Using the folder camera to reprocess images

[roblucas]

I've been trying to investigate how to improve my live stacked images by using the folder camera on previous sets of raw images from live stacks and adding in a master flat for example. In doing this I noticed that the stacked images from using the folder camera were nothing like the original stacked image. Commonly the colours would be quite different with a much greater blue and less red appearing. I have made sure to use the exact same set up, i.e. the set of raw images from the live session, along with the master dark frame that I used at the time. The camera is the SvBony 705c. I have tried this on several sets of images and always get this strange result. The results seem to apply to both the saved png and the fits file . If I try to process the fits files in Siril, the one that I saved live processes OK, and I can get much the same image as the corresponding png file; but the fits file saved from the folder camera stack, seems to be largely missing the red channel, (in Siril, when I look at the histogram stretch, the red appears to be missing).

I feel I am going round in circles with this, so any help would be appreciated.

[admin]

Hi,

the most obvious thing to check would be that the individual raw images seem to have the right sort of colours in them when you reload them into the folder monitor camera. You should be able to get at least a hint of this colour with a strong display stretch on the individual frames. This is dependent on the folder monitor camera picking up the correct bayer pattern for the image from the FITS headers of the FITS files it is loading (if you are loading TIFF or PNG then that needs to be set manually in the 'Debayer Preview' option - otherwise you will get a grey image with a fine grid pattern on it). Out of the four possible patterns, one will be correct, one will give the red/blue swapped and two will give odd pink/purple colour casts with a grid visible at the pixel level.

It's probably worth making sure that you are on the latest version of SharpCap as there have been bug fixes to the automatic loading of the bayer pattern at various times, so older versions might not be right by default.

If that doesn't help, then I think the next step would be for you to post some images showing the sort of results you are getting and to share a couple of raw frames and calibration files so that I can experiment to see if I get the same and/or find out what might be going wrong.

cheers,

Robin

[roblucas]

Thanks for that, it is the Bayer pattern that is the problem even though the raw images are saved as FITS. I searched the pattern on the internet for this chip and got RGGB from several sources, but that still gave me the wrong colours. It didn't occur to me that this could still be wrong. I set the pattern to Force GRBG which is what the FITS header says, and the colours are fine.

Thanks for the help.
Rob

[roblucas]

Thanks for that, it is the Bayer pattern that is the problem even though the raw images are saved as FITS. I searched the pattern on the internet for this chip and got RGGB from several sources, but that still gave me the wrong colours. It didn't occur to me that this could still be wrong. I set the pattern to Force GRBG which is what the FITS header says, and the colours are fine.

Thanks for the help.
Rob