图片提示词

{
  "subject_pose": "A young woman standing in a relaxed, confident pose, leaning her back against the wooden balcony railing. She has one leg straight and the other leg bent at the knee, resting her foot slightly behind her on the lower rail structure. Her body is angled slightly towards the camera, and one hand is holding a bag handle down by her side.",
  "outfit": "She is wearing a glamorous all-white winter set consisting of a soft, fluffy white fur crop top that reveals her midriff, and a matching white mini skirt with a thick, plush fur trim at the hem. She is also wearing a large, round white fur hat that covers her forehead.",
  "background": "The scene is set on a wooden balcony with a breathtaking winter view. Behind the woman, there are tall pine trees heavily covered in snow and a majestic, snow-capped mountain range stretching into the distance under a cold sky.",
  "lighting": "Soft, natural ambient light typical of a winter evening, creating a cool tone with no harsh shadows, highlighting the texture of the white fur.",
  "atmosphere": "Cold, luxurious, and cozy winter aesthetic with a high-fashion influencer vibe."
}

An astonishing hyper-realistic 8K photograph capturing the moment {argument name="character name" default="Zenitsu Agatsuma"} from "{argument name="manga name" default="Demon Slayer"}" steps out of a giant physical manga book.

Zenitsu in the real world:
Zenitsu's image is photorealistic, with messy gold and orange hair, wearing a realistically textured Demon Slayer Corps uniform and his iconic yellow haori with white triangle patterns. His expression is determined, his eyes sharp, full of resolve, tightly gripping his Nichirin blade in its sheath. Half of his body has stepped onto the wooden bedroom floor, while the other half remains within the manga pages.
The giant manga book and interaction:
A giant, thick Japanese edition of "Demon Slayer" manga book is spread open on a slightly worn wooden floor. Zenitsu breaks the boundary of the paper, and the pages are violently torn, curled, and burst where he steps out, with paper scraps and toner flying in the air.

Details within the manga page (picture-in-picture):
The pages display an extremely authentic black-and-white Japanese manga style.
Style: Rough black ink lines, rich screentone textures for shading, exaggerated speed lines.

Layout: Strict Japanese right-to-left reading order, a surreal four-panel manga (yonkoma) spread.
Content: Inside the broken panel, a trapped, manga-style Zenitsu can be seen. He is wearing the same clothes, shedding noodle tears in the panel, frantically knocking on the borders of the manga panel, accompanied by large Japanese onomatopoeia like "出られない!" (Can't get out!), "えええ!?", and "ドンドンドン" (Knocking sound). The manga version of him is watching the real-world version of himself tear the frame in terror.

Environment and lighting:
The bedroom background is soft-focused, with domestic items like bed legs and bookshelves. Natural afternoon light streams in from the side window, illuminating the edges of the torn paper and the fabric texture of Zenitsu's clothes, emphasizing the strong contrast between reality and the two-dimensional world.

[Purpose] Define the role first
[Ratio] Determine the composition
[Product] Only write key materials
[Background] Specific color, not vague
[Lighting] Uniform natural soft light
[State] Focus on the main subject
[Text] Minimalist, does not distract from the image
[Constraint] Commercial use priority

