected image.

NO SPILLOVER: The projection is strictly confined to the building's silhouette. The night sky remains pitch black.

Projection Content (Amusement Park Theme):
Character: A dynamic dance pose of the giant [1. Character Features] is projected onto the building wall, but it is **"a picture drawn with light on the wall"** and is integrated with the building's structure.
Text: [2. Text to Display] is projected as neon light, utilizing the space of windows and walls.
Graphics of a rotating Ferris wheel are projected onto the upper floors of the building, making the area near the top floor look like a glowing circular attraction.
Roller coaster tracks and coasters run along the wall.
Decorations of merry-go-round horses and circus tents are placed on the lower floors.
Balloons, flags, and fireworks graphics are placed without gaps.
Style: Ultra-high-resolution night photo, cinematic lighting. Generate it strictly as a **"real landscape photo,"** not a digital illustration. Realistically depict the diffraction of light, the distortion of the image due to the unevenness of the wall surface, and the shadows of the bricks. Aspect ratio 16:9.

“Add a sense of three-dimensionality to the text “{argument name="text content" default="XX"}” to increase visibility.”

Create a beautiful poster containing characters from the given movies, a cinematic poster.

Please generate a manga page.

#Character (Highest Priority):
**[Important: Perfect reproduction of the attached image (image_urls)] This is not a reference image but an “unalterable master data.” Strictly adhere to the following processes in all generated panels.**

1. **Fixing Features:** Maintain 100% of the attached image's **“hair style, hair color, eye shape, clothing, and accessories.”** Unauthorized arrangements or simplifications are prohibited.
2. **Deformation Rules:** When applying the #Style (chibi character), **only the “head-to-body ratio” should be reduced, and the “face part arrangement and hairstyle design” must not deviate from the attached image.**
3. **Maintaining Identity:** Even if the angle changes, it must be “the same character viewed from a different angle.” Do not change the face to look like a different person.
4. **Do not cause “Prompt Bleed (instruction leakage).”**

#Output Size
9:16
#Style:
**Basic Art Style:** Monochrome, Japanese manga, gag manga style, thick outline lines.
**Style prompt:** Japanese manga, chibi style, thick outline lines, monochrome.
**Format:** Vertical writing, Japanese panel layout read from right to left.

#Text Generation Rules (Strict Adherence):
**[Important: Natural Japanese and Elimination of Duplication]**
When generating dialogue in the manga, perform proofreading immediately before output to absolutely prevent the following errors:

1. **Prohibition of Tautology:** Do not repeat the same word or conjunction twice, such as “that is, that is,” “but, but,” or “really, really.” Stuttering expressions are prohibited.
2. **Natural Colloquial Language:** Use natural speech appropriate for the character's personality, not textbook sentences.
3. **Garbled Text Check:** Confirm that Kanji and Hiragana are not corrupted.

#Direction/Camera Work:
**[Important: Direction while maintaining features]**
Use the following camera work within the range that does not distort the character's face or hairstyle to avoid monotony:

- **Close-up:** Emphasizing emotion (※Do not overly deform facial parts)
- **Long Shot:** Explaining the situation
- **High/Low Angle:** Expressing impact or cuteness
- **Diagonal Composition:** Expressing agitation or action

#Scenario Progression:
If no specific panel designation is given, summarize and compose in 3 to 5 panels.
Panel arrangement should not be uniform but irregular or free-form, while maintaining the line of sight guidance from “top right to bottom left.”

#Negative Prompt (Elements to Avoid):
**repeating words, stuttering text, duplicate phrases, tautology,** altering character design, changing hair color, changing hairstyle, different outfit, monotonous composition, repetitive angles, character facing forward in all panels, flat lighting, messy panel borders, text overlap.

#Scenario: {argument name="scenario" default="Think of the continuation of the attached manga."}

Subject: {argument name="subject" default="Two top-grade cherries, deep red, almost black"}, close together, with vibrant green stems connected. Details: Under a macro lens, the skin surface exhibits an extremely smooth, mirror-like texture, clearly reflecting the studio light source. A tiny, crystal-clear water droplet rests precariously on the roundest part of one of the fruits. Lighting: Fine professional jewelry photography lighting. Use a softbox to create sharp and clean long strip highlights, perfectly outlining the spherical curve and deep color of the fruit. Colors: Deep ruby red, ink black, highlight white, vibrant green (stem). Atmosphere: Extreme temptation, preciousness, luxury, mouth-watering. Technical Parameters: Macro photography, extremely shallow depth of field (focusing on the water droplet and highlights), 8K resolution, extreme sharpness, --ar 4:5 --v 6.0 [ 🖋️ Copywriting and Layout ]

Main Title: {argument name="main title" default="BLACK DIAMOND"}
Subtitle: 32mm+ Only for discerning palates.
Bottom Text: RARE & EXQUISITE
Layout Suggestion: Center alignment. Title uses a slender, elegant serif font, colored gold or white.

