“Gold tone” is the vaguest claim in this industry. Every shop says they match. Almost none say to what, measured how, against what reference, or within what tolerance, which makes the claim unfalsifiable and therefore worthless.
Here is our method in full. The parameters we withhold are the ones that would let a competitor reproduce the deposit itself; the method, the instrument conditions, and the acceptance criterion are published because a standard nobody can check isn’t a standard.
The reference is a modern genuine Rolex 18K yellow gold component in original factory finish, not a photograph, not a swatch, not a published figure. Colour on gold shifts with alloy, surface finish, and age: pre-1980s Rolex yellow ran measurably warmer than current production, and a piece that has spent thirty years developing copper-oxide and sulphide surface films no longer reads as the alloy underneath. We therefore reference current-production material only, and re-verify against it rather than against a stored number.
The reference’s coordinates are not published. That figure is the one piece of information that would let another shop skip the work.
Measurement is by spectrophotometer, not by colorimeter or by eye. The distinction matters: a spectrophotometer records reflectance across the visible spectrum and computes coordinates from it, so two samples that match under one light source but not another are correctly identified as a mismatch (metamerism that a three-filter colorimeter cannot see).
Conditions, held constant for every reading:
| Illuminant | D65 (standardized daylight, 6500 K) |
|---|---|
| Standard observer | CIE 1964 10° |
| Colour space | CIE L*a*b* (CIELAB) |
| Difference formula | [CONFIRM WHICH THE INSTRUMENT REPORTED / VALIDATE] |
| Measurement geometry | [CONFIRM / VALIDATE] |
| Aperture | [CONFIRM / VALIDATE] |
Readings are taken in a controlled D65 booth rather than under room light. Ordinary domestic and workshop lighting sits between 2700 K and 4000 K, which flatters warm golds and hides exactly the deviation being tested for.
Gold is a specular surface on a curved, finished component, and a single reading off a polished case flank is not a colour measurement; it is a reflection. So:
The reported figure for a piece is the average of those readings, not the best one.
A piece passes at ΔE < 1.5 against the genuine reference.
Context for that number: ΔE ≈ 1.0–1.5 is broadly cited in colour-science literature as the just-noticeable difference range for trained observers viewing two samples side by side under controlled light. Actual perceptibility varies with surface finish, colour region, and viewing geometry, but under standardized conditions this range represents the approximate threshold of reliable human discrimination.
We publish the threshold, not per-piece values. A customer’s measured deviation is reported to that customer privately, with their piece. Publishing individual figures would turn a QC gate into a marketing number, and it would let a competitor reverse-engineer the target from a large enough sample.
Colour measurement governs colour. It does not govern how a finish wears, and it says nothing about adhesion or thickness. Those are separate gates with separate tests. A piece can match perfectly and still fail QC on coverage or finish quality, and it will not ship if it does.
Matching is offered for Rolex yellow tones, where we hold a verified reference. Rose and white are developed against the same gate and released only when they clear it. Where you have a specific reference or a custom target in mind, tell us at intake and we will work to it.