Ai Image Prompts

I tried creating a 4-panel manga about recent daily life. Banbi & SMA Mom Remote Control Episode

prompt to create the storyboard grid below

STRICT ADHERENCE REQUIRED:
  - Follow this YAML specification precisely. Do NOT deviate, add unauthorized elements, or make 'creative' interpretations.
  - Character identity consistency from reference images (if provided) is PARAMOUNT.
  - Thorough image analysis is mandatory to maintain this consistency. Any result that fails character identity checks is unacceptable.

Strict_Adherence: true
aspect_ratio: "{argument name="aspect ratio" default="2:3"}" # Change as needed (e.g., "1:1", "16:9")
resolution: "{argument name="resolution" default="2K"}"    # Change as needed (e.g., "4K")
Character_Count: {{ Total number of characters appearing }}
Total_Panel_Count: {{ Total number of panels (usually 4-8 for a 4-tier structure) }}

Summary: >
  {argument name="summary" default="Overall summary of the story. Briefly describe the plot flow, setting, main events, and conclusion."}

Style_Guide:
  # Panel division style is fixed to "no outer frame, display to canvas edge"
  comic_style: "4-tier, multi-panel manga strip with no outer framing border. The panels should extend to the very edge of the canvas, separated only by the inner white dividing lines."
  lighting: "{{ Specify lighting. Time of day (day/night), light source (sunlight/electric light/moonlight), atmosphere (bright/dark/dramatic shadows), etc. }}"
  background_elements: "{{ Main setting, specific description of the location, important props or environmental elements placed. }}"
  dialogue_bubbles: "Standard white bubbles for speech, thought clouds for internal thoughts, dashed bubbles for whispers, and rectangular boxes for narration/monologue. {{ Add definitions for special speech bubbles if necessary (e.g., spiky bubbles for shouting) }}"
  visual_effects: "{{ Specify manga visual effects. Manga symbols (sweat, veins, sparkle), effect lines (concentration lines, speed lines, tremor lines), visualization of sound effects, etc. }}"

Characters:
  # Character definition. Alphanumeric IDs recommended.
  {{ Character ID_1 }}:
    description: "{{ Appearance, hairstyle, color, clothing, distinctive items, personality and expression tendencies. If reference images exist, write the file name here (e.g., reference image_0.png). }}"
  {{ Character ID_2 }}:
    description: "{{ Second character details... Describe for the number of characters. }}"
  # Add as needed...

Storyboard:
  # 4-tier panel division definition
  Tier_1:
    {{ Panel ID_T1_P1 (e.g., Panel_1_Large_Top) }}:
      scene: "{{ Specific situation description for this panel. Who, where, and what they are doing. Composition and camera work can also be specified. }}"
      dialogue:
        - speaker: "{{ Speaker's Character ID }}"
          type: "{{ Type of line: speech(normal), thought(thought), whisper(whisper), shout(shout), monologue(monologue/narration) }}"
          text: "{{ Specific content of the line. Use \n if a line break is needed. }}"
        # Add list if there are multiple lines in the same panel
        # - speaker: "..."
        #   type: "..."
        #   text: "..."
    # Write here to add panels to Tier_1
  Tier_2:
    {{ Panel ID_T2_P1 }}:
      scene: "{{ Scene details }}"
      dialogue:
        - speaker: "{{ Speaker ID }}"
          type: "{{ Type }}"
          text: "{{ Content }}"
    # Add panels as needed
  Tier_3:
    {{ Panel ID_T3_P1 }}:
      scene: "{{ Scene details }}"
      dialogue:
        - speaker: "{{ Speaker ID }}"
          type: "{{ Type }}"
          text: "{{ Content }}"
    # Add panels as needed
  Tier_4:
    {{ Panel ID_T4_P1 }}:
      scene: "{{ Scene details }}"
      dialogue:
        - speaker: "{{ Speaker ID }}"
          type: "{{ Type }}"
          text: "{{ Content }}"
    # Add panels as needed

Transform the original photo into a dramatic, photorealistic, ultra-detailed set of 4 different styles characters are included , each a mid close up wide-angle shot with an extreme, dynamic camera angle with  powerful pose in a consistent, expanded version.

Real men and women are sucked into a comic book, transform into comic characters, and strike the final pose of the image against a futuristic city background.

Prompt for creating a manga of over 10 pages using nanobanana pro.

