DSS help

I use DSS for calibrating and stacking images, in some images I noticed poor stacking (one corner where the coma is more evident). Any idea how to improve the quality of stacking in DSS? Any special setting?

Adel Kildeev:
I use DSS for calibrating and stacking images, in some images I noticed poor stacking (one corner where the coma is more evident). Any idea how to improve the quality of stacking in DSS? Any special setting?

Hello Adel,

DSS is a fairly intuitive software and very few settings need adjustment - maybe the number of detected stars is the main one. Other than that, the standard settings are ok.
In the case you mention, the problem may not be the stacking but the frames itself - I don't see a way that one specific area has problems but not the other areas - do you have issues in other corners or in the center?

André

Andre, thanks for your reply. Please see the example https://www.astrobin.com/fr272y/. The problematic is the right lower corner where the coma is more visial. May be I should check the focuser?

Adel Kildeev:
Andre, thanks for your reply. Please see the example https://www.astrobin.com/fr272y/. The problematic is the right lower corner where the coma is more visial. May be I should check the focuser?

I really don't think this is stacking issues but most likely field distortion. By the way, refractors don't have coma but instead field curvature and to avoid that, you need to add a field corrector, unless you have a quadruplet or above. For what I see, you have a triplet and you don't mention any field corrector so my guess is that DSS is trying to align the Ha and Oiii frames but due to different field curvatures, it can't do it, at least in full frame. Also, I see the same issue in other photos of yours.

If I'm correct and you really don't have a field corrector, I'd suggest you to get one to minimize field curvature. But if you do, pay attention to backfocus distance. See this article for further information: https://optcorp.com/blogs/deep-sky-imaging/how-to-set-the-correct-back-focus

I hope this helps but let me know should you have any doubt.

Excellent article, very usefull, great thanks! I use Orion flattener, sorry, omitted to mention in the details. Without flattener field curvature (you are right again) is visible in all corners. With the flattener the only one is still suffering. I have almost 55-56 mm back focus, so It may be the scope problem. Do I need callimation or focuser adjustment? I doubt that I should mess up with the lenses by the myself.

Hi
Just an observation: I had similar and even more weird problems when trying to stack star fields with many, many stars.
The problem disappeared, when I deselected the flag "Align RGB channels in final image" (or similar … I just see a german version …

in the settings for stacking.
The reason in that case was obvious: if you have too many stars and the channels are a little bit displaced in parts of the image the fitting algorithm might run towards into the wrong direction and into a local minimum. Your picture does not look like to crowed with stars … but maybe it is worth a try.

Meanwhile I am stacking within PixInsight which also provides a "Distortion correction" option, which I found very useful in case of distorted corners.

Christoph Nieswand:
Hi
Just an observation: I had similar and even more weird problems when trying to stack star fields with many, many stars.
The problem disappeared, when I deselected the flag "Align RGB channels in final image" (or similar ... I just see a german version ...) in the settings for stacking.
The reason in that case was obvious: if you have too many stars and the channels are a little bit displaced in parts of the image the fitting algorithm might run towards into the wrong direction and into a local minimum. Your picture does not look like to crowed with stars ... but maybe it is worth a try.

Meanwhile I am stacking within PixInsight which also provides a "Distortion correction" option, which I found very useful in case of distorted corners.

Hello Christoph,

How are you using distortion correction in PI? I have been looking at that but can't figure out how can I make it work.

Thanks,
André

In DSS, be sure to choose "bicubic" alignment when stacking, because there is some kind of distortion present for sure, so "bilinear" will not be adequate. Also, designate one of your best subs as a reference frame, and align ALL of the stacks to that reference, regardless of color. The reference frame should be one that has a high score after a preliminary stacking. With these settings, DSS should minimize the distortion problem to the best of its ability.

In DSS, be sure to choose "bicubic" alignment when stacking, because there is some kind of distortion present for sure, so "bilinear" will not be adequate. Also, designate one of your best subs as a reference frame, and align ALL of the stacks to that reference, regardless of color. The reference frame should be one that has a high score after a preliminary stacking. With these settings, DSS should minimize the distortion problem to the best of its ability.

