RGB channels with humps in DSS

[pixueto]

Dear all,

I'm struggling with the way my pictures come up from deep sky stacker. The green, red and blue come up with two humps each. I did't had this problem before. Anybody has any idea why this is happening and/or whether I can fix it? I would really appreciate it that you could help me with that as I'm going mad over it. I'm using an astronomik clip LP filter in case that has to do something with it.

Thanks for your help.

[Digz]

Mine often display this trait.

Switch to the RGB/K tab and adjust the green and blue sliders so the peaks sit over the red one - you should see some improvement there. I then go back to the luminance tab and tweak the sliders until the upswing on the curve is shifted to sit over the aligned peaks.

HTH.

[allcart]

Good tip Lee. i hadn't worked out how to use the luminance slider yet. I'm new to DSS.

[pixueto]

Thanks, I always do that but any idea as to why I'm getting these double humps for each channel? That didn't use to happen before. I use a canon 350D with a LP clip filter.

Thanks

[Digz]

I must confess, I dont really know what Im doing and its all trial and error, but the above workflow seems to work for me 9 times out of 10 :-D

[Digz]

Thanks, I always do that but any idea as to why I'm getting these double humps for each channel? That didn't use to happen before. I use a canon 350D with a LP clip filter.

Thanks

Were you using the LP clip filter before? If the answer is no then this may be the cause. Unfortunately Im not sure exactly what would cause the 'split' peaks.

[pixueto]

Were you using the LP clip filter before? If the answer is no then this may be the cause. Unfortunately Im not sure exactly what would cause the 'split' peaks.

Yes, and I didn't get those before. I recently opened the camera to try to remove the IR filter but didn't succeed. I put it all as it was but I wonder if I might damage something. Nevertheless, when I don't use the filter, now it seems to be OK.

Thanks

[dph1nm]

Note that DSS can align the peaks automatically for you (switch on RGB Background Calibration).

NigelM

[ollypenrice]

It may be that you have some edge artefacts creeping in. I wouldn't do any stretching or post processing in DSS. All of that kind of thing I do in Ps. If you look at Levels while the whole image is active you get a false reading. It's best to select an inner region just a bit smaller than the whole image so that the borders are ignored. Then you get a meaningful Levels readout. Don't crop because if you do that you won't be able to do a restretch for a different part of the image at a later stage and drop it onto to what you have done so far. (I do this a lot.)

Once you have excluded the outer edge use the Levels box to align the top left of the peak in all channels. This doesn't have to be set in stone but it's a good starting point.

Olly

[martin_h]

What does the histogram look like in Photoshop? Just save the image as a 16 bit TIFF and take a look...Like Olly says its probably the edges of the stacked subs creating a border.

[jgs001]

You will get a colour cast caused by the CLS clip filter, and the image will be moved to blue. Use the RGB background calibration, and DSS will apply a custom white balance to the image, eliminating the colour cast, and you should get aligned colour channels.

In the RAW/FITS settings... uncheck both white balance options

Then in the lights tab, set the RSB Background calibration

The image in was shot with the CLS clip filter and stacked using the above settings, I didn't have to deal with the blue cast.

[pixueto]

I've been away for a couple of days and did not see all your replies. Thank you for all your help guys. I'll do what you say now and see what happens.

[pixueto]

Thanks Olly, I've used the rectangular marquee tool and selected the central area, leaving the edges away. The histogram does not present the double humps now. How should I go about processing this? Curves and levels as normal with the marquee tool selected?

Thanks

[pixueto]

I'll run the subs again through DSS with the setting that you are suggesting guys and see if the humps disappear in the final image. I'm not bothered about channels not aligned as I always get this result; the main problem is that I used to have a single peak curve for each channel and now I'm getting two. I wander why this is happening now and not before.