Create a historical diagram-style infographic, 16:9 horizontal composition, titled "Visual Deconstruction of the Ancient Chinese Official Attire Ranking System." Adopt the aesthetic of ancient canonical diagrams and draw in the style of illustrations from ancient books. The image should display a comparison of the official uniforms of nine ranks of officials from the {argument name="dynasty" default="Qing Dynasty"}, arranged from left to right by rank: nine standing figures from the first to the ninth rank. Above each figure, use fine-line painting (Gongbi) to label the "Buzi (Rank Badge) Pattern" (First Rank Crane, Second Rank Golden Pheasant, Third Rank Peacock... Ninth Rank Quail; civil officials use flying birds, military officials use fierce beasts), with a close-up magnifying glass effect highlighting the embroidery details of the Buzi. Below the figures, label the "Chaozhu (Court Bead) Material" (First Rank Eastern Pearl, Second Rank Coral, Third Rank Sapphire... decreasing by rank), "Dingdai Hualing (Hat Ornament)" (Ruby/Coral/Sapphire/Crystal, etc., distinguished by color blocks), and "Mangpao (Dragon Robe) Pattern" (Nine Dragons/Five Claws/Four Claws, using diagrams to illustrate the difference in claw count). The top of the image features a "Rank Identification Quick Reference Chart": a table comparing the differences across 5 dimensions (Buzi, Dingdai, Chaozhu, Mangpao, Waist Belt) for all 9 ranks, with each cell marked by a small icon and text. The bottom of the image features "Historical Trivia": labeling "Why civil officials used flying birds and military officials used fierce beasts" (civil service requires soaring high, military service requires bravery), "The political meaning of the Hualing" (being granted a Hualing signifies imperial favor), and "Color Restrictions" (commoners were forbidden from wearing yellow or bright yellow). The very bottom includes a timeline showing the "Evolution of Official Attire through Dynasties": thumbnail sketches illustrating the style changes from the Han, Tang, Song, Yuan, Ming, and Qing dynasties. All text must be in Chinese: the top title uses the Small Seal Script font "Official Attire Rank Diagram: The Power Code Worn on the Body," and the bottom caption states "The clothing system is the visual representation of the feudal hierarchy." Use traditional official attire colors for the palette. The overall presentation should focus on historical institutions and clothing culture, suitable for historical education and cultural research.

Use this character to explain the {argument name="product name" default="Nano Banana Pro"} in a 4-panel comic!

{
  "subject": {
    "core": "{argument name="subject" default="bright orange low-profile streamlined sports car"}",
    "action_pose": "accelerating rapidly along a right-upward spiral arc, body slightly leaning forward, wheels showing motion blur"
  },
  "environment_set_design": {
    "overall_environment": "aerial spiral highway above New York City",
    "details_props": [
      {
        "object": "spiral asphalt suspended racetrack",
        "surface": "gray asphalt texture with white arrows and road markings",
        "position": "foreground extending diagonally to the right and upward"
      },
      {
        "object": "stationary skyscraper clusters",
        "arrangement": "vertically mirrored symmetrical layout along horizontal central axis, upper half inverted skyscrapers and blocks linked with lower half upright city buildings",
        "density": "dense geometric architectural blocks on both sides"
      }
    ]
  },
  "composition_camera": {
    "camera_angle": "high aerial top-down view",
    "composition_logic": "spiral highway and white guide lines form diagonal leading lines, main subject positioned at right golden ratio section",
    "shot_type_focal_length": "medium close-up, equivalent 35-50mm mid focal length",
    "focus_point": "front half of car body and front wheels, background buildings moderately blurred to emphasize subject"
  },
  "atmosphere_mood": {
    "overall_mood": "dynamic, tense, cinematic realism",
    "time_of_day": "morning",
    "weather_air": "clear sky, warm sunlight, no clouds"
  },
  "lighting_color": {
    "lighting_setup": "key light is direct sunlight from front-side, city buildings reflect natural fill light",
    "light_color": "natural daylight white with warm tint"
  },
  "photography_technical": {
    "aperture": "f/1.4",
    "camera": "ARRI ALEXA Plus",
    "lens": "Zeiss prime lens"
  },
  "art_style": {
    "reference": "{argument name="art style" default="Blade Runner 2049"}",
    "render_quality": "cinematic still frame archive style"
  },
  "color_grading": {
    "grading_strategy": "overall cool blue tone with warm golden highlights, shadows in charcoal gray to enhance contrast",
    "film_simulation_lut": "cinematic LUT with cool tones and neutral darks",
    "highlight_tone": "cool blue",
    "shadow_tone": "deep blue and charcoal gray",
    "saturation_contrast": "medium saturation overall, high contrast"
  },
  "negative_prompts": {
    "avoid": ["noise", "film grain", "clouds"]
  }
}

(masterpiece, best quality, photorealistic:1.3),
A whole watermelon carved in the center with [ {argument name="character description" default="brown hair, short twin tails, red eyes, slightly pointed teeth (giza teeth)"} ],

