Colour, Grayscale and Calibration Targets

Specify the image pipeline

A colour requirement needs more than a gamut name. Record source colour encoding, transfer function, signal range, bit depth, processor settings, display mode, white point and working luminance. Distinguish the content's intended colour space from the display's available gamut and from the calibration target.

ITU-R BT.709 defines HDTV production parameters; BT.1886 describes a reference electro-optical transfer function for studio displays. Their roles differ. A delivery instruction that names BT.709 should also identify the intended display response and viewing conditions. A room used for signage or presentations needs a documented operating target appropriate to that application.

Match signal range at both ends

For BT.709 nominal RGB and luma signals, reference black and white are codes 16 and 235 at 8-bit, or 64 and 940 at 10-bit. Chroma has different nominal endpoints: 16–240 or 64–960. See BT.709-6, table 4.6 (printed p. 5). Match the source output range and receiver interpretation; a full-range computer output requires the corresponding input setting. Use near-black and near-white patches to check clipping and lifted blacks.

The source encoding transfer function and display light-output response describe different stages. BT.1886, Annex 1 defines the display response using measured white and black luminance with a 2.4 exponent. Its black-level compensation means that selecting a generic “gamma 2.4” menu item does not establish a match.

Separate gamut, accuracy and uniformity

Gamut describes a region of reproducible colours. Accuracy describes how closely measured output matches target values under a stated error metric. Uniformity describes how output varies across the image or with viewing position. A large gamut percentage cannot substitute for an accuracy or uniformity test.

Preserve detail at the intended brightness

Dark ramps reveal problems that a vivid demonstration may hide. Check for collapsed steps, visible bands, colour casts and module boundaries in the actual operating mode. Compare a known external test file with internal processor patterns while recording which parts of the pipeline each bypasses.

For an N-bit component there are 2ⁿ possible code values: 256 for 8-bit and 1,024 for 10-bit. That arithmetic describes code capacity. Effective precision depends on source encoding, conversion, dithering, processing and display drive. A high-bit-depth processing label alone does not establish the number of distinguishable low-light levels on the screen.

For small coloured text, test chroma sampling through the full signal path. Reduced chroma resolution can soften fine colour detail while larger luminance features still look sharp. Use native-scale patterns and retain the actual output format in the acceptance record.

Make calibration recoverable

Save the factory or commissioning baseline before changing settings. Store calibration files together with receiving-card and module identifiers. After a replacement, confirm both optical matching and compatibility with the existing correction data. Recheck the working luminance, neutral ramps and boundary areas before returning the system to service.

Define numerical tolerances in the project plan, with the measurement method and uncertainty. This library provides the record structure; limits must correspond to the actual application and equipment.