Analyze this image in detail and output a JSON object describing what you see. Include any relevant details about:

Here is an example output:

{
  "scene": "overall 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 }.