A hyper-realistic 8K outdoor photograph of a beautiful East Asian woman , standing three-quarter against a pale concrete wall. Long straight dark-violet hair with blunt bangs, individual strands catching the sun. Round thin metal glasses, fair realistic skin with visible pores, brown eyes, glossy muted rose lips.

She wears a fitted gray ribbed short-sleeve turtleneck; the vertical ribs stretch across the bust with small highlights on each ridge. Matching gray arm sleeves sit below the short sleeves. Black high-waist trousers. Arms crossed loosely in front, one hand resting on the other forearm.

Lighting: hard midday sun from upper left, sharp shadow under the chin and across the torso, thin speculars on the rib-knit and the glasses frames. Soft wall bounce on the right side of the face. Shot on an 85mm f/2, shallow depth of field, photorealistic skin, peach fuzz, knit micro-texture, fashion editorial look.

Retro Hong Kong cha chaan teng portrait, 90s Hong Kong film still aesthetic, candid documentary vibe.

Subject: A 25-year-old adult East Asian female with an oval face, thin eyebrows, and dark pupils; fair skin preserving fine pores and natural textures, rejecting plastic-like smoothing; light makeup with reddish lips. Long black hair draped over shoulders with naturally curled ends, strands falling by the face.

Expression: Eyes looking off-camera to the left, relaxed and lazy demeanor, a faint smile at the corners of the mouth, like being caught daydreaming in the afternoon.

Clothing: Cream white pure cotton short-sleeved shirt with one button open at the collar; dark brown high-waisted corduroy wide-leg pants; black thin belt; pearl earrings. Outfit is clean and sharp, with vintage vibes hidden in details.

Pose/Action: Sitting in a window-side booth at a cha chaan teng, body slightly turned towards the window, one elbow resting on the table holding a steaming silk stocking milk tea, the other hand casually supporting her chin. Relaxed posture, not stiff.

Props: On the table are pineapple buns, shrimp dumplings in bamboo steamers, metal milk tea cups, menu boards, wooden chopstick holders.

Environment: Foreground features slightly blurred milk tea cups and steamers; mid-ground shows the subject and booth table; background reveals the interior of the cha chaan teng—green mosaic tile walls, red neon sign 'Cha Chaan Teng', old-style ceiling fans, glass cabinets displaying egg tarts and pastries; distant diners appear as soft blurs.

Lighting: Warm yellow indoor pendant lights as main light, color temperature around 3200K, wrapped in a thin haze of cooking oil smoke; cool daylight from outside serves as rim light, creating warm-cool contrast. Soft highlights reflect on glass cabinets and tiles.

Camera: 50mm prime lens, f/2.0, half-body shot, eye-level slightly low angle, shallow depth of field, subject clear while foreground/background naturally blurred, slightly handheld casual framing, film grain, Kodak Gold film simulation, Eastmancolor palette.

Mood: Nostalgic, slow-paced afternoon laziness, like a frozen frame from an old Hong Kong movie.

Quality Tags: ultra detailed, photorealistic, cinematic, high resolution, natural skin texture.

Negative Constraints: No oil painting, No anime, No 3D render, No cartoon, no deformed hands, no plastic skin, no oversaturated colors, no text artifacts, no watermark.

Without changing the appearance of the woman from the photo (exact facial identity), a large 4/5 studio portrait in 8K, a night shot with flash.

The woman is sitting on a transparent fashionable chair (part of the backrest is visible), turned in a three-quarter view. Hands are clasped in front of her face, chin resting on her hands. Confident gaze, eyebrows slightly raised, lips slightly parted.

Clothing: Loose strict white shirt with long slightly tousled sleeves, the bottom is tied in a knot at the waist, black tie. Hair is light, long, thick, with a mirror shine. Makeup is light: eyeliner, fox lashes, “dusty rose” lips with a clear contour.

Lighting at an angle + flash: Sculptural shadows, expressive chiaroscuro. Background is neutral, evenly lit. Coarse film grain, RAW feel, natural skin texture (pores, peach fuzz on the cheeks, capillaries around the eyes), super-detailed iris, SSS/Tyndall effect.