(oblique angle:1.4), (3/4 view:1.3), (side view:1.2),
Composition capturing the watermelon from an oblique front angle, stem not visible,

(Intricate Thai fruit carving style:1.4),
(sculpted eyes made of watermelon skin:1.6),
(eyes are expressed only by carving lines and negative space:1.5),
(no color in eyes, monochromatic eyes:1.5),
The eye area is expressed only by the remaining skin and the step difference of the white part,
Eye color is the natural color of the watermelon material itself, without color,

(deep carving exposing red flesh in the background patterns:1.5),
Gradation of deep-carved red flesh in the pattern areas,
Contrast of three colors: green skin, white mesocarp, and red flesh,
Elegant rose and geometric relief patterns surrounding the character's face, arabesque patterns and flower carvings,
Three-dimensional spiral patterns utilizing the contrast between the green skin and the white layer,
Artisan skill, texture of the carving knife,

Placed on a silver tray on a tablecloth,
Various fresh fruits such as strawberries, grapes, oranges, and berries are lavishly arranged around the watermelon,
Depth of field, bright natural light, vivid colors, close-up, high-resolution photo, 8k

Negative Prompt:
(stem:1.5),iris, pupil color, makeup, painted, skin tone, colored hair, illustration, drawing, 2d, smooth skin, plain watermelon skin, simple background, text, watermark, low quality

Make this damaged finger look natural

Based on the uploaded "realistic interior space" reference image, generate a "transparent miniature house display model."

[Core Display Concept]
- Transform the original realistic interior space into a "miniature-scale transparent architectural model"
- Like an architectural cross-section model displayed in a museum
- The house is completely enclosed in a transparent acrylic / glass shell
- The interior structure and furniture are fully visible

[Structure and Layout (Must be Strictly Retained)]
- Strictly inherit the interior layout and proportions from the reference image
- Wall positions, furniture placement, and circulation paths must be identical
- The room is scaled down entirely to a miniature proportion (e.g., 1:10 / 1:12)
- This is proportional scaling, not redesign

[Transparent Casing Design]
- Exterior is high-transparency acrylic or glass material
- Rounded corners, clearly visible thickness
- Local use of cutaway / angled structures for easy viewing of the interior
- The transparent casing does not obstruct the main interior view

[Interior Representation (Miniature but Realistic)]
- All furniture is scaled down proportionally
- Materials remain realistic: wood grain, fabric, metal
- Retain a sense of real life, not a toy feel
- Lighting fixtures, plants, and ornaments are clearly visible

[Lighting System (Key Focus)]
- Warm-toned micro-lights inside (like model lights)
- Neutral or cool-toned display lights outside
- Creates a high-end display contrast of warm interior and cool exterior
- Gently glowing but not dazzling

[Display Environment]
- Placed on a minimalist pedestal / museum display stand
- Clean background (light gray / dark gray / black)
- Includes a real scale reference (like the stand's shadow)

[Visual Style]
- Hyper-realistic
- High-end industrial design aesthetic
- Looks like a display piece from Apple / Muji / an architectural firm
- Not toy-like, not cartoonish

[Photography and Composition]
- Professional product photography or architectural model photography
- Medium-close shot composition, slightly overhead view
- Shallow depth of field, blurred background
- Clear focus on the transparent miniature house itself

[Technical Requirements]
- Ultra-high definition details
- High refractive index, realistic glass representation
- Correct reflections and perspective
- No people, no text, no watermarks

Displayed as an art installation, possessing contemporary design and collectible value
Healing, exquisite, visually striking effect that immediately captures attention
Avoid toy texture, avoid plastic feel, emphasize architectural model and industrial design texture
High transparency, clear refractive edges, clean glass
Interior lighting clearly visible, like a miniature model lighting system

Page, Panel, Panel Size, Scene Description, Character, Dialogue, Manga Part Prompt, Live-Action Cooking Prompt, Notes 1, 1, Large, Yuta plates Japanese food in a Western style, Yu...

