|
previous | next |
click for a larger image |
Colorimetric Veil This image was the result of a collaboration with Mike Cook, who was brave enough to try my earlier color work (HOT color space) on some of his CCD image data. It is a composite of two color transforms, neither of which is HOT, but were inspired by discussions with Mike on how to accurately present his multiband recordings. I consider this image to be colorimetrically correct, within the accuracy of my model for Mike's filter and detector, and the limits of the display to portray an sRGB image. It is quite different from most images of the Veil, and my color processing will have to be confirmed in other tests, but the approach is from well-established principles in color science. |
|
|
Mike uses an SBIG ST-10 sensor on an Astrophysics 130mm f/6 refractor, usually from his suburban driveway. He has fitted his filter wheel with the CFW8 red, green, blue wideband filter set, and also has installed three narrowband filters to record spectral emission lines H-alpha (656nm), ionized oxygen OIII (501nm) and H-beta (486nm). His target was the Veil Nebula (NGC6992/6995), a supernova remnant that glows in these wavelengths. The Veil and nebulas like it are a color imaging challenge because the OIII and H-beta emission lines fall between the blue and green sensitivities of most color films, and also are missed by the blue and green analyzing filters in many RGB tricolor imaging systems. As a result, most pictures of the Veil Nebula show a brilliant red cloud hanging among the stars, the red H-alpha light easily captured by film and red-sensitive CCD cameras. A carefully obtained tricolor image however, can show hints of the delicate bluish-green and greenish-blue strands that weave through the nebula. This was Mike's goal, and he thought the HOT color space would reveal it from his image data. His result is posted at his website, where I was pleased to see an actual application of HOT, but I felt that there was still something missing. I became interested in how to translate his wide- and narrowband image frames to a conventional RGB display space, and Mike was kind enough to provide me with his raw data (in fact he insisted I take all 150MB of it!) In working with his six channels of the image I have learned quite a bit about the nature of filtered CCD sensors and invented or re-invented some methods for color correcting it. Here are a few of the lessons along the way. |
 Tricolor film image by Robert Bickel
|
|
 home |
previous | next |