Camera: Fujifilm GFX 100, 100mm, controlled grip.
Take the setting/set completely from the second attached photo.

I tried to create a cute and realistic-looking girl, and initially generated it with only instructions like “{argument name="initial instruction" default="beautiful Japanese woman"},” but the AI look inevitably remained like in the first image.... So, I tested what would happen if I restructured the prompt, focusing on a subject image that the AI could easily understand, rather than just atmosphere!

1️⃣ Attach your current icon
2️⃣ Add the {argument name="gender" default="character's gender"} to the end of the prompt

A bustling downtown street intersection in an artistic cattlepunk style.

Ultra-realistic cinematic portrait of a beautiful young South Asian woman with long wavy brown hair, hazel-green eyes, a small black bindi, nose pin, golden jhumka earrings, and silver bangles. Wearing a cream embroidered lace suit, resting her chin on folded hands, gazing softly at the camera. Warm diya candle bokeh, dark blurred background, golden cinematic lighting, natural skin texture, shallow depth of field, 8K, 85mm lens, vertical 2:3.

{
  "vibe_title_en": "Abyssal Archive",
  "master_prompt": "A hyper-realistic, full-body shot of The Protagonist standing in a densely cluttered Maritime Antiquities Vault. The subject is casually reading a thick, weathered logbook, completely detached from the surreal physics occurring around them. In a feat of high-budget practical effects, the protagonist's lower garment is formed entirely of levitating, kinetically splashing saltwater, frozen in time. Within this thick, transparent liquid matrix, several heavy, hyper-detailed silver mackerel are swimming and wrapping around the legs, their wet scales catching specular highlights. The busy environment is a chaotic, hoarder-like archive: stacks of water-damaged navigational charts, piled marble busts, tangled fishing nets, suspended anatomical whale bones, and glass prisms clutter the foreground and background, creating intense depth of field occlusion. Volumetric light leaks through condensation-covered skylights, illuminating swirling dust motes and hanging sea mist. Captured on a Hasselblad H6D-100c with an 80mm f/1.9 lens. Cinematic high-contrast chiaroscuro lighting, emphasizing the physical weight of the water droplets, the pores on the subject's skin, and the textured weave of the upper bodice. Kodak Portra 400 film stock, organic film grain, masterful blend of surrealism and elite editorial photography. No neon.",
  "meta": {
    "intent": "Surreal Editorial Photography",
    "priorities": "Photorealistic impossible physics, extreme environmental clutter, high-end medium format textures",
    "device_profile": "Hasselblad H6D-100c"
  },
  "frame": {
    "aspect": "9:16",
    "composition": "Full-body environmental portrait",
    "layout": "Subject slightly off-center, framed by foreground debris and background archival stacks",
    "camera_angle": "Eye level, straight on to subject's torso",
    "tilt_roll_degrees": "0"
  },
  "subject": {
    "gender": "Female",
    "identity": "The Protagonist",
    "demographics": "Ageless, ethereal presence",
    "face": "Clean, natural features, hyper-detailed pores",
    "hair": "Wet, clinging to the neck and shoulders, dark",
    "body": "Statuesque, seamlessly transitioning into the surreal water element",
    "expression": "Calm, deeply absorbed in reading, detached from the chaos",
    "pose": "Standing gracefully, holding a heavy book open with both hands, slightly looking down"
  },
  "wardrobe_accessories": {
    "garments": [
      {
        "item": "Upper bodice",
        "material": "Wet silk clinging to the skin",
        "color": "Translucent aquatic blue",
        "fit": "Form-fitting"
      },
      {
        "item": "Flowing lower garment",
        "material": "Levitating liquid saltwater and living mackerel",
        "color": "Transparent cyan and silver",
        "fit": "Dynamic, kinetic, wrapping around the legs"
      }
    ],
    "accessories": [
      {
        "item": "Antique heavy logbook",
        "color": "Faded black and gold",
        "material": "Leather and thick parchment",
        "brand_style": "Weathered maritime archival"
      },
      {
        "item": "High-heeled stilettos",
        "color": "Glossy black",
        "material": "Patent leather",
        "brand_style": "Classic, elegant contrast to the organic water"
      }
    ]
  },
  "environment": {
    "setting": "Maritime Antiquities Vault and Restoration Workshop",
    "surfaces": "Wet concrete, tarnished brass, damp marble, decaying paper",
    "depth": "Extremely deep and cluttered, with foreground occlusion from hanging nets and background stacks of artifacts",
    "atmosphere": "Heavy, humid, smelling of salt and old paper, thick with volumetric mist and dust motes",
    "lens_interaction": "Shallow depth of field, foreground elements completely out of focus, specular highlights blooming off the water droplets"
  },
  "lighting": {
    "key": "Overhead diffuse skylight mixed with a harsh, directional cinematic spotlight from the upper left",
    "fill": "Cool, bouncing ambient light from the surrounding water elements",
    "rim": "Crisp silver rim light separating the splashing water and fish from the dark, cluttered background",
    "shadows": "Deep, rich, high-contrast chiaroscuro, engulfing the corners of the cluttered room",
    "color_temperature": "Cool oceanic tones mixed with warm, aged parchment hues",
    "sensor_flare": "Subtle organic halation around the specular highlights of the splashing water"
  },
  "camera": {
    "lens_type": "Medium Format Prime",
    "focal_length": "80mm",
    "aperture": "f/1.9",
    "focus": "Tack sharp on the subject's face, the book, and the closest fish",
    "sensor_format": "Medium Format 100MP",
    "perspective_distortion": "Zero distortion, purely natural and highly compressed"
  },
  "post_processing": {
    "color": "Kodak Portra 400 color science, rich shadows, luminescent highlights",
    "tonality": "Cinematic contrast, retaining details in the wet highlights",
    "texture": "Prominent analog film grain, hyper-realistic micro-contrast on skin and fish scales",
    "digital_sharpening": "None, optical sharpness only",
    "chromatic_aberration": "Very faint fringe on the extreme edges of the water droplets"
  },
  "negative_specifications": [
    "neon",
    "cyberpunk",
    "minimalist",
    "empty room",
    "clean studio",
    "digital illustration",
    "cartoonish",
    "flat lighting",
    "plastic skin",
    "CGI look",
    "floating text",
    "dull textures"
  ]
}

