The reinterpretation of the digital image on the left using primary colorants derived from emission nebula and black-body radiators (the HOT primaries, right) does NOT result in a "calibrated" color rendition. Here's why. The short answer is that the appropriate levels of green are missing! The film that recorded this picture was almost blind to the oxygen-3 green spectral line and so it is under-represented in this scene. A proper view of the light coming from the center of M42 would show a yellowish hue, the result of mixing the green O3 and red H1 emissions from this region of the nebula. In general, the levels of each primary, not just green, are incorrect, and here begins the long answer. Consider the process that was involved in reproducing the light from the original scene (the view in the telescope) to its display on your monitor. A color film recorded the light according to the sensitivities of its three emulsion layers, and represented the exposure as concentrations of three colored dyes. A scanner measured how much light from an incandescent light bulb was transmitted through the superposition of these dyes (i.e., through the film), and then represented what it saw using digital levels intended to excite three colored phosphors. We then took that result and, rather arbitrarily, declared that the three colorants were those of the HOT primaries. In order to simulate what they would look like, a color transform modified the digital values in the image so that a Trinitron monitor could display them (to the best of its ability). Now what is the likelihood that the perception of the light levels at the end of this tortured path are the same as would be perceived by amplifying the light at the telescope? Almost zero! Not that we couldn't, in principle, obtain the right levels. This would require careful measurement of: film spectral sensitivities film dye spectral transmissions scanner sensitivities phosphor primaries The first of these is particularly difficult because of reciprocity law failure and untangling the three sensitivity bands and the three dye layers in a color film. It would be a bit easier to accomplish with color separation methods using black and white film, or CCD exposures. It is possible to acquire HOT images however. Steven Lee, at the Anglo-Australian Observatory, carefully created the image below using the correct filters for each primary. Here one can appreciate the emission regions of this nebula in its "correct" color (as best it can be displayed on your monitor). Compare the image which is displayed directly (left) to one which has been modified to simulate its appearance on an HOT display. It's not my opinion that a calibrated image is what we want for astrophotos (unless we are doing a scientific measurement). Instead we want to portray a beauty that cannot be seen. If applying the HOT primaries to an image improves your perception of this beauty, then it is the right thing to do!   A true HOT image rendered in sRGB. The same digital image interpreted as HOT primaries (simulated).

Copyright 2000-Jun-11 Thor Olson

home

previous | next contents | techniques