One-shot color image processing: HELP!!!

Hello everyone,

I'm trying to do some DSO imaging for the first time and I have some problems prcessing my ST-2000XCM images.

First of all, dark frames: since I don't like the way CCDOPS process the one shot color images, I decided to process the single color channels one by one and then joined them after processing. Should I do the same thing with the dark frame and then dark subtract each channel with the corresponding dark?
I have already tried to dark subtract my images (both systems, i.e. all cahnnels with their own dark frame and the LRGB image with a single dark) but my final image always comes out with too many colored specks here and there and I can't understand why. I attach an image of M33 taken last night (20 min at F6) that shows what I'm talking about. It has been dark subtracted with a median of 3 darks and no other processing: as you can see, there are a lot of green blue and red pixels that I don't know how to eliminate.

Anyone out there that can help?

Ciao all

Luca

Attached Image:

Hello Luca,
Other than verifing that the RGB and Lum frames have been aligned and combined properly you might want to talk to the folks on the Yahoo, sxv-osc group. They discuss mostly sx cameras but have questions answered on other cameras. Try a test processing without the darks to see if the darks are causing the problem. Also, I don't think this is a problem with your setup as it is a nice widefield, but if you are getting less than one-pixel resolution with a one-shot on smaller star the one-shot RGB matrix will not get sampling on all the colors and will come out as either a red or green or blue star. Either a dither guide or higher sampling rate should correct that. But again I don't think thats the problem. Hope I helped more than confuse. I like the image and you will only master this as you have done with your beautiful solar images.
Ciao

Ken Tydeck

Gianluca Valentini said:
It has been dark subtracted with a median of 3 darks and no other processing:

You need more than three darks to make a good master dark, Luca.
The software needs more images to properly identify what it bad data.
I would combine at least ten subexposures to make the master dark if I were you.
Sorry, I can't help you with your other questions.

dan kowall

Luca,I use the ST-2000xcm as well,with the same spots.99% of the spots disappear when I do my dark subtract.Try a single dark of the same length and temp.If some spots still remain the remove hot pixel tool in CCDops usually will kill the rest.

I dont see any need to split and dark subract the channels seperatly.
Stacking more shots also greatly reduces the amount of speckles.
I have a master dark of about 15 shots,it still doesnt work as well as a single dark of the same length and temp does, so I will keep taking darks to add to it.

Hello everyone. Thanks for your help guys: I believe that I found a way around the problem. At least it looks like it from the picture attached, which is not great, but at least it doesn't show the same speckles

Thanks again

Ciao
Luca

Attached Image:

Hello everyone!

Here is my first DSO image ever
I am quite satisfied of the final outcome, although it is far from being a good result in the standards of this forum.
Please feel free to comment and suggest.

Sum of 4 20 minute shots with Tak 102 @F6 and ST-2000XCM. I hope to be able to get some more data as soon as the weather gets any better (my area is currently getting FLOODS :C ).

Ciao all

Luca

Attached Image:

Well thank you Tyler, Ivan and Louie

I'm sure that, if you keep on helping me, I'll be able to meet the very high standards set by all of you guys! Sooner or later

Ciao all

Luca 8)