People with color vision deficiency still discriminate lightness differences

What it is

Even when hue discrimination collapses — red-green or blue-yellow colors becoming confused — people with color vision deficiency retain essentially normal ability to discriminate lightness (brightness) differences. That is because the deficiency affects a specific cone type, or the circuitry that compares different cone types against each other, rather than the pathway that computes overall brightness — which stays largely intact. For this population, once hue fails, light-dark contrast remains a reliable channel.

Why it happens

Color vision deficiency typically arises from a missing or spectrally shifted photopigment in one cone type (protan/deutan/tritan). The pathway that computes brightness, however, works by summing signals across whatever cone types are present (plus rods under low light), largely independent of any one cone type's exact spectral tuning. As long as at least two cone types with overlapping-but-different sensitivities remain (the anomalous trichromacy case), or even with only one cone type fully functional and another entirely absent (the dichromacy case), this "sum the light signals to get brightness" computation keeps working. What actually breaks is the separate computation of "compare different cone types' responses to extract hue" — not the "sum responses to extract brightness" computation.

Studying it

This conclusion comes from luminance contrast sensitivity testing administered to clinically diagnosed dichromats and anomalous trichromats, compared against normal trichromat controls under matched conditions. Studies consistently find the luminance contrast sensitivity function is largely comparable between groups (small differences are reported under some specific conditions or eccentricities, but the magnitude is far smaller than the gap in chromatic discrimination), while the same participants' chromatic discrimination is markedly reduced. This dissociation — impaired chroma, spared luminance — is itself the key evidence behind the design recommendation to rely on lightness rather than hue for color-vision-safe design.

Where it stops holding

• "Largely normal" is not "identical." Some experimental conditions still report subtle differences; the magnitude is far smaller than the chromatic-channel loss and can be treated as a reliable design fallback, but it is not a zero-risk guarantee.
• This describes congenital color vision deficiency. Most cases involve a specific cone type's function; it does not necessarily extend to certain acquired eye conditions that also damage luminance/contrast sensitivity — in those cases the luminance channel itself may be impaired, so it cannot be assumed to always be a safe fallback.
• The lightness gap has to be large enough to be reliable. Very subtle lightness differences are hard for anyone to discriminate, regardless of color vision status; color vision deficiency does not make lightness discrimination better than normal, it just does not make it worse.

Applying it

• When information must reliably tell color-deficient users "these are different" or "these are opposite," make lightness difference the primary distinguishing channel, with hue difference as a supporting cue rather than the only one.
• When designing a semantic color system, avoid placing colors that carry different meanings at the same lightness level — no matter how far apart their hues are, matched lightness reads as no distinction at all to a color-deficient user.
• How to check: convert the color scheme to grayscale (removing hue and saturation, keeping only lightness) and check whether elements that were meant to be distinguished by color remain distinguishable. If they do, the lightness channel is doing its job; if they collapse in grayscale, the scheme relies on hue alone and will fail for color-deficient users.

Related

• Same group: A1.06.1 Red-green deficiency is the most common type, far more prevalent in men · A1.06.2 Red-green semantic opposition fails for this population · A1.06.4 Color-blindness simulators only validate part of the picture
• Nearby: A1.04.1 Discriminability is set by luminance contrast, not hue difference · U4.04 Color-vision-deficiency-safe palettes
• Search terms: luminance discrimination · color vision deficiency · lightness channel