The web page outlining our color of the universe idea is here.
Here are some more details given in answer to an email I received. You might call this an FAQ.
Is it just speculation to pick out the colour of the Universe?
We do not think it is a matter of speculation. There are a few ways one could define an average color of the universe. We picked one that we think is appropriate and is described on the web page. First, we determined the average spectrum of galaxies (luminosity weighted and in the rest frame). This means it is the average spectrum in the visible range that escapes from galaxies. It is like looking down on an average galaxy and seeing the light combined. This is well defined and shown as the first plot on our web page. The CIE coordinates are also well defined. The colour of an illuminated source depends on the white point. And we picked the E white point (flat flux per unit wavelength) and we state clearly this is a choice. The one point we did not make clear is that this only gives a 'hue' and we chose to maximise the RGB values while keeping the same 'hue'.
What about light from the Sun, Moon, planets, indeed the entire Milky Way,
and all terrestrial light sources?
In a cosmic volume, the light from our Sun and its planets is negligible. We actually also ignore our own galaxy and assume we are outside it; this is called correcting for 'Galactic extinction' but this effect is small because extragalactic surveys tend to look away from the Galactic plane (where most of the dust is).
Does not the Universe consist mostly of transparent space, which is actually
close to black in color because it contains far more interstellar and
intergalactic non-luminous gas, dust, and perhaps dark matter, that absorb
light than it contains stars that radiate light?
Black is just the absence of light. When we look at the night sky with our eyes, it is black nearly everywhere (assuming away from light pollution) because our eyes are not sensitive to faint galaxies. However, we can now measure the spectra of faint galaxies and imagine that light boosted to maximum while maintaining its 'hue'. Almost no visible light is absorbed by the inter-galactic medium.
But still, the average color of the entire Universe is probably close to
black, isn't it?
The extragalactic sky is black to the naked eye. However, black contains no information other than galaxies are too faint to see to the naked eye. The way we have defined the colour, the hue, derived from the cosmic spectrum is related to the history of star formation in the universe. It would not be interesting or informative to say the universe is black.
What about hue shifting due to interstellar absorption and re-radiation
We consider the average spectrum of the universe to be observed outside a galaxy. The interstellar absorption within each galaxy is implicitly included because absorbed light does not escape. Our own Galaxy's absorption is not included. Re-radiation tends to be in the infrared, which is not visible to the naked eye.
How do you chose to average the light?
There is only one way to determine our cosmic spectrum as defined, and that is a luminosity-weighted mean. This is defined in the paper and is the logical choice for a 'cosmic spectrum' in our opinion: the contribution from each galaxy is proportional to the amount of light it gives out. In other words, it "represents all the sum of all the energy in the local volume of the universe emitted at different optical wavelengths of light". See abstract of our scientific paper for another description. It is possible to define such a 'cosmic average' because the universe is homogeneous on (very very) large scales. Imagine a box 100 million light years on each side! Wherever you place the box in the universe you will get the same cosmic average spectrum. In fact, the box could be smaller, 10 million light years on each side, and the colour would be the same or very similar wherever you placed the box.
Why was the CIE system chosen, which, although a color measurement standard,
is not obviously 'best' nor is it free from distortions caused by the
attempt to represent colors as three-dimensional values and the further
attempt to represent those three-dimensional values in two dimensions?
The spectrum was the main point of our research. We acknowledge the fact that representing a colour to the human eye is not straightforward, the CIE system allowed us to define a 'hue' from a spectrum.
Will the true average color depend critically on the distribution of
chemical elements throughout all the stars and nebulae, which is currently
Ours is not a theoretical spectrum but an observed one so does not depend on the details in this way.
Why tell the public that the average star color is beige, when the 'true'
average color depends rather heavily on the particular averaging and
measuring methods that are chosen?
We thought it was interesting to make a choice in order to promote the concept of the cosmic average spectrum, which is well defined, and the constraints on the history of star formation. This is explained on our web site and was followed up by a number of articles originally in 2002. Though admittedly not all journalists have chosen to promote the science side of our concept.
Ivan Baldry, 2010