{

  "type": "image_generation_prompt",

  "language": "English",

  "style": "ultra-realistic cinematic film still",

  "aspect_ratio": "16:9",

  "subject": {

    "gender": "female",

    "pose": {

      "position": "standing alone",

      "orientation": "centered, facing the camera",

      "posture": "still, composed, slightly tense"

    },

    "appearance": {

      "hood": "wearing a hooded jacket with the hood up",

      "face": {

        "skin_texture": "photorealistic, natural human skin",

        "cyborg_transition": {

          "stage": "early-stage",

          "details": [

            "faint bioluminescent circuitry barely visible beneath the skin",

            "subtle glow along one cheek and the side of the neck",

            "restrained and realistic, not exaggerated"

          ],

          "eyes": {

            "left_or_right_eye": "one eye reflects a cold, mechanical glint while remaining mostly human"

          }

        }

      },

      "expression": "neutral but intense, anticipatory, calm with underlying tension"

    }

  },

  "environment": {

    "location": "inside a moving subway train",

    "interior": {

      "walls": "brushed metal",

      "poles": "vertical steel poles",

      "floor": "worn industrial flooring",

      "lighting": "soft, flickering fluorescent lights"

    },

    "windows": {

      "outside_view": "motion blur indicating speed and movement"

    }

  },

  "lighting": {

    "type": "natural cinematic lighting",

    "contrast": "soft contrast",

    "mood": "cool, tense, restrained"

  },

  "color_palette": {

    "primary_tones": ["desaturated blues", "cool grays"],

    "saturation": "low, muted"

  },

  "camera": {

    "depth_of_field": "shallow",

    "focus": "sharp on the woman’s face and upper body",

    "style": "high-end cinematic film still"

  },

  "atmosphere": {

    "mood": "tense, anticipatory, pre-transformation",

    "genre": "sci-fi realism",

    "aesthetic": "grounded, subtle, realistic"

  },

  "quality": {

    "realism": "photorealistic",

    "detail_level": "high",

    "skin_detail": "visible pores and natural texture"

  },

  "constraints": [

    "No exaggerated sci-fi elements",

    "No fantasy glow or neon effects",

    "No cartoon or illustration style",

    "No overexposed lighting",

    "Cyborg elements must remain subtle and realistic"

  ],

  "output_goal": "Create a photorealistic cinematic image of a woman standing alone inside a moving subway train, capturing a subtle early-stage cyborg transformation with a tense, anticipatory mood and grounded sci-fi realism."

}

A close-up shot of a man with scissors, claymation, stop motion animation, miniature diorama, realistic clay texture, visible fingerprints, warm lighting, high detail, 8k

The subjects from the uploaded image rendered in a stark, high-contrast graphic novel style, heavily inspired by Frank Miller's Sin City aesthetic. Deep, unforgiving shadows define the muscular structure and harsh expressions, utilizing heavy cross-hatching and thick ink outlines instead of traditional shading. Dramatic chiaroscuro lighting emphasizes the tension between the two figures, with only stark white highlights cutting through the predominantly black composition, against a pure void background, appearing like a single, powerful comic book panel.

{
  "concept": "a quiet moment that feels remembered rather than photographed",
  "subject": {
    "type": "young man",
    "age": "early 20s",
    "presence": "subtle, almost unnoticed within the frame",
    "expression": "emotion held back, eyes suggesting unfinished thoughts",
    "skin_tone": "natural muted tone with real-life unevenness",
    "features": "unposed face, relaxed posture, human imperfections preserved"
  },
  "scene": {
    "environment": "abandoned outdoor bus stop near a dried riverbed",
    "location": "cracked concrete shelter, faded signage, tall wild grass reclaiming the area",
    "props": "old bench, weathered backpack, torn paper timetable",
    "time_of_day": "late evening just before blue hour fades",
    "weather": "still air, faint dust particles floating"
  },
  "composition": {
    "style": "5-image narrative grid, non-linear storytelling",
    "shots": [
      "extreme close-up of fingers resting on the bench, shallow focus",
      "reflection of the subject in a scratched metal surface, face partially distorted",
      "medium shot from behind, subject seated and slightly slouched",
      "side profile framed through tall grass in the foreground",
      "wide static shot where the subject appears small, almost dissolving into space"
    ],
    "camera_angles": "observational, imperfect, slightly off-center",
    "framing": "intentional imbalance, negative space dominating the frame"
  },
  "lighting": {
    "type": "natural residual daylight",
    "quality": "soft, fading, low-energy light",
    "color_temperature": "cool desaturated blues with hints of warm decay",
    "shadows": "subtle, undefined, realistic"
  },
  "visual_style": {
    "mood": "nostalgic, suspended, emotionally restrained",
    "color_grading": "washed-out tones with gentle color separation",
    "contrast": "very low, almost flat",
    "film_look": "experimental indie film / memory archive aesthetic",
    "realism": "raw, anti-gloss, intentionally imperfect"
  },
  "camera_settings": {
    "lens": "40mm–58mm documentary-style lens",
    "depth_of_field": "selective focus, occasionally soft",
    "focus": "priority on atmosphere over sharpness"
  },
  "quality": {
    "resolution": "4K",
    "detail": "natural textures, visible wear, realistic surfaces",
    "noise": "organic grain resembling scanned film",
    "artifacts": "none, no AI distortions"
  }
}

I use GPT-image for drawing combat scenes ( ´∀｀)
Left: GPT-image
Right: nano banana pro
Same prompt, same reference image

[Forbidden Items]
- Left-to-right reading, bottom-to-top reading composition is forbidden
- English dialogue is forbidden
- Horizontal writing of dialogue is forbidden
- Monochrome, black and white, line art is forbidden