{
  "title": "Imaging material data for AI lot and reference-target review",
  "source_page": "https://fiund.com/industries/imaging-material-lot-reference-target-data",
  "authorship": "Original fictional Fiund editorial exercise. No actual operating records or third-party source assets are included.",
  "example": {
    "title": "A gray step belongs to the capture configuration",
    "context": "An owned grayscale target is captured twice under fictional device presets. There is no claim about material sensitivity.",
    "records": [
      [
        "1",
        "Capture protocol",
        "Use preset M1 and lot L2 for the comparison."
      ],
      [
        "2",
        "Frame A",
        "L2 image under M2; dark step appears lighter."
      ],
      [
        "3",
        "Frame B",
        "L2 image under M1; target and white reference both visible."
      ]
    ],
    "question": "Which frame qualifies and what can it show?",
    "answer": "B matches both specified keys in records 1 and 3 and includes a visible reference. A can be retained as M2 evidence, but the different preset in record 2 prevents attributing its lighter step to material variation in this comparison.",
    "failureModes": [
      "Comparing grayscale values across presets",
      "Using a lot match to ignore capture conditions"
    ],
    "fields": [
      [
        "material_lot",
        "Identifies the specimen batch"
      ],
      [
        "capture_preset",
        "Defines comparable imaging setup"
      ],
      [
        "reference_visibility",
        "Flags frames without the required target context"
      ]
    ],
    "holdout": "Provide two M1 frames with different white reference positions; require correct alignment before comparing steps.",
    "boundary": "No proprietary imaging chemistry, device settings beyond mock names or calibration claims."
  }
}