OSC colour calibration

[alan4908]

I've always found achieving the correct colour balance very difficult when post processing images. I think that I might have found the origin of the problem. In my case, it would appear to be the RGB colour conversion coefficients that are used when debayering my OSC image. For my particular camera (Trius SX26C) I'd previously had these set at the default values of R=1, G=1, B =1 in my stacking software programs (Maxim DL and CCDstack2+). Recently, I came across a freeware program called eXcalibrator (http://bf-astro.com/eXcalibrator/excalibrator.htm) that automatically provides a colour calibration for a given plate solved image. I tried it on a M31 stacked image set and the result for my camera using Maxim DL was R =1, G = 1.18, B = 1.58. In subsequent post processing, the colour balance did seem much better. I was quite surprised about the relative sensitivity difference insensitivity (58%) between the Red and Blue channels.

I presume that the optimum RGB coefficients are a function of the debayering program, camera and "average" amount of atmosphere through which the image was taken. So, if you want to have the correct colour balance, it implies that you should colour calibrate every image prior to post processing. Is this correct ?

Alan

[ollypenrice]

I don't think it matters whether your data was from OSC or mono-with-filters. It's all the same beast. (Of course you must correctly attribute the OSC Bayer colour positions so the software knows what's blue, what's green and what's red but that is just a one-off trial and error test. If it's wrong you'll know it's wrong! I did it by taking a daytime pic in my kitchen.)

Personally I combine my separate colours in AstroArt all equally weighted at 1:1:1 to make an initial RGB.  (AstroArt does offer to colour calibrate them and set its own weighting but it does this very badly so I decline the offer.) I then take this linear image into Pixinsight and run Dynamic Background Extraction. While the primary aim of this routine is to give a neutral background sky (R, G and B all at the same brightness) it also gives a superb colour balance right up the brightness scale. I rarely touch it thereafter.

I don't think there's any need to be surpirsed by the weightings you found. When I look at an equally weighted RGB image from our setups it isn't disastrous but it certainly isn't right. DBE sorts it out, though. As an aside I often found my OSC camera gave a distinct red-green imbalance either side of  a diagonal line dividing the chip. I've no idea why it did this and, again, DBE dealt with it perfectly.

Olly

[alan4908]

Olly - thanks for the reply. I don't have Pixinsight but maybe I should try it, even if it is just to compare it with PS equivalent. The DBE function in Pixinsight sounds interesting !

Alan