The HOT Color Space

Chromaticity diagram for two display technologies: sRGB, and HOT.

More on chromaticity diagrams...

The primaries of most display systems are based on the chromaticities of phosphors or of absorption dyes. The HOT (Hydrogen, Oxygen, Thermal) color space uses primaries found in deep space.

You can build a color profile based on these primaries using Photoshop, or Colorsync, or several other ICC profile-making tools. The information you need is:

Primary

Source

Chromaticity

red

Hydrogen-alpha emission line,
656 nm

x =0.7287, y=0.2713

green

Oxygen-III emission line,
501 nm

x=0.0082, y=0.5384

blue

B-filtered, infinite-temperature,
black-body radiator

x=0.1491, y=0.0544

You will need two more things. First, the whitepoint:

white

Infinite-temperature, black-body radiator

x=0.2399, y=0.2341

And finally, a tonal response. Gamma 2.2 is appropriate for PC displays, 1.8 is a better choice for Macs.

Save your profile with a distinctive name. My choices were HOTwTinfG22.icm and HOTwTinfG18.icc.

The rationale for these choices is fairly simple. Emission nebulas are dominated by the light from energized hydrogen (H-alpha) and doubly ionized oxygen (O-III). These are strong red and blue-green components of the light coming from these objects.

Reflection nebulas, because of the scattering behavior of dust and gas are largely blue. They are not line sources, but rather more like the the dispersive color of our own sky. The scattering mechanisms are most effective for short wavelengths, and so they are primarily bluish. The brightest blue-light sources are high-temperature stars. These have broadband spectra with their peak at the extreme blue end of the visible range. As the temperature of the star increases, its color converges toward a particular bluish-white appearance.

I find it quite amusing that we can calculate the appearance of an infinitely hot star, even though such a star would be impossibly bright to look at! Nevertheless, its color is quite similar to stars we can look at, and so I have taken this conceptual color to be the whitepoint of this color system. It turns out to be pretty close to an uncalibrated CRT monitor (~10K degrees).

Further, if we filter that source by the B-filter of the astronomer's UBV filter set, we obtain a blue that works quite well as a primary. It is a happy coincidence that the B-filter matches the blue channel response of the human visual system, at least when looking at infinite-temperature stars!