subject:
    description: "The chibi character in the attached image is performing the 'Tehepero' pose"
    character_style:
      - "chibi character"

  expression_and_pose:
    facial_expression:
      - "winking the right eye"
      - "tongue slightly sticking out upwards from the corner of the mouth ('Tehepero')"
    pose:
      - "lightly tapping the high side of the head with the right fist"
      - "left arm is stretched straight downwards"

  body_orientation:
    torso:
      - "The front of the body is not completely facing the camera, but slightly angled (3/4 view)"

  gaze:
    direction: "looking at the camera"
    quality:
      - "soft frontal gaze"
      - "gazing at the viewer"
      - "gentle eye contact"

  camera_and_composition:
    angle:
      - "high angle"
      - "looking down from above, but the face is clearly visible"
    perspective:
      - "exaggerated perspective"
      - "strong foreshortening (foreground appears large, background appears small)"
    lens:
      focal_length_equivalent: "24mm equivalent"
      distance: "camera is close to the subject"
    framing:
      - "full body included"
      - "centered"
    background:
      - "simple white background"

Write the given question on a green board with chalk and solve it.
There is a number y. Multiply that number by 3. Add the same number y to it. Then it becomes 6×x.
When x = {argument name="value of x" default="2"}, what is y?

Please create it in a loose and comforting drawing style, like a casual sketch drawn by a professional manga artist.

meta:
  title: "CALENDAR 2026 — {argument name="month name" default="FEBRUARY"} (Winter Alpine / Flat Teal Splitter / Sundays Red)"
  purpose: "image_generation_prompt"
  language_in_image: "English"
  output: "single calendar poster"
  quality: "commercial, crisp typography, no artifacts"
  strict_text_and_dates: true

canvas:
  size: { width: 2048, height: 1536 }
  aspect: "4:3 landscape"
  base_bg: "#FFFFFF"
  safe_margin_pct: 6

style:
  keywords:
    - "premium modern minimal"
    - "corporate clean"
    - "cool winter calm"
    - "high legibility"
  palette:
    deep_teal: "#083D57"
    teal: "#0E7890"
    mid_teal: "#6A90A0"
    navy: "#0B2433"
    text_dark: "#2C2F33"
    border_light: "#D7DDE2"
    icy: "#EAF2F6"
    saturday_blue: "#1F7FA0"
    sunday_red: "#C93B3B"
    holiday_red: "#C93B3B"
  typography:
    font: "Noto Sans (fallback: Helvetica/Arial)"
    rules:
      - "no warping, no jitter, no random glyphs"
      - "high anti-aliasing + optical kerning"

composition:
  layout:
    - "Left ~52%: winter alpine photo inside a large smooth curved mask"
    - "Right ~48%: clean white negative space with month + calendar card"
    - "Bottom-left: deep teal layered wave band hosting 'CALENDAR 2026'"
  photo_left:
    subject:
      - "snowy mountains, conifer forest, calm valley, overcast sky"
      - "no people, no buildings, no animals, no text"
    grading: "cool-neutral, slightly desaturated, moderate contrast, preserve snow detail"
    overlay: "very subtle snow/bokeh particles mainly over photo area (keep right area clean)"
  divider_splitter:
    shape: "thin smooth curved band near the boundary of photo/white"
    thickness: "5–7% of canvas height"
    fill_gradient: { from: "#083D57", to: "#0E7890" }
    optional_rim:
      enabled: true
      thickness: "1–2% of canvas height"
      fill_gradient: { from: "#EAF2F6", to: "#FFFFFF" }
    flat_rules:
      - "no 3D, no bevel, no inner shadow, no glass, no facets, no specular"
  bottom_waves:
    coverage: "bottom 22–28% of canvas"
    layers:
      - { gradient: { from: "#083D57", to: "#0B2433" }, opacity: 1.0 }
      - { gradient: { from: "#0E7890", to: "#6A90A0" }, opacity: 0.55 }
      - { gradient: { from: "#EAF2F6", to: "#FFFFFF" }, opacity: 0.45 }
    rule: "flat premium (no facets/bevel), minimal noise only to prevent banding"

text_blocks:
  month_number:
    text: "2"
    position: "top-right of right area"
    style: { weight: 800, color: "#0E7890", size: "very large" }
  month_label:
    text: "FEBRUARY"
    position: "upper-center of right area"
    style: { weight: 800, color: "#2C2F33", size: "large", tracking: "slightly wide" }
  footer_left:
    position: "bottom-left on teal wave band"
    lines:
      - { text: "CALENDAR", style: { weight: 700, color: "#FFFFFF", size: "medium" } }
      - { text: "2026",     style: { weight: 900, color: "#FFFFFF", size: "very large" } }
    alignment: "left"

calendar_card:
  placement: "center-right"

