Artifacts on lunar image processed by AutoStakkert

[CaptainKingsmill]

I sometimes get artefacts on lunar images that I stack in AutoStakkert, they are usually manageable, so I play around with the settings and sometimes I can make them go away.
However my process is quite unscientific and I've not managed to nail down what it is that causes it. As a result I sometimes end up scraping the data and recapture is when I next get some clear skies.  

This latest one is a total mess, the worst I've ever seen, with similar settings and processes that I usually use. 
50% of 800 frames. 
The below is an example, one with AP size of 24 and one of 200. a noticeable difference, but a bad one.... 

The left if the AP200 the right is AP24. and below is a single frame, just for reference. 

Any points, thoughts, or directions would be very much appreciated. 

 

Original frame

Edited January 18, 2022 by CaptainKingsmill
sentences.

[raadoo]

Can you post a screenshot of your AS! interface, please? There may be settings other than APs that are causing this.

[CaptainKingsmill]

Certainly can Sir.

this was the settings from last night

This morning I have also tried setting it to surface out of curiosity. I've tried AP 312, Best 10% of frames. 
All with similar results. 

 

[rob_r]

I'm not that experienced with AS!3 but it just looks like way too many alignment points. Taking nearly 7 minutes to stack a little over 800 frames seems excessive to me. Nothing in the screenshots is jumping out at me as incorrect though. Puzzled. Have you tried checking 'manual draw' and place say a couple of dozen APs then rerun the stacking process?

[ONIKKINEN]

I would at first glance agree with the above, that there are too many alignment points. But if the same happens with an AP size of 312 this is really weird.

This is what my settings look like and no such issues happen. Image Stabilization set to surface as that is meant for lunar recordings and the improved tracking selected. Dont know what the improved tracking thing does, but cant hurt right?

[raadoo]

I echo what the others have said. Nothing stands out as strange, other than the far too many APs (I usually shoot for <500).

I'd also recommend going for a lower percentage of frames to stack. Looking at the quality graph, I'd try something more in the line of 15%-20%.

[neil phillips]

Small AP box can produce box artifacts but never seen them to this level. I find lots of APs not a problem on my data. Others may vary of course. BTW planet is ticked here. You should be on surface. 

I suspect the camera is different to a dedicated cmoss camera. Not sure of course. What camera was this again ? 

[CaptainKingsmill]

Hello all,

Thanks for your help in getting to the bottom of this. - I would usually have discarded the data by now, but I'm determined to find out what goes on that causes it.
I'll try and reply to all your points.
 

4 hours ago, rob_r said:

 'manual draw' and place say a couple of dozen APs then rerun the stacking process?

I have tried this just, but got totally garbled data, so I don't think I added enough - I'll try this again tomorrow. 

 

4 hours ago, raadoo said:

I'd also recommend going for a lower percentage of frames to stack. Looking at the quality graph, I'd try something more in the line of 15%-20%.

I also have a 10% from the same data set which is below. same issue. 

 

48 minutes ago, neil phillips said:

BTW planet is ticked here. You should be on surface. 

I did actually run both just to see if it made much difference, just uploaded the above screenshot

 

49 minutes ago, neil phillips said:

I suspect the camera is different to a dedicated cmoss camera. Not sure of course. What camera was this again ? 

This is an unmodified canon 2000D, which has worked fine before, such as for my profile picture.

I initially put all the frames into PIPP to crop them to the right size, however I have also tried directly using the frames I captured into AS! to rule out some funny business in PIPP - Same result.

I'm going to have another play around tomorrow morning, trying some more of your suggestions, I can't see how it could be the data itself but i'm running out of other options.
 

[neil phillips]

3 hours ago, CaptainKingsmill said:

Hello all,

Thanks for your help in getting to the bottom of this. - I would usually have discarded the data by now, but I'm determined to find out what goes on that causes it.
I'll try and reply to all your points.
 

I have tried this just, but got totally garbled data, so I don't think I added enough - I'll try this again tomorrow. 

 

I also have a 10% from the same data set which is below. same issue. 

 

I did actually run both just to see if it made much difference, just uploaded the above screenshot

 

This is an unmodified canon 2000D, which has worked fine before, such as for my profile picture.

I initially put all the frames into PIPP to crop them to the right size, however I have also tried directly using the frames I captured into AS! to rule out some funny business in PIPP - Same result.

I'm going to have another play around tomorrow morning, trying some more of your suggestions, I can't see how it could be the data itself but i'm running out of other options.
 

I am thinking some kind of settings on the camera is not agreeing with AS/2 i cant be sure. But it kind of looks like some kind light settings you captured with the Canon,  are not working with the AP mapping on AS/2 ive never seen it this bad. Is the reason i am suspecting this. Let us know if you trace the problem. Ive not used that type of camera on the moon. So hard to be certain of course. Good luck

[raadoo]

Shot in the dark, but can you share your data here so we can try it ourselves and see if we get the same result?

[CaptainKingsmill]

Apologies for the delay, Work has been quite busy this week. 
So I've had a couple more attempts, and I found something interesting.
For speed I started processing just half the images.... and as if by magic...

I've not done much processing on it, but there are no artefacts at all. 
I genuinely didn't expect that halving the data simply to speed up processing would make any difference. 
It's then that I took a deeper look at the data and noticed a cast change as shown below,  possibly a condensated secondary?, as I didn't notice any mist, fog, or high level cloud. I'm really not sure
The detail in them looks about that same as does the focus, it just appears to be a change in colour, is something like this going to have an effect on the stacking process?  (we wont talk about how lazy my tracking is if I'm shooting lunar)

 

 

[Tommohawk]

First, well done for sticking with it - there will be a solution.... but I'm not sure I have it!

You've tried al the things I would have, but one thing I notice is the position of the discs in those last two images is quite different - maybe beyond the scope of AS to align?

When you ran it through PIPP, did you get a cropped result with the disc central in each frame?