5 April 15
The Perils Of Color In The Digital Age
Last fall we bought a digital DSLR, and this has led me to dabble in a bit of astrophotography, even buying a good equatorial mount to facilitate this. Astrophotography is an interesting hobby. Practiced at its upper echelons (which I have no aspirations towards), it calls for extremes of technique, patience, and expensive gear. The aesthetic which many astrophotographers aim for is a quite colorful image showing many subtleties in the details of a nebula or other deep-sky object.
The use of color in this manner is in most ways an illusion, at least from the point of view of human visual perception. Cones, the anatomical structures in the retina that respond to color, are not sensitive to light at low levels, so looking through even a very large telescope at the nebula one will not discern the color brought out in the photograph. The color in the image may be useful in terms of scientific visualization, for instance illustrating different emission spectra, but it’s not what the eye can see.
Astrophotography is perhaps an extreme example, but I’m philosophically confounded by the question of representation of color in the digital age. Consider what might be the modal life of a digital photo shared with others. A snapshot taken with one’s smartphone, then posted to the web via Facebook or Instagram. The issue is of calibration. How does one ensure that the color intended by the photographer is what is seen by the recipient? Without both the creator’s and recipient’s monitors calibrated to a common standard, one cannot. These standards exist, but 99.99% of the time both parties are not so calibrated. Even enthusiastic photographers tend to spring for new lenses in lieu of color calibration equipment. One can adjust color curves in image software to one’s heart’s content, but without calibration both sides, one’s artistic intent in terms of color cannot be reliably shared.
And what if the intent is simply to record the colors one sees? The problem persists and is refracted twicefold. First, how faithfully do the levels of color recorded by the digital sensor represent the spectra of light being reflected by the object? Second, what are the colors being displayed by one’s monitor?
A solution I often adopt is to directly record the color in watercolor paint. With sufficient practice in color mixing, this can work reasonably well. Soil scientists have a more scientific answer to this problem — they go to the field armed with Munsell color swatch books to match and note the color of a soil unit. And the astrophotographer in me wishes there was a market for black-and-white consumer digital cameras: these problems go away, and cameras without RGB filters are more sensitive to light.
It’s also not clear where octarine fits in any color calibration system I know of.
[Image below is one of my attempts at astrophotography, being of the Orion Nebula.]