Analyze the input images. I want you to merge them into one scene using the following logic:
Take the Character from Image 1: Describe this person in detail (use only "man" or "woman").
Take the Art Style/Background from Image 2: Apply this aesthetic and setting to the character.
Final Output Requirements:
Create a cohesive visual description of this merged scene. Include specific details on:
Lighting: (e.g., rim lighting, softbox, golden hour)
Materials: (e.g., fabric textures, skin pores, material of background objects)
Camera: (e.g., macro shot, wide angle, depth of field)
The result should read like a high-end photography brief.


Here is an example output:

{
  "scene": "overall merged scene description",
  "subjects": [
    {
      "description": "detailed subject description",
      "position": "where in frame",
      "action": "what they're doing"
    }
  ],
  "style": "artistic style",
  "color_palette": ["#hex1", "#hex2", "#hex3"],
  "lighting": "lighting description",
  "mood": "emotional tone",
  "background": "background details",
  "composition": "framing and layout",
  "camera": {
    "angle": "camera angle",
    "lens": "lens type",
    "depth_of_field": "focus behavior"
  }
}

CRITICAL: Output ONLY the JSON object. No explanations, no markdown code blocks, no extra text. Start with { and end with }.