Munsell: the painter who put colour in order

Albert Munsell was a portrait painter and art teacher from Boston, and he didn't trust colour names. So he spent two decades turning what his eye saw into numbers. His 1905 book and his 1915 Atlas gave every colour a written address, which means you could describe a colour without ever naming it. Scientists re-measured his system after he died. And soil surveyors still use it today.

Who was Albert Munsell?

Before any of the colour work, Albert Henry Munsell was a painter. He was born in Boston on 6 January 1858, and by 1881 he was teaching at the Massachusetts Normal Art School, the same school where he had trained. He stayed for more than three decades, teaching drawing and painting from the figure, composition and anatomy for artists.

From 1885 to 1888 he studied in Paris, first at the Académie Julian and then at the École des Beaux-Arts, and he showed work at the Salon. He painted portraits, and he painted the sea. He loved boats. In fact, years later, the first printed charts for his Atlas reached him while he was out cruising on his own boat.

The colour idea itself goes back to 1879, when he was still a student. As he remembered it later, he read Ogden Rood's book Modern Chromatics and built a spinning model out of two pyramids joined base to base. He coloured the corners red, green and violet. Those are the primaries of light, not the red, yellow and blue of the school paintbox. Even his 1885 painting Chloris Calls arranged the colours of the spectrum around a circle.

Why did a painter want colours written as numbers?

Because colour names kept letting him down. His book opens with Robert Louis Stevenson in Samoa, writing a letter and stuck on a red he just couldn't put into words: "I'll be hanged if I can describe this red."

Then Munsell lists the names people actually used: baby blue, Nile green, lemon yellow, royal purple, plum, and even automobile. Names like these, he wrote, end up "conveying different ideas to different persons". He made fun of an old German scale of blues from 1772, with steps called Blue being born and Hellish blue. And when he showed people a true middle red, they told him it wasn't red at all. It was terra cotta.

His answer was to treat colour the way music treats pitch. Music had a written notation long before colour did, and he kept coming back to that comparison. A colour, he thought, should have an address you could post to a stranger in another city. A nickname only works among friends. If you've ever sent a client "a warm teal" and got back something that wasn't your teal at all, that's the problem Munsell was trying to fix.

Mind you, he wasn't above a nickname himself. On 2 April 1900, debating in his diary whether yellow-green needed a single word, he suggested one: lettuce.

A side note: Munsell's book prints the date of the Stevenson letter as 1902, and that's a misprint. The letter, to Colvin, is dated 8 October 1892 in the published Vailima Letters. If you quote the line, use 1892.

Measuring what the eye sees

Around 1898 Munsell started experimenting in earnest. He spun a child's globe painted with colours, and he built a light meter, which he later patented. What he found was that the grey scales printed for schools disagreed with each other by more than 10 per cent. His tests even turned up a difference between a person's left eye and right eye.

That's why he didn't pick the middle of his grey scale himself. He chose it by "the average choice of artists, dyers, salesmen and students". In other words, he measured what people see and averaged it across a crowd, rather than trusting one person's eye, even his own.

On 29 March 1900 Munsell carried his spinning spheres to Columbia University to show Ogden Rood, the physicist whose book had set him going back in 1879. The spheres turned on a machine at up to 3,000 turns a minute. Rood called them very successful, and the diary records his verdict: "You have put an artistic idea (sequences) into scientific form."

That same year his first colour patent was issued, for a colour sphere for teaching with six main hues. He soon settled on five instead, so the system could count in tens. Then a Boston retailer, William Filene, told him shops wanted a standard system too. In 1905 Munsell published A Color Notation, with an introduction by H. E. Clifford of MIT, and a second edition followed in 1907.

The parts that weren't his

The tidy parts of the system weren't Munsell's, and he said so himself, in his own book. In a footnote he explains that he came up with the colour sphere without knowing others had used it years before. It was common ground among psychologists, he wrote. Philipp Otto Runge had described one in 1810, and Lambert had suggested a pyramid.

The three words his system is famous for weren't his either. The glossary at the back of his book takes its definitions straight from the Century Dictionary. And when the US Bureau of Standards examined his system in 1920, it pointed out that Bezold, Rood and Abney had all described colour by three qualities before him. (What colour is made of explains what those three are.)

So what was his? Historians generally credit Munsell with the measuring. He set separate scales by eye, in steps that look equal. He made colour strength a measured scale with no fixed end. And he gave the whole thing a practical written code. In his 1915 Atlas, a bright vermilion is 5R 4/10, which you read aloud as five red, four over ten. That's why it's fair to call his the first widely adopted colour system built from equal-looking steps, even though it wasn't the first colour solid.

Why is Munsell's colour solid a tree and not a ball?

