How Close Can Two Colors Be Before They Become “The Same”?

Two paint cards can look identical in the store and separate clearly on a wall. A fabric panel can pass visual inspection indoors and fail beside a window. Color science calls the smallest reliably detected change a discrimination or perceptibility threshold, but that threshold belongs to specified conditions. It is not a magic HEX-code distance. The problem combines the lighting sensitivity seen in visible wall-paint repairs with the limits of the observer and measurement system.

First decide what “same” needs to mean

Two spectral reflectance curves can differ while producing the same apparent color under one light; that is metamerism. Two measured coordinates can differ while an observer cannot reliably tell the samples apart. Two visibly different panels may still be acceptable on separate products but unacceptable when they touch.

A perceptibility threshold asks when a difference becomes detectable. An acceptability threshold asks when the difference becomes objectionable for a task. The second is usually wider and depends heavily on expectations.

What ΔE measures—and what it does not

ΔE expresses distance between colors in a defined color space. The original CIELAB ΔE uses straight-line distance between L*, a* and b* coordinates. Later formulas, including CIEDE2000, add corrections because equal geometric distances in CIELAB do not look equally different everywhere.

A smaller ΔE generally means a closer match, but quoting “ΔE 1” without naming the formula, instrument conditions and material is incomplete. Published thresholds vary across studies and applications. Dental ceramics, textiles, displays and automotive finishes do not share one observer task or surface geometry.

Why one pair is easier to separate than another

Sensitivity varies across color space. A small lightness difference may stand out on a smooth neutral surface while an equal numerical hue change hides in a textured, saturated material. The size of the samples, their separation and the surrounding field also affect detection.

Side-by-side samples reveal small boundaries that memory loses. Place a gap between them and the task becomes harder; show them sequentially and color memory adds its own bias. This is why industrial visual assessments specify how samples are arranged.

Light can make a passing match fail

Two materials that match under one illuminant can separate under another because their spectra differ. Gloss and texture change the balance between diffuse color and highlights. A spectrophotometer setting that excludes or includes specular reflection can therefore answer a different question from a person viewing the finished object.

For practical matching, check the intended material under the intended lights and angles. Numerical tolerances are most valuable when paired with a repeatable viewing protocol.

A simple home comparison

1. Place samples directly beside each other, then swap left and right positions.
2. View them under daylight and the actual indoor lamp.
3. Compare from the normal use distance, not only inches away.
4. Ask whether you can detect a difference and whether that difference matters; record the answers separately.

This will not produce a laboratory threshold. It will produce a decision tied to the real object and context, which is often more useful than an unsupported universal number.

The honest answer is a test condition

Colors become “the same” when a defined observer, under defined conditions, cannot distinguish them reliably or when the remaining difference is acceptable for the task. Color-difference formulas help organize that judgment. They do not replace the need to state what was compared and why.