A Box of Colored Caps Still Outperforms Your Monitor

A well-calibrated monitor tells you nothing about whether you can actually see color well. The industry-standard test for that skill is a box of eighty-five colored caps, built on a system a painter invented decades before anyone needed a profile at all, and it still has nothing to do with your gear.

A well-calibrated monitor tells you nothing about whether you can actually see color well. That claim sounds strange until you consider that the tool built to test human color perception is not a monitor at all. It is a box of eighty-five colored plastic caps, arranged by hand under controlled light, unchanged in principle since a U.S. Navy color scientist built it in 1943 to screen pilots. The Farnsworth-Munsell 100 Hue Test has nothing to do with sensors, gamuts, or profiles, and it is still the industry standard used by governments and manufacturers today. It measures the one variable no calibration process has ever touched: the photographer's own eye. And it descends directly from a color system a painter invented decades before anyone needed to calibrate anything at all.

The Test Nobody's Gear Can Pass For Them

You can still order the Farnsworth-Munsell 100 Hue Test today, from Pantone, largely unchanged from its original design. Four trays, eighty-five removable colored caps spanning the visible spectrum, arranged under standardized daylight conditions. The task is simple to describe and genuinely difficult to do well: place the caps in each tray in a smooth, unbroken progression of hue, with two fixed reference caps anchoring either end. There is no ambiguity to hide behind. Either your eye can detect the small perceptual step from one cap to its neighbor or it cannot, and the errors get scored on a chart that shows exactly where your color discrimination weakens.

Dean Farnsworth, a color scientist working for the U.S. Navy, built the test in 1943 to solve a specific operational problem: screening pilots and other personnel for color vision deficiencies that a simple color-blindness chart might miss. The original design used one hundred disks; Farnsworth trimmed it to eighty-five after early testing showed that number gave more consistent, uniform results without losing diagnostic value. The test has been an industry and government standard for more than sixty years since, still used today by manufacturers, printers, and anyone whose work depends on someone's ability to actually perceive fine color differences accurately, not just own equipment capable of displaying them.

Munsell Built The Foundation Decades Earlier

Farnsworth did not invent the underlying color logic his test relies on. He built it directly on top of a system Albert Munsell had developed nearly four decades earlier, for reasons that had nothing to do with pilots, industry, or measurement standards of any kind. Munsell was a painter and longtime instructor at the Massachusetts Normal Art School, and what drove him was a teaching problem: color names were useless for precise communication. Two people looking at the same shade might call it sapphire, navy, or deep blue, and none of those words told you anything exact.

Starting around 1898, Munsell began working out a rational alternative, arranging color samples according to three independent, measurable properties: hue, the color's name in the traditional sense; value, how light or dark it is; and chroma, its purity or intensity. He published the system in 1905 as A Color Notation, and refined it further in a 1915 atlas that gave it visual form as a three-dimensional structure, later nicknamed the Munsell color tree. None of this depended on physics, optics, or any instrument beyond human observers comparing painted chips and agreeing on when the perceptual steps between them felt equal. Munsell died in 1918, but the system outlived him by design. It was built to describe how people actually see color, not how any particular tool renders it.

The CIE Myth, Corrected

A common shorthand claims that modern digital color spaces descend directly from Munsell's system, that CIE 1931, the foundational coordinate system behind sRGB and Adobe RGB, took Munsell's perceptually-calibrated tree and simply converted it into mathematics. That version of the story gets the relationship backwards, and it is worth correcting plainly rather than repeating.

CIE 1931 was built independently, by an entirely different research tradition. In the late 1920s, W. David Wright and John Guild ran separate color-matching experiments, asking human observers to mix three fixed primary lights until the mixture visually matched a test color. Their results, combined and standardized by the International Commission on Illumination in 1931, became the CIE color-matching functions, a system built from physics and psychophysics, with no reference to Munsell's work at any point in its creation. Munsell's chip-based system and the CIE's light-matching system were solving the same underlying problem, how to specify color precisely, from two completely different directions, at almost the same time, without either one growing out of the other.

The two systems only met later, and even then through deliberate reconciliation rather than descent. In 1943, the Optical Society of America undertook what became known as the Munsell renotation, translating Munsell's existing color chips into CIE coordinates and adjusting their spacing based on new visual experiments, producing a corrected version of Munsell's system expressed in CIE terms. That is a committee spending years lining up two independently built maps of the same territory. It is not one system giving birth to the other.

Farnsworth Builds On Munsell, Not On CIE

Here is what makes the timing worth noticing. The same year the Optical Society of America was busy reconciling Munsell with CIE mathematics, Dean Farnsworth was building an entirely different tool on Munsell's foundation, one aimed at the opposite problem. The renotation project existed to make Munsell's system speak the CIE's coordinate language, useful for manufacturing and instrumentation. Farnsworth's test existed to measure something no coordinate system could ever substitute for: an individual human being's actual capacity to perceive hue differences, using colored caps with constant value and chroma so that hue was the only variable in play.

That is the real inheritance worth naming. Munsell's most lasting contribution was never a coordinate system waiting to be absorbed into digital color management. It was a rigorous, human-perception-first way of organizing color that turned out to be useful for two entirely different purposes, mathematical translation on one hand, and a hands-on test of a person's own eye on the other. Only one of those two purposes required a computer. It was not Farnsworth's.

Take The Real Test, Not The Free One

If you want to know how good your own color discrimination actually is, order the Farnsworth-Munsell 100 Hue Test and take it under controlled daylight conditions, the way it was designed to be taken. It is worth doing once, seriously, the way you would take any diagnostic seriously.

One caution, and it matters. A quick search turns up X-Rite's free online version of the same test, and a lot of photographers stop there, score reasonably well, and walk away assuming their color acuity is proven strong. X-Rite's own page undercuts that assumption directly. Their online challenge states plainly that it is "not a replacement for the full test," and lists the reasons why in the same breath: your score shifts with the lighting around you, the colors on your screen and desk, your fatigue level, even your age. A monitor test, taken in whatever ambient light happens to be in the room, with a screen that is itself a variable rather than a controlled instrument, cannot isolate the one thing the physical test is built to isolate: your eye, and only your eye. Doing well on the free version tells you your eye survived a fairly forgiving, uncontrolled screen-based approximation of the real thing. It does not tell you what the actual test, under the actual conditions it was designed for, would show.

That gap is not a technicality. It is the entire argument of this series in miniature. A digital approximation of a perceptual test, taken on uncontrolled equipment, will always flatter you a little. The real test does not care what monitor you own.

The Point Of Owning The Box

A calibration target tells you whether your monitor is behaving correctly. It has never once told you, and was never built to tell you, whether the person sitting in front of that monitor can actually see the difference between two closely related hues. Munsell built the vocabulary for that question in 1905, working entirely by hand, decades before anyone needed a coordinate system for it. Farnsworth built the test for it in 1943, working directly from Munsell's foundation, entirely independent of the mathematics a different set of scientists were simultaneously working out to translate Munsell into CIE terms. Neither of them needed a profile. A photographer chasing better color today still doesn't, not for this particular question. What they need is the box of caps, the controlled light, and an honest accounting of what their own eye can and cannot do.

Where This Comes From

A Colorful Life, Science History Institute
Munsell colour system, Encyclopaedia Britannica
Farnsworth-Munsell 100 Hue Test, X-Rite
Farnsworth-Munsell 100 Hue Test, Pantone product page
The X-Rite Color Challenge and Hue Test, X-Rite, free online version referenced for its own stated limitations
Understanding Modern Art, Module 03, Scott Parker Photo, 6-part course