*** Great thanks, I used your methods, looks much better now, see C20 North America and Pelican Nebula (NGC 7000)-SHO ( Adel Kildeev ) - AstroBin

Sure looks like the channels are aligned in the revision. Now the next step is to figure out what optical problems you might have, as software will not always bail you out when there is an optical aberration.

In your revision, the corner stars are all pulled to the edges. You may want to try adjusting the back-focus, or distance between your camera sensor and the telescope optics. The sensor is either too far away, or it's too close to the optics. You should get some 1.0 mm spacer rings and get to work fine-tuning the spacing. Getting the back-focus exactly right, and eliminating sensor tilt at the same time, is probably my least favorite job in this hobby.

Thanks, I will play with the spacers

Adel Kildeev:
Thanks, I will play with the spacers

I agree with @rhedden. But I would recommend getting also 0.5 mm spacers. In my flattener made difference and don't forget that if you have filters, you should increase the backfocus about 2/3 of the thickness of the filter (as a rule of thumb).

Cheers.

Andre Vilhena:
Adel Kildeev:
Thanks, I will play with the spacers

I agree with @rhedden. But I would recommend getting also 0.5 mm spacers. In my flattener made difference and don't forget that if you have filters, you should increase the backfocus about 2/3 of the thickness of the filter (as a rule of thumb).

Cheers.

I am experimenting tonight: 1. Decreased backfocus- bad-bad-bad, abandon idea. 2.Increased backfocus by 3 mm-good-good, never had such flat field

Sure looks like the channels are aligned in the revision. Now the next step is to figure out what optical problems you might have, as software will not always bail you out when there is an optical aberration.

In your revision, the corner stars are all pulled to the edges. You may want to try adjusting the back-focus, or distance between your camera sensor and the telescope optics. The sensor is either too far away, or it's too close to the optics. You should get some 1.0 mm spacer rings and get to work fine-tuning the spacing. Getting the back-focus exactly right, and eliminating sensor tilt at the same time, is probably my least favorite job in this hobby.

*** I adjusted the focuser with the Allen's key to be sure that the focuser is not the problem (by the way Meade APO 6000 has a very nice focuser) and experimented with decreasing the back-focus first by 5-6 mm (no way, a really bad idea), then I increased the back-focus by 3-4 mm accordingly. I have never had such flat field before

Some small curvature is still visible in one corner but it is MUCH better now. Great thanks, Rhedden, for your advises, clear skies, my friend! ** See example C27 Crescent Nebula-Ha-testing new spacer ( Adel Kildeev ) - AstroBin

That's great news, Adel! The small curvature in one corner means there is probably a little bit of tilt of your camera sensor with respect to the optics. Often, this problem can be addressed by inserting a thin piece of foil or tape (shim) near the face of your camera to tip it slightly, and moving the shim until you find the right corner. Eliminating tilt is much harder than getting back-focus correct, so good luck if you try it.

Adel Kildeev:
Andre Vilhena:
Adel Kildeev:
Thanks, I will play with the spacers

I agree with @rhedden. But I would recommend getting also 0.5 mm spacers. In my flattener made difference and don't forget that if you have filters, you should increase the backfocus about 2/3 of the thickness of the filter (as a rule of thumb).

Cheers.

I am experimenting tonight: 1. Decreased backfocus- bad-bad-bad, abandon idea. 2.Increased backfocus by 3 mm-good-good, never had such flat field

That's really nice to hear!

All the best and clear skies.

Andre Vilhena:
Christoph Nieswand:
Hi
Just an observation: I had similar and even more weird problems when trying to stack star fields with many, many stars.
The problem disappeared, when I deselected the flag "Align RGB channels in final image" (or similar ... I just see a german version ...) in the settings for stacking.
The reason in that case was obvious: if you have too many stars and the channels are a little bit displaced in parts of the image the fitting algorithm might run towards into the wrong direction and into a local minimum. Your picture does not look like to crowed with stars ... but maybe it is worth a try.

