Color Maker: Film Curves → Charts (fixed exposure)

1. Prompt Example (JSON blueprint)

Use the following prompt to ask an AI to extract curve data as JSON. Include a screenshot of the film manufacturer’s datasheet when you submit the prompt.

You are an image processing engineer and data scientist.
Analyze the attached film datasheet image and extract"intermediate data" for downstream numerical simulation (scanner model).

## Goal and scope
- Goal: Collect raw data for density → transmittance → RGB conversion on the user side.
- Scope: Read values from charts in the image and output JSON only. Do not attempt color rendering or spectral synthesis.

## Important (optional fields)
- If the process type (C-41 / ECN-2 / RA-4 / E-6 / K-14, etc.) is not clearly readable, fill `meta.film_process`
 so the user can override it manually.
- Even if OCR detects a process in `characteristic_curves.ocr.process`, store a normalized value in `meta.film_process`.
- Allowed values for `meta.film_process` (examples):"C-41","ECN-2","RA-4","E-6","K-14","UNKNOWN"

## Task steps
1. **OCR:** Read all text in the image accurately (exposure, process, axis labels, etc.).
2. **Axis definition:** Determine min/max for X and Y axes from tick marks and labels.
3. **Curve tracing:** Visually trace each curve and convert it into numeric points.
  - **Important:** This is an AI visual estimate. Do not extrapolate unclear regions.
  - **Important:** If curves overlap or are ambiguous, lower the `confidence` score.

## Output requirements (JSON)
- Output raw JSON only (no code fences).
- Use numeric values (floats).
- Sampling interval: Prefer uniform spacing on X (e.g., Log E every 0.25, wavelength every 10nm).
- Keep each RGB curve separate to improve readability and accuracy.

## JSON schema
{"meta": {"source":"Film model name detected via OCR","iso": 200,"film_process":"C-41","notes":"Comments about graph distortion, readability, and confidence","digitize_info": {"method":"AI_visual_estimation","sampling_strategy":"fixed_interval_if_possible"
  }
 },"characteristic_curves": {"ocr": {"exposure":"Example: Daylight 1/100 sec","process":"Example: K-14","densitometry":"Example: Status A","text_dump":"Other text near the chart"
  },"axes": {"x": {"label":"LOG EXPOSURE","min": -4.0,"max": 1.0 },"y": {"label":"DENSITY","min": 0.0,"max": 4.0 }
  },"data": {"blue_layer": {"confidence": 0.9,"points": [ [-3.0, 0.1], [-2.75, 0.2] ]
   },"green_layer": {"confidence": 0.9,"points": [ [-3.0, 0.1], [-2.75, 0.2] ]
   },"red_layer": {"confidence": 0.9,"points": [ [-3.0, 0.1], [-2.75, 0.2] ]
   }
  }
 },"spectral_sensitivity_curves": {"ocr": {"conditions":"Example: Effective Exposure 1.4 sec","notes":"..."
  },"axes": {"x": {"label":"WAVELENGTH (nm)","min": 250,"max": 750 },"y": {"label":"LOG SENSITIVITY","min": -1.0,"max": 2.0 }
  },"data": {"yellow_forming": {"confidence": 0.8,"points": [ [380, 0.1], [390, 0.2] ] },"magenta_forming": {"confidence": 0.8,"points": [ [380, 0.1], [390, 0.2] ] },"cyan_forming": {"confidence": 0.8,"points": [ [380, 0.1], [390, 0.2] ] }
  }
 },"spectral_dye_density_curves": {"exists": true,"ocr": {"notes":"..." },"axes": {"x": {"label":"WAVELENGTH (nm)","min": 400,"max": 700 },"y": {"label":"DENSITY","min": 0.0,"max": 2.5 } },"data": {"midscale_neutral": {"points": [ [400, 1.0], [410, 1.1] ] },"minimum_density": {"points": [ [400, 0.4], [410, 0.45] ] }
  }
 }
}

Use this prompt to generate JSON, then load it in the next step.

2. Load JSON

Paste JSON or load a file.

ISO Film Type Scan Look Preview EV 0.0 Highlight protection auto exposure Soft shoulder enabled Black offset (D6) enabled

3. Preview

Classic Color Chart 24 + Gray Scale 20 Steps (single output)

LUT Input Export clean no auto exposure / shoulder