Help with Gradient in Group Stacking DSS NGC 6359

[Ramsees]

Hi Folks,

I have never attempted stacking images over several nights,I have been trying to capture NGC 6357 and done so over 3 nights with over 5 hours of Data using Hyperstar on my Celestron CPC 9.25 HD Edge . However ,I have made the group stack with DSS but for some reason I am getting this Gradient that cannot be removed thru Gradient Xterminator or anything else for that matter. It looks like a rectangle shape. Possibly something to do with the stacking, not sure but I used Median for flats and Darks and Kappa sigma for light frames. However when I stack the nights separately they seem to be fine. Any ideas what is happening here. Is there a way we can manually stack the 3 together with photoshop to generate the same end result.

Group Stacked with all 3 nights together Gradient showing once you stretch the image

https://www.dropbox.com/s/4yg6mfu69eh1ovu/NGC-6357-All.TIF?dl=0

First night all good (No Rectangular Gradient)

https://www.dropbox.com/s/7dqinmzfliohmjt/NGC-6357-1.TIF?dl=0

Second night all good (No Rectangular Gradient)

https://www.dropbox.com/s/n3fkkjym6yfm74n/NGC-6357-2.TIF?dl=0

Third night all good (No Rectangular Gradient)

https://www.dropbox.com/s/yajw7iwje2b0ma8/NGC-6357-3.TIF?dl=0

I have put a massive amount of work here but this funny looking gradient will not go away frown.gif

Any ideas please assist

Thanks

[luka]

Your links are not working so it is hard to say.

[Ramsees]

I will fix them , not sure why as they are Dropbox links . My picture showing the gradient is attached above

[Ramsees]

I have fixed the links, hope you can help. I followed this procedure as below but the end result gave me this gradient :

File Groups
File Groups may be used to simplify multiple nights on the same object file management by logically grouping files for each imaging session.
If you use only the Main Group DeepSkyStacker is working exactly like before the introduction of File Groups.

There are two kinds of File Groups: the Main Group and all the others groups.

Light Frames from the Main Group may only be associated with Dark, Flat and Offset/Bias Frames from the Main Group.
This is the behavior of DeepSkyStacker before the introduction of the File Groups.
Dark, Flat and Offset/Bias Frames from the Main Group may be associated to Light Frames of any group.
Dark, Flat and Offset/Bias Frames from others groups may be associated only with Light Frames of the same group.
You can create as many File Groups as you want knowing that a file may belong to only one File Group.
When you start DeepSkyStacker only the Main Group is available. As soon as you add a file to the last available group a new empty group tab is created.

Example:
You shot the same subject two nights in a row.
For each night you have a set of Light, Dark and Flat Frames but the temperature was not the same each night and the Dark Frames are not compatible and the orientation was slightly different so your Flat Frames are also different between the two nights.

To associate each Light Frame with the good Dark and Flat Frames you just need to put all the Light+Dark+Flat Frames from the first night in one File Group and all the Light+Dark+Flat Frames of the second night in another File Group.
Since Offset/Bias Frames are common to all the nights they must be put in the Main Group.

DeepSkyStacker will automatically associate the Light Frames of the first night with the Dark and Flat Frames of the first night and the Light Frames of the second night with the Dark and Flat Frames of the second night.
The Offset/Bias Frames from the Main Group will be associated with the Light Frames of the first and second night.

[luka]

As far as I can see it has to do with the rotation of the images. Not all images cover the same area of the sky. For example, if you rotate image 2 to match the orientation of the image 1 you will see that it gets cut off where your left-side gradient-line is.
In other words, some parts of your final image have stacked only 1-night worth of data, some parts have stacked 2-nights worth of data while the rest (middle section) has all 3 nights of data.

In DSS can you try to go to stacking options and on the first page of stacking parameters select "Intersection mode" instead of "Standard mode". I have a feeling that will fix the problem, although your final image will be much smaller.

[Ramsees]

Hi Luka, really appreciate your help here and comments. I will try your recommendation but I don't like the idea of making the image smaller as that defeats the purpose of Hyperstar imaging with the wide field. I am not sure how everyone else does it but I have to assemble and dismantle my gear every night hence it seems I cant really take the advantage of imaging over several nights if that is the case.

Is there any other method ?

[luka]

I have attached the image from the 2nd night (left) rotated to match the final image (right, extremely stretched). In the left image you will see how it ends along the gradient-jump line of the right image.

What I always do when imaging over multiple nights is to try to rotate the camera to the identical orientation relative to the scope. Assuming the scope always goes onto the mount the same way, you will be imaging the sky in the same orientation. This way the rotation between images is kept minimal. You seem to have about 110 degrees rotation between the image 2 and the final image which gives you quite a significant area of non-overlapping parts. The final image matches the orientation of the first night.

Your nights 2 and 3 have almost the same orientation but your first night is quite different. If you don't like the extreme cropping you will get by choosing "Intersection mode" you could try stacking only nights 2 and 3.

Keep in mind that I may not be correct. Try stacking first and see the result.

[Ramsees]

Hi Luka,

Really appreciate your help here, I see your point now. I just tried your method but it still did not work so now I will try and stack nights 2 and nights 3 and will let you know. I thought DSS matches all the stars in the image and makes the rotation so they overlap but it seems this is not happening as intended

[luka]

I was just testing the "Intersection mode" here with my data and, after stacking, it did not crop out the areas where data from all images was not present. Actually I don't think it was any different from the "Standard mode".
So it is not doing what I thought it should and what I think is shown in the mode-selection image. See how you go with nights 2 and 3.

[Ramsees]

It worked , stacking nights 2 and 3 did the job with some slight gradient which can be cropped , I will now target to keep the images the same over the nights by centrting an obvious star or capturing the screen so next night I have it very close . Thanks heaps for pointing this out to me . A valuable lesson learnt here indeed from fine people like yourself 👌🏼

[luka]

Glad it worked!!!