Meanwhile I am stacking within PixInsight which also provides a "Distortion correction" option, which I found very useful in case of distorted corners.

Hello Christoph,

How are you using distortion correction in PI? I have been looking at that but can't figure out how can I make it work.

Thanks,
André

Hi André
I was talking about WeightedBatchPreprocessing Script in PI. In the Lights Tab go to Image Registration and there you find a flag "Distortion correction". It really helped a lot when I stacked 500 untracked subs imaged with a 24mmm lens. It takes considerably longer ... these 500 subs (full frame) took about 12 hours to stack on my computer which has quite some power ... but the result was worth it.
Regards

Chris

I have the same issue with my subs. One corner Only!!!! It's not just that all kinds of things show up here and there. Skywatcher Evostar 72ed, Orion flattener for short fast refractors and last but not least, Canon 60da.

field curvature

I have the same issue with my subs. One corner Only!!!! It's not just that all kinds of things show up here and there. Skywatcher Evostar 72ed, Orion flattener for short fast refractors and last but not least, Canon 60da.

field curvature

*** Interesting-the same Orion flattener- may be this is the flattener’s fault? As I see, you have both right sideswith curvature. Did you try different spacers?

I continue to play with the spacers, added two thin plastic spacers (1mm total)- looks better now, curvature is almost gone

https://www.astrobin.com/833lur/

Adel Kildeev:
I continue to play with the spacers, added two thin plastic spacers (1mm total)- looks better now, curvature is almost gone https://www.astrobin.com/833lur/

Adel,

It is really great - congratulations! I am truly happy with that progress as I believe that solving issues at the source is better than trying to do it in processing. And nice photo, by the way. It was my first photo ever (an it was lousy). This year I tried again (https://astrob.in/93bsj3/C/) and if you look, field curvature is not better that yours.

Cheers,
André

Andre, great thanks, your photo is also a very nice one, I could barely see curvature in the corners

Friends, please remind me if stars gets elongated in all picture in the same direction-is it flexure of the focuser or overcorrected curvature? If I remember correctly, the elongated stars from the center of the picture to the corners means too short backfocus.

Adel Kildeev:
Friends, please remind me if stars gets elongated in all picture in the same direction-is it flexure of the focuser or overcorrected curvature? If I remember correctly, the elongated stars from the center of the picture to the corners means too short backfocus.

In my understanding, elongated stars in the same direction all over the photo is tracking related - may be flexure in the guidescope system or poor tracking by someother reason, such as bad polar alignment.
Too short backfocus causes elongated stars in a radial pattern and too far makes them in a circular pattern - but the center is usually good - have a look at the optcorp link i mentioned earlier.

Cheers,
André

Andre Vilhena:
Adel Kildeev:
Friends, please remind me if stars gets elongated in all picture in the same direction-is it flexure of the focuser or overcorrected curvature? If I remember correctly, the elongated stars from the center of the picture to the corners means too short backfocus.

In my understanding, elongated stars in the same direction all over the photo is tracking related - may be flexure in the guidescope system or poor tracking by someother reason, such as bad polar alignment.
Too short backfocus causes elongated stars in a radial pattern and too far makes them in a circular pattern - but the center is usually good - have a look at the optcorp link i mentioned earlier.

Cheers,
André

*** Great thanks! I was listening the Devil's whispering to continue to increase back-focus to reduce curvature, so I made a radical change: instead adding thin spacers (it helped a bit) I added 10 mm extra length of back-focus, in total 66 mm. I expected horrible distorted stars but suprisely the image was fine.

The Orion flattener's manual states 55 mm required back-focus. But reading small print to found out interesting note:
“X”mm of additional spacing may be required to achieve 55mm, depending on make and model of camera." What does it mean? I have no idea.

I think that with 66 mm back-focus the image is pretty flat, all stars are round and sharp. Pixel scale has changed from 1.622 to 1.616 arcsec/pixel and focuser 's settings accordingly. See: C27 Crescent Nebula-Ha-testing new spacers ( Adel Kildeev ) - AstroBin ***

Related discussions