A cabinet gloss number needs an angle before it can settle a finish comparison

A gloss reading by itself cannot establish that two cabinet finishes match. The report needs the measurement angle and an identified reference sample, and acceptance needs an agreed limit. First determine whether the numbers describe the same measurement. Otherwise a dispute about the coating may really be a comparison of different tests.

Why the angle belongs to the result

BYK-Gardner describes gloss measurement at 20°, 60° and 85°, with different geometries suited to different gloss ranges. It also distinguishes repeatability from agreement between instruments. These are measurement considerations, not a universal cabinet acceptance limit.

Reading note: BYK-Gardner: Guide for Reliable Gloss Data.

The angle defines the geometry used to illuminate and receive reflected light. Changing it changes the observation. A value taken at one geometry should not be subtracted from a value taken at another and described as a production difference. Keep the angle next to the gloss-unit result in the report; burying it in a separate email makes an apparently precise table easy to misread.

The same discipline applies to the sample. A retained approval panel and a newly supplied door may have different measurement areas, textures or surface conditions. Establish a suitable common method before treating their readings as comparable. The measurement location should represent the surface being accepted. A number from a convenient flat patch does not automatically describe a profiled recess or the appearance of the whole door.

Use the agreed angle when comparing readings

In a hypothetical project, the approved flat sample reads 38 GU at 60°. The project has expressly agreed on a tolerance of ±2 GU at that angle, producing an acceptance band of 36-40 GU. A replacement panel report says 48 GU at 85°. The buyer does not reject it as 10 GU above target, because the geometries differ. The panel is remeasured at the agreed 60° using the agreed method and reads 39 GU. Its difference from the reference is 1 GU, within the hypothetical band, so the numeric gloss item is accepted. Color, texture, defects and the visual match remain separate acceptance items.

That result would change if the remeasured panel fell outside the agreed band, but a limit cannot be invented after seeing the result. Nor should the sample target drift because a production batch happens to have a different average. Store the approved sample identity and the acceptance rule together so the receiving team and the fabricator use the same reference.

Use repeated readings to understand a disagreement

If repeated measurements on the same suitable location fluctuate, first examine the measurement process. A stable difference between two instruments raises a different question from a variable reading on one instrument. Those patterns do not diagnose the cause by themselves, but they help decide whether to review the instrument setup, the sample location or the coating before requesting refinishing.

A batch average also needs context. It can hide a conspicuous panel if most other panels cluster around the reference. Record the part identities and the agreed sampling method, rather than allowing an average to silently replace a per-part requirement. Whether the contract accepts a mean, individual results or another rule should be explicit before inspection begins.

For a homeowner, a physical sample viewed in the room may be the practical selection tool; owning a glossmeter is not a prerequisite for choosing cabinets. Instrument data becomes useful when the supplier and purchaser have defined what it will resolve. Even then, a gloss result does not replace a color or texture decision. Two finishes can satisfy a gloss criterion while differing visibly in another way.

Keep the final record short and usable: the actual part, the reference sample, geometry, readings, acceptance rule and disposition. This prevents a supplier report containing a single bare number from being treated as proof of a complete finish match. The numeric criterion settles the numeric question that was agreed, while the remaining appearance decisions stay visible.