The chibi character in the attached image is performing the Tehepero pose.
She is making a fist with one hand and placing it on her temple.
Her tongue is sticking out upwards from the corner of her mouth.
She is winking with one eye.
Gaze: soft direct gaze, looking at the viewer, gentle eye contact

Camera angle: high angle (not overhead), looking down at the subject, face visible
Perspective: exaggerated perspective, strong foreshortening
Lens: wide-angle (24mm), close to camera
Composition: full body, centered, simple white background
Proportions: big head, small body (super-deformed)

Create a high-resolution low-angle, full-body shot of a beautiful woman (as attached image) with long brown hair and glossy pink lipstick sitting on outdoor concrete steps against a backdrop of a tall apartment building. She is wearing an oversized grey and brown plaid flannel shirt over a black top, dark trousers, and chunky white sneakers, accessorized with round sunglasses and a silver chain necklace. The lighting is soft and moody,

{ 
"prompt_type": "photorealistic gym mirror selfie", "main_composition": "full-body mirror selfie of a single young woman standing confidently in a modern gym, taking a photo with her iPhone held in right hand at chest level, left hand resting near waist adjusting shorts, subtle neutral expression looking at phone/camera, legs slightly apart in relaxed athletic stance, dim purple ambient lighting with wooden gym floor and equipment in background", "subject": { "description": "beautiful young woman in her early 20s, attractive feminine features with light natural makeup (defined eyeliner, glossy lips), long straight dark brown or black hair with full blunt bangs framing face and falling past shoulders, fair to tanned skin with toned athletic physique, visible defined abs and strong legs", "clothing": "fitted black long-sleeve crop top with high neck and thumb holes, high-waisted very short light pink seamless athletic shorts emphasizing curves, white crew socks pulled up, clean white athletic sneakers with laces", "details": "large white over-ear headphones worn on head, visible script tattoos on both upper thighs (mirror writing, artistic font), black manicured nails, confident casual pose with slight hip tilt and core engagement" }, "environment": { "foreground": "polished light wooden gym floor with reflections and subtle wear", "midground": "woman centered in large floor-to-ceiling mirror reflection", "background": "modern gym interior with dim purple/blue ambient lighting, racks of colorful medicine balls and kettlebells on left, weight plates and bars on right, wall clock visible, water fountain and trash bin on left wall, spacious open studio feel" }, "lighting_and_atmosphere": "dim indoor gym lighting with strong purple/blue ambient glow from overhead LEDs, cool tones contrasting warm skin highlights, subtle floor reflections, moody athletic vibe with high contrast on clothing and tattoos, photorealistic detail", "technical_quality": "highly detailed, sharp focus on subject with realistic fabric textures (seamless shorts shine, cotton crop top matte), skin details, tattoo script, and phone case, 8k resolution, professional mirror selfie style with visible mirror edge and slight wide-angle lens effect", "negative_prompt_suggestions": "blurry, deformed, extra limbs, phone screen visible, mirror distortion, overexposed highlights, underexposed shadows, low quality, cartoon, anime, text watermark, extra people, cluttered equipment in focus" 
}

アップロードした人物を完全再現
その人物が「Nano Banana Pro」を紹介する note記事の見出し画像
16：9の横長サイズ
スタイル、カラー：シンプル、手描き風、斜体、青と緑のグラデーション
タイトル：Google の新AI「Nano
Banana Pro」徹底解説

A high-end editorial photo of a {argument name="product" default="[PRODUCT]"} placed flat on a {argument name="surface type" default="[TEXTURED SURFACE]"}, captured from a direct top-down view. The surface is gently disturbed to suggest recent motion or interaction. The front of the product is fully visible and properly oriented upright. The area around the product is intentionally left open to optionally place {argument name="complementary objects" default="[COMPLEMENTARY OBJECTS]"} that visually enhance the scene. Natural sunlight from the upper left casts warm, realistic shadows. 3D realism, luxury product photography, shallow depth of field, 1:1 format.