The short version is that every hue reaches its greatest strength at a different lightness. By the 1907 edition of his book, the sphere was still there, but only as a simple shape a child could grasp. His fuller model was the lopsided "colour tree". Its branches stick out unevenly, and he left them open, so they could grow longer if anyone found stronger pigments.

By 1912 he'd given up on neat outlines altogether, writing that no regular shape would fit. His 1915 Atlas put it plainly: "branches are of uneven length because pigments vary." In the Atlas, vermilion runs ten steps out from grey at a middle-dark level. Yellow runs nine steps, but only near the top, where it's light. Blue manages six. A sphere is roughly what you'd get if you sawed every branch back to the length of the shortest one.

Munsell's last years

They started busy and ended hard. In December 1911 he lectured to the American Psychological Association in Washington, and they asked him to give the talk again before the meeting was over. He took the system to a congress in Groningen in 1913, then to Paris and Berlin in 1914. On that trip he fell ill and had an operation in London. He came home, as the company's later history puts it, more or less an invalid.

The Atlas came out in 1915. Then in 1917 an operation for appendicitis nearly killed him. His diary notes that he wasn't expected to survive, and that two nurses stayed all summer. He gave up his Boston studio, which John Singer Sargent later took over. The A. H. Munsell Color Company was proposed in March 1917 and incorporated on 6 February 1918, and Munsell died a few months later, on 28 June 1918, aged 60.

Who checked Munsell's work after he died?

Government scientists did, and his eye mostly passed. Only weeks after his death, the company asked the US Bureau of Standards to look over the system. The Bureau's 1920 report praised "the colossal task of producing an atlas". But it also found that his lightness numbers followed the square root of reflected light, not the straight percentages his Atlas had printed.

His son, A. E. O. Munsell, took over the company in 1921. That October a young woman named Dorothy Nickerson joined as his secretary and lab assistant. She went on to become a leading colour scientist, and later she wrote the company's history. It was never a money-spinner, either. In 1940 she wrote that the company "never has made a profit, nor has it ever employed a salesman".

The Munsell Book of Color replaced the Atlas in 1929, and then came the big re-measurement. From 1935 to 1940, 41 observers judged the colour chips, each of them over a year or more, which added up to what the report called "roughly three million" judgments. In 1943 an Optical Society committee of S. M. Newhall, Nickerson and D. B. Judd smoothed all three scales and tied every notation to a measurement in the 1931 international colour system, the one the 1920s experiment behind every screen colour produced. After that, the chips were re-made to match. Think of an instrument tuned by ear, then checked with a meter and a few strings re-tuned, and the tune itself survived.

Worth knowing: Munsell's eye beat his arithmetic. His 1915 Atlas said its middle grey reflects 50 per cent of the light white does. By the Bureau's 1920 rule, it was nearer a quarter. And under today's standard, middle grey reflects about 19 per cent. The grey his eye picked was darker than half the light, and the later science kept that dark grey, not the printed number.

Where is the Munsell system still used today?

Its most visible job these days is naming the soil. The US Department of Agriculture's Soil Survey Manual records soil colour in Munsell notation, such as "brown (10YR 5/3)". A surveyor matches a sample by eye against about 250 colour chips and writes down the code. Archaeologists still rely on the same charts.

The government also used his system to referee the very names he distrusted. In 1955 the Bureau of Standards published a dictionary of about 7,500 colour names, and it defined all 267 of its name blocks in Munsell terms. That's the dictionary that shows "royal purple" landing in five different categories across four colour references. Which means the system built to escape colour names ended up keeping them honest.

The ownership moved on as well. The company passed to a foundation in 1942 and sold for US$235,000 in 1970. Then in 1983 the foundation was wound up to fund a colour science laboratory at the Rochester Institute of Technology. That lab now hosts Munsell's diaries (opens in a new tab), and you can read them for free. The student's spinning toy from 1879 is in there, in his own words.

Matching by eye has limits, and the soil manual says so. The USDA manual notes that field readings repeat only to within about two and a half hue steps, and one step of lightness or strength. The code still beats a name, though, because two people can see exactly how far apart their readings are.

What should you do differently with colour after Munsell?

Give your colours an address, and then check your steps with your own eyes.

1. Never hand over a colour by name alone. "Teal" means something different to your printer, your developer and your client. Send a code with every colour in a brief or brand guide.
2. Use the right code for the job. Screen colours get a screen code, but paint, fabric and print need a physical reference, matched under good light.
3. Expect paper and screens to disagree. There are colours paint can make that a screen can't show. Why your brand colour looks different on your phone explains why.
4. Check that equal steps look equal. Munsell's own printed numbers were off while his eye was right. Screen colour has modern scales chasing the same goal, and Why "same brightness" colours don't look the same and How one engineer's blog post changed colour on the web cover them.
5. Read the original. A Color Notation (opens in a new tab) is free and short, and Munsell is good company on the page.

