Key takeaway
A low color-difference value under one condition does not establish a match under another.
Define the comparison before choosing a tolerance
A textile color export can contain precise-looking differences and pass flags without the conditions that gave them meaning. The reviewer needs to know which standard was compared, how the sample was prepared and which lighting, observer and difference method the result represents. A blank context cannot be repaired by selecting the smallest number from several tests.
Konica Minolta explains metamerism: a pair can match under one light source and differ under another, and spectral reflectance allows evaluation under different illuminants. Datacolor’s technical discussion recommends controlled primary and secondary visual lighting and describes spectral comparison across illuminant and observer combinations. These manufacturer explanations were checked 11 October 2026; they establish no universal textile acceptance limit.
Complete a comparison register
The following hypothetical results use an invented T9 standard and local acceptance policy. Primary evaluation requires D65 illumination, a 10-degree standard observer and a Delta E 2000 limit of 1.0. Secondary evaluation uses illuminant A, the same observer and a limit of 2.0. The numbers are invented observations, not calculated from a supplied real fabric spectrum.
The example assumes a qualified instrument and consistent geometry, backing and specimen preparation. Sample A matches the stated primary limit at 0.8 but misses the secondary limit at 2.2. Sample B’s 0.7 uses a different observer, so it cannot establish acceptance under the primary policy as stated. Sample C’s 0.5 compares the wrong target version. Both need the specified comparison rather than a favorable pass flag.
| Result | Comparison conditions | Invented Delta E 2000 | Review decision |
|---|---|---|---|
| Sample A, primary | T9; D65; 10-degree observer | 0.8; local limit 1.0 | Matches the stated primary condition |
| Sample A, secondary | T9; illuminant A; 10-degree observer | 2.2; local limit 2.0 | Does not meet the secondary condition |
| Sample B | T9; D65; 2-degree observer | 0.7 | Repeat or recompute the required 10-degree comparison |
| Sample C | T8; D65; 10-degree observer | 0.5 | Resolve standard version; not approval against T9 |
Preserve measured spectra and the appearance context
If spectral reflectance is retained, identify its wavelength coverage, instrument, geometry and measurement settings. Keep derived color coordinates, the difference formula and the illuminant/observer selection separately. Recalculation under another condition needs compatible source measurements and an approved method; renaming a field cannot change the underlying observation.
Record the specimen side, orientation, layers, backing and relevant preparation. Distinguish an instrument result from a controlled visual assessment and retain which one the actual specification requires. Do not infer the chemical composition of a dye or diagnose the cause of a mismatch from the four invented numbers. They demonstrate conditional acceptance and missing comparison context, rather than a measured metameric mechanism in an actual textile.
Keep the decision history when a standard changes
A replacement physical standard, revised digital reference or customer approval can change the comparison target. Record the effective standard version and reviewer decision beside the lot or specimen record. A result against T8 should remain an observation against T8; silently relabeling it T9 destroys the evidence for the earlier decision.
Separate accepted primary results, accepted secondary results, unreviewed context changes and actual rejection reasons. A missing secondary result is unknown, not a failed test or an assumed match. Conversely, Sample A’s supplied secondary result prevents this example from receiving the combined acceptance required by its invented policy. The archive may still be useful for conditional color analysis if those limitations travel with it.
Release an interpretable color package
The example’s completed deliverable is a conditional comparison register with one primary match, one secondary mismatch and two unresolved comparisons. The operating owner should add the applicable customer specification, reference-control history and measurement records before making a real acceptance decision. A generic calibration certificate alone cannot decide whether two color results used the same illuminant, observer or target.
Inventory can map standards, spectra and lot-level decisions; Due diligence can ask which appearance conditions a recipient intends to study. Start with a synthetic register when explaining the structure. Customer designs, formulas and specimen identifiers may be confidential and need a separate rights review. Permission for a named recipient to receive a description does not itself authorize the release of actual spectra or customer approval history.