Create an image at 31.7785° N, 35.2296° E, on April 3 of the year 33, at 3:00 PM

Anchor: [Uploaded Image]::3
Morphology: Intricate paper folding, sharp creases, layered topographic map style, polygonal mesh simulation::3
Material Physics: Textured cardboard, washi paper grain, fibrous edges, semi-translucent layering::2
Illumination: Hard paper-cut shadows, warm ambient occlusion, top-down architectural lighting::1.5
Render Stack: Tilt-shift photography, macro lens, stop-motion aesthetic, tangible texture::1
Negative: [Liquid, cloth, metal, realistic skin, rendering errors, blur]:: -0.5

Ultra-realistic candid smartphone photograph,
9:16 vertical.

Use the attached image only as a strong visual inspiration for general facial proportions and vibe,
while ensuring the generated subject remains fully original, non-identifiable, and not a face copy.

Scene
Nighttime on a quiet Tokyo city sidewalk beside a building with white ceramic tile walls.
The subject is walking past the camera, partially side-on, captured mid-step in a rushed, accidental moment.

Poses & Gestures (combined naturally)
The subject reacts instinctively to being photographed:
turning her head back mid-walk as if surprised,
raising one hand close to the lens to partially block the camera,
shy, restrained smile forming as she suppresses laughter,
body angled in half-profile while continuing to walk forward,
motion carries her slightly out of frame, imperfect framing.
Expressions are subtle and spontaneous — embarrassed, playful, fleeting — never posed.

Camera & Motion
Shot on a modern smartphone, handheld, rushed capture. Strong uncontrolled camera shake. Heavy directional motion blur across the raised hand, hair, and face. Facial features appear smeared and streaked by movement, faintly recognizable but distorted. Partial ghosting around the body. Imperfect framing, subject nearly leaving the frame.

Lighting
Flash-only lighting fired mid-motion. Harsh flash highlights on skin and hand. Deep surrounding darkness. Uneven exposure, blown highlights, hard shadows typical of hurried night phone photography.

Background
White ceramic tile wall stretches into streaked light bands due to motion blur. Street surroundings fade into darkness with minimal readable detail.

Mood
Candid, shy, slightly playful. Feels intrusive, raw, imperfect, and authentic — like a fleeting memory accidentally captured and never meant to be perfect.

Image Quality
Extremely noisy. Heavily blurred. Raw smartphone look. Not cinematic. Not editorial. Not polished.

Negative Prompt
anime, illustration, painting, stylized,
studio lighting, beauty lighting, soft portrait,
sharp focus, clean face, perfect anatomy, smooth skin,
fashion editorial, posed portrait, cinematic grading, film still,
tripod shot, professional photography,
identity match, face copy, real person replication.

– no anime, illustration, cartoon, or doll-like features
– no beauty filters, no airbrushed skin
– no face reshaping or eye enlargement
– no exaggerated body curves
– no extra or missing fingers
– no distorted arms, shoulders, or neck
– no studio or cinematic lighting
– no fashion posing

**CORE INSTRUCTION FOR THE MODEL:**
Analyze the personality traits and potential conflicts of the CHARACTERS previously provided as input. You must infer how these specific individuals would interact and place them into the four scenarios below based on their likelihood to be involved in that specific type of scene (e.g., put gossips in the confessional, aggressors in the fight scene).

**MANDATORY VISUAL AESTHETIC ("Real Broadcast Look"):**
The images must NOT look like staged photography. They must look like raw, unpolished TV broadcast footage.
* **Crucial Details:** Every panel must include on-screen graphics: timestamps (e.g., "DAY 12 | 14:32"), camera location identifiers (e.g., "CAM 1 - LIVING AREA"), and a small, semi-transparent channel logo in a corner.
* **Textures:** Expect video grain, slight lens distortion, harsh fluorescent lighting, and occasional digital compression artifacts.