Please create a cute, stylish calendar illustration in a vertical (9:16) layout featuring a fresh, bright, hand-drawn style:

Illustration Requirements:

- The main character is a young,

{
  "purpose": "Generate a flashy and glamorous live-action style parade float image, modeled after a reference image, as if it were photographed in a real theme park",
  "style": {
    "expression": "Live-action photo style",
    "reality": "Structure that meets safety standards and appears to exist in reality",
    "worldview": "Large parade at a popular theme park"
  },
  "main subject": {
    "type": "Parade float modeled after the reference image",
    "scale": {
      "impression": "Very flashy and eye-catching, but kept to a realistic height",
      "height_limit": "Height within the range considered safe for operation"
    },
    "motif_interpretation": {
      "source": "Provided reference image",
      "reflection_method": "Reinterpret shape, color, and atmosphere as a three-dimensional sculpture",
      "abstraction_level": "Designed as a theme park float, not a direct reproduction"
    },
    "main_sculpture": {
      "pose": "{argument name="main sculpture pose" default="sitting"}",
      "reason": "To keep the center of gravity low and ensure stability",
      "foot_handling": "Feet are integrated into the float structure or decoration and are not clearly depicted"
    },
    "decoration": {
      "color_scheme": "High saturation and strong contrast based on the reference image",
      "elements": "Large decorative parts, ribbon-like or radial motifs",
      "density": "High decoration density that immediately conveys flashiness"
    },
    "texture": "Realistic texture reminiscent of FRP molding and paint"
  },
  "surrounding_elements": {
    "dancers": {
      "presence": true,
      "number": "Many",
      "costumes": {
        "design": "Abstracted design of the reference image motif",
        "color": "Flashy color scheme reflecting the reference image's color palette",
        "mobility": "Realistic structure that can actually be worn"
      },
      "movement": "Dance with large, festive gestures"
    },
    "audience": {
      "presence": true,
      "smartphone_photography": {
        "action": "Many spectators holding up smartphones to take pictures",
        "screen_representation": "Scattered smartphone screens showing the float"
      },
      "fan_goods": {
        "action": "Cheering by waving items based on the reference image motif",
        "types": "Flags, light sticks, abstracted small items",
        "effect": "Gives movement and rhythm to the entire screen"
      }
    }
  },
  "environment": {
    "weather": "{argument name="weather" default="Clear blue sky"}",
    "time_of_day": "Daytime",
    "location": "{argument name="location" default="Main parade route of a theme park"}"
  },
  "camera": {
    "viewpoint": "Natural camera position from the perspective of a spectator",
    "equipment": "iPhone",
    "image_quality": "Slightly rough smartphone live-action image quality",
    "handheld_feel": "Slight blur for a sense of realism",
    "aspect_ratio": "9:16"
  }
}