In 2002, astronomers averaged the light of 200,000 galaxies and declared that the universe was pale turquoise. They then traced the color to a software error and corrected it to beige, calling it cosmic latte.

The universe is the color of weak coffee

In 2002, Johns Hopkins University astronomers Karl Glazebrook and Evan Baldry averaged the light of more than 200,000 galaxies and announced for the first time that the collective color of the universe is Pale turquoise. Within weeks they had it back. An error in their color software skewed the result, and the corrected answer was an off-white color that the public quickly dubbed the cosmic latte.

Color was never the goal of research. It began as a footnote to a serious survey of light-emitting galaxies, and briefly became one of the most talked-about results in astronomy. The idea that the universe has one average shade spread in newspapers, and home decorators paid attention to the color derived from each galaxy simultaneously. When the retraction happened, it was almost as widely publicized as the original claim.

How do you calculate the average color of the universe?

The two were studying the cosmic spectrum, created from the Australian 2-degree Galaxy Redshift Survey, which measured the spectra of more than 200,000 galaxies. By combining tens of thousands of those spectra, they produced a single, average-sized, luminosity-weighted result, which is actually the average color of light emitted by the nearby universe. Their real goal was to read the history of star formation from this shared light, where young blue stars and old red stars leave different fingerprints.

This combination is why the spectrum fell where it did. It is a mixture of hot young blue clusters and cool old red clusters, and the average of the two is neither strongly blue nor strongly red. The paper noted that the resulting color was not the kind that a single incandescent object could produce, which was the first hint that converting the spectrum into a single perceptible shade is less straightforward than it seems.

On a whim, Glazebrook decided to figure out what this spectrum would look like to the human eye, using standard color matching functions that describe how people perceive different wavelengths. Combining the cosmic spectrum through these functions resulted in a blue-green color that the team declared to be “a few percent greener than pale turquoise.” Announced at a meeting in January 2002, the idea that the universe has an average color, and that the color is green, immediately captured the public’s imagination.

White point error

Not long after news broke of the turquoise claim, Mark Fairchild of the Munsell Color Science Laboratory at Rochester Institute of Technology reached out. Fairchild wrote, almost verbatim, the book on the subject, a 1998 text called Models of Color Appearance. He and his colleagues found that the free software the astronomers used had been tuned “white dot” Incorrectly. White point is light that the eye treats as normal white, and shifts with the surrounding environment, slightly yellow under household lamps, and a little blue in some monitors. The wrong setting has colored the entire result green.

A color scientist in the same laboratory, Francis Imai, put the problem in everyday terms. “Tungsten lights, for example, can make the white dot a little yellow,” he explained. “Some screens have a bluish white dot. Your visual system tries to adapt, so you assume the color is white, but it’s actually yellow or bluish.” The freeware effectively chose the wrong white reference, and the error shifted the entire result towards green.

When the white dot was fixed to a person viewing the light in a dark room, the color came out beige. After setting it to daylight, it turned dim red, and in indoor lighting it turned blue. Baldry made the dependence vivid: “If you looked at all the light in the universe from a room with red neon light, it would appear turquoise.” “But this is not a standard perspective.”

Glazebrook was not afraid of the mistake. “It’s our fault because we didn’t take color science seriously enough,” he said. “I’m so embarrassed. I don’t like being wrong, but once I found out I was wrong, I knew I had to get the word out.”

The astronomers posted their spectrum data online and invited anyone to recalculate the color. Among a series of proposed names, the cosmic latte stuck, chosen for its milky off-white color.

What does color actually mean

The “color of the universe” is not a property of space such as temperature or density. It’s a perceptual quantity, and the turquoise blunder showed exactly why that’s important: the same spectrum reads beige, red, or blue depending on the lighting in which you imagine yourself standing. There is no neutral place to view all galaxies from, so there is no one true color, only a color for a specific set of assumptions.

It is also an average, weighted in terms of luminosity across single scan galaxies at modest distances, not a count of every photon the universe has ever emitted. The galaxies in the 2dF sample are relatively close in cosmological terms, so the figure describes the light of the present universe rather than the redder glow of its former distant regions. This was, by the authors’ account, a fun aside to a paper on how quickly stars form over cosmic time. Beige is real enough, but it describes a mathematical and conventional process, not a wall you can paint to match the sky.

The naming was also unofficial. Cosmic latte was once a popular choice rather than a scientific term, appearing more in color value exhibits than in astronomical literature.

Beige is worth keeping

What’s left is the color-less patch. Two researchers made a clear, widely repeated claim, and learned within weeks that they had gotten it wrong, saying so in plain language rather than letting the nicer answer stand.

The universe, when measured and observed in the dark, is approximately the color of milky coffee. The most lasting lesson is how quickly beige is replaced by turquoise once someone examines the color science.