**GRID BREAKDOWN (2x2 Layout):**

**[TOP ROW]**
* **[PANEL 1: Top-Left] Domestic Friction (Kitchen/Living Area).** Wide-angle fixed CCTV camera shot. Harsh, unflattering overhead lighting. Based on inputs, show a tense moment of daily life—two characters with incompatible habits arguing over cleaning or food, while others sit uncomfortably in the background. The composition is messy and naturalistic.
* **[PANEL 2: Top-Right] The Confessional Booth (El Confesionario).** Tight, claustrophobic fish-eye lens close-up. The background is a simple, soundproofed wall. Select the character most likely to vent or gossip from the inputs. They are looking directly into the lens, highly emotional (crying or angrily gesturing), breaking the fourth wall.

**[BOTTOM ROW]**
* **[PANEL 3: Bottom-Left] Night Vision Secrets (Bedroom).** Monochrome green or grey infrared night-vision shot. Grainy, low fidelity. Total darkness in the room. Show two characters (strategic allies based on inputs) whispering conspiratorially in a bed, barely visible under covers.
* **[PANEL 4: Bottom-Right] The Explosion (Garden/Smoking Area).** Chaotic, slightly shaky "handheld camera" feel, as if an operator ran to catch the action. High drama. A screaming argument at a party or nighttime gathering. The most volatile characters from the inputs are being physically separated by others. Flashing lights or dramatic shadows.

**STYLE:** Raw reality TV broadcast screenshots, 2x2 grid, highly detailed, 4k

{
  "generation_request": {
    "meta_data": {
      "tool": "NanoBanana Pro",
      "task_type": "photoreal_artistic_street_photography",
      "version": "v1.0_STREET_ART_REALISM_SER_EN",
      "priority": "highest"
    },
    "output_settings": {
      "aspect_ratio": "4:5",
      "orientation": "portrait",
      "resolution_target": "ultra_high_res",
      "render_style": "ultra_photoreal_candid_film_still",
      "sharpness": "crisp_but_natural",
      "film_grain": "subtle_35mm",
      "color_grade": "restrained_cinematic_realism",
      "dynamic_range": "natural_not_hdr",
      "skin_rendering": "real_texture_no_retouch"
    },
    "creative_prompt": {
      "scene_summary": "Candid artistic street photograph on location in {argument name="city or neighborhood" default="[CITY/NEIGHBORHOOD]"}. A real, unposed moment that feels documentary and emotionally true.",
      "subject": "A {argument name="age" default="[AGE]"} {argument name="gender" default="[GENDER]"} wearing {argument name="clothing" default="[CLOTHING]"}, caught mid-action: {argument name="micro action" default="[MICRO-ACTION]"}. Natural facial expression, slight asymmetry preserved.",
      "camera": "35mm lens equivalent, eye-level, close but respectful distance, imperfect framing like a real street photographer, focus on the eyes, shallow-to-medium depth of field.",
      "lighting": "Motivated natural light from {argument name="light source" default="[SOURCE: shop window / street lamp / sunset / neon]"}, deep but detailed shadows, no studio lighting.",
      "environment_details": "Street textures: worn posters, scratched metal, rain-speckled pavement, subtle reflections in glass, real urban clutter, no clean studio background.",
      "authenticity_markers": "Subtle motion blur on background only, slight halation on highlights, tiny film gate weave, realistic grain, no beauty retouch.",
      "mood": "{argument name="mood" default="[MOOD: melancholic / hopeful / tense / quiet joy]"}, cinematic but believable."
    },
    "negative_prompt": [
      "studio lighting",
      "beauty retouch",
      "plastic skin",
      "over-sharpened",
      "HDR",
      "AI glow",
      "perfect symmetry",
      "overly clean background",
      "fake bokeh shapes",
      "text",
      "logo",
      "watermark",
      "painterly",
      "cartoon"
    ]
  }
}