Frequently asked questions

Is the Munsell colour system still used?

Yes, and most visibly in soil science. The US Department of Agriculture's Soil Survey Manual says most soil survey organisations use Munsell soil colour charts. A surveyor matches a sample by eye against about 250 colour chips and records a code such as 10YR 5/3. Archaeologists use the same charts for soil and pottery, so a reading in one field notebook means the same colour to a reader anywhere.

What's the difference between Munsell and Pantone?

Munsell is a map, whereas Pantone is a catalogue. Munsell gives any surface colour an address made of its hue, how light it is and how strong it is, on scales spaced to look even to the eye. Pantone is a numbered library of specific ink colours that printers mix to a set formula. So they answer different questions, though at the time of writing X-Rite's own timeline lists both brands as owned by X-Rite.

For the curious

Munsell's three scales are hue, value (lightness, 0 for black to 10 for white, with the painted steps 1 to 9) and chroma (strength). The notation writes hue first, then value over chroma. The 1943 re-measurement is known as the Munsell renotation, defined in CIE 1931 terms, and its data files sit at RIT's lab. The Bureau's 1920 finding was that value was proportional to the square root of luminous reflectance. Today's grey curve is in ASTM D1535.

The lightness scale of CIELAB (1976) descends from cube-root formulas fitted to Munsell's value spacing (Glasser and colleagues, 1958). For greys, CIELAB's lightness is close to ten times Munsell value. That link covers lightness only, though. OKLCH lightness is not value divided by ten: middle grey N5 is about oklch(0.578). Ottosson used the value-5 ring as a check on Oklab, not to build it.

Japan's national standard, JIS Z 8721, uses the same notation and reprints the 1943 table without using the name Munsell. Later systems chose other goals. OSA-UCS (1974) aimed for equal steps to each colour's 12 neighbours, and the Natural Colour System (1979) aimed for resemblance to a few elementary colours.

Sources (17)

• A. H. Munsell, A Color Notation (opens in a new tab), 2nd edn, Geo. H. Ellis, Boston, 1907 (© 1905), introduction by H. E. Clifford; footnote 11, paras 2, 33–34, 65, 131, glossary
• A. H. Munsell, diary, typed copy 1920–23 from six handwritten volumes (1899–1918), RIT Munsell Color Science Laboratory (opens in a new tab); Vol. A1, entries for 29 March and 2 April 1900
• A. H. Munsell, "A Pigment Color System and Notation" (opens in a new tab), American Journal of Psychology 23, 236–244, 1912
• A. H. Munsell, Atlas of the Munsell Color System (opens in a new tab), Wadsworth, Howland & Co., Malden, 1915
• A. H. Munsell, US Patent 640,792, "Color-sphere and mount" (opens in a new tab), 9 January 1900
• R. L. Stevenson, Vailima Letters (opens in a new tab), letter to Colvin, 8 October 1892
• D. Nickerson, "History of the Munsell Color System and Its Scientific Application" (opens in a new tab), JOSA 30, 575–586, 1940; and "History of the Munsell Color System, Company, and Foundation" (opens in a new tab), Color Research & Application 1, 1976 (Wayback copy)
• National Cyclopaedia of American Biography, vol. XII (opens in a new tab), p. 316, 1904
• I. G. Priest, K. S. Gibson and H. J. McNicholas, An Examination of the Munsell Color System. I (opens in a new tab), NBS Technologic Paper 167, 1920
• S. M. Newhall, "Preliminary Report of the O.S.A. Subcommittee on the Spacing of the Munsell Colors" (opens in a new tab), JOSA 30(12), 617–645, 1940
• S. M. Newhall, D. Nickerson and D. B. Judd, "Final Report of the O.S.A. Subcommittee on the Spacing of the Munsell Colors" (opens in a new tab), JOSA 33(7), 385–418, 1943 (abstract)
• K. L. Kelly and D. B. Judd, The ISCC-NBS Method of Designating Colors and a Dictionary of Color Names (opens in a new tab), NBS Circular 553, 1955
• USDA NRCS, Soil Survey Manual (opens in a new tab), Agriculture Handbook 18, ch. 3 "Soil Color", 2017, amended 2018
• Glasser et al., "Cube-Root Color Coordinate System" (opens in a new tab), JOSA 48(10), 736–740, 1958 (abstract)
• B. Ottosson, "A perceptual color space for image processing" (opens in a new tab), 23 December 2020
• RIT Munsell Color Science Laboratory, home and history (opens in a new tab)
• X-Rite, "Our brands" timeline, retrieved 23 September 2026