Integration causing nasty rings

@mackiedlm Can you post a screenshot of an unstretched flat including the histogram? In APP load a flat and then double click it. Then make sure that the stretch checkbox at the upper right is unchecked and then take a screenshot.

@wvreeven, attached as requested, with and without background neutralised. Now I do see some rings in the flat - which are likely the source - but not sure why they are there or what to do when capturing my flats to stop them - I use CCD Flats Tool in APP and target about 50% ADU for the flats.

Also confused as to why its not an issuee in pixinsight.

thanks for your help.

@mackiedlm How do you take the flats? What light source do you use?

I normally use a LED tracing panel with several layers of white t shirt. But wondering if this was the cause I took some with just the white T and using the dawnsky for illumination and got the same result.

@wvreeven @vincent-mod

Does anyoneone have any input in relation to this problem - if its an issue with my flats (as per those I posted, on request, above) can you suggest what I need to do to capture flats that will not cause this problem. If its in the APP settings what do i need to do.

As already mentioned this issue does not manifest when using Pixinsight on exactly the same data set but I am reluctant to move to PI for all my preprocessing as i like the Extraction algorithms in APP. But if I cant resolve this then I will need to go to PI fully.

Thanks for any help.

David.

@mackiedlm David, sorry for the late reply. Do you see the rings as well if you do an integration without using the flats? Perhaps you can upload a few lights, flats and other calibration files to a directory called mackiedlm_rings on

https://upload.astropixelprocessor.com/

using upload5 for both username and password? Then we can have a look and see if we can find out what's causing this. Thanks.

Thanks for getting back Wouter.

I had actually just run a short integration, 5 subs, without flats this morning and it does not appear that the rings are showing when the flats are not used. But it still confuses me as to why there is no issue when integrating in PI!

So I have uploaded 10 each of Lights, Flats, Dark Flats and the Master Dark for this set-up. Also included the MF and MDF which came out of my process and the first integration.

Thanks for your help - I really want to work this out.

@mackiedlm David, I did a quick integration in APP and indeed the rings show up. Looking at the flats they already show colored rings which seem to be cause of the rings in the final stack. You can see that by loading a Flat in APP and displaying it. Then enabling neutralize-BG and saturation, choosing 30% BG and setting the saturation to the max will bring out the color aspects of the rings:

Not sure why they don't show up in PI because I cannot do it in PI since there are no darks between the files you uploaded and I cannot use the master dark of APP in PI.If you want I can have a look with PI as well. My suspicion is that the background level in the PI integration is higher than in the APP integration but I can only be sure of that once I get the darks.

Thanks Wouter -  darks iploaded now.

OK, so what do you think could be causing these coloured rings - I'm back to thinking its the LED tracing pad but I've never had issue with it on other images - Then again, this was the first time I used exactly this set up which is using a different FF/FR and giving 1976 FL instead of the usual 1200 with this scope and the old reducer..

Is there a way it can be mitigated in APP?

@wvreeven Any thoughts on this - clear tonight so I'll be capturing data and new flats so would like to be able to get it right.

Thanks

David

@mackiedlm I am discussing this internally with Vincent and Mabula and this may take a few days more.

Ok thanks.

@mackiedlm 

Thanks for posting this topic as I'm having very similar issues.

I've only been using APP for a few months, but this only started happening very recently and only under certain circumstances.

I'm seeing occasional large rings and even the dust motes from my flats don't appear to be fully removed. It's almost as if they are not correctly registered and everything is just a few pixels off, causing this ring effect... sometimes in several places.

It shows up more when I use broadband filters and I don't notice it at all on my narrowband shots.

I'm also using an ASI2600 (mono version though) and also using one of those LED tracing pads + t-shirt for shooting flats. I've experienced it through my FLT132 but have not tried this camera/filter combo with any other scope yet. I also have an ASI2600MC, but have also not tried it with the setup that is exhibiting these issues.

I keep thinking that it has something to do with a specific setting that I'm using. I keep re-shooting flats but nothing is working. Going to try stacking in another software for comparison.

Hi David @mackiedlm, @wvreeven, @drano

I will have a look as well, to see if I can understand what is happening here 😉

Will report back later once finished...

Mabula

After reviewing my earlier post and finding another topic elsewhere, I think I'm having two separate issues. One has to do with flats not properly calibrating my image and the other one is closer to this issue that David has raised. At first I wondered if it was some form of optical issue being caused by the IR-cut protective window on the 2600 series cameras. It mostly appears for me when using luminance or green filters with my 2600MM and with very bright objects. For me it was the core of the Iris nebula and the rings seem to be centered around the core and radiate outwards from there. 

Posted by: @mabula-admin

Hi David @mackiedlm, @wvreeven, @drano

I will have a look as well, to see if I can understand what is happening here 😉

Will report back later once finished...

Mabula

Thanks Mabula, I appreciate your help.

I'd ask if we could focus primarily on the issue I raised rather than being diverted by the extra issues being brought in by @drano. My problem is not with mono, filters, bright objects, dust motes etc. The rings are not centred on an object in the image but seem, as shown by Wouter, to be related to the flat frames.

My issue is ASI2600mc, coloured rings in a stack produced by APP,  but not when the same data is stacked in PI.

Thanks again

David

Hi David @mackiedlm & @drano & @wvreeven & @vincent-mod,

Good news!

I have located the source problem that causes the rings and it explains 100% why Pixinsight did not create the rings. And there is a simple solution to get it right in Astro Pixel Processor.

The issue is that Astro Pixel Processor by default creates 16bit masters and Pixinsight creates 32bits masters. Your calibration masters (the MasterDark & MasterDarkFlat) really suffer when you create 16bits masters. Severe posterization is the result... this is due to the sensor characteristics clearly. So you really need to create 32bits masters in APP by setting:

2) Calibrate, enable create 32bits masters

Then all should go well and give a proper result without rings. I checked myself and the result is now good 🙂 !

Compare the histograms of the 16bits MasterDark:

to the histogram of the 32bits MasterDark:

You can even see the difference visually.

This is the first time for me, that I clearly see that 32bits masters are needed over 16bits masters. So I have changed APP's current code to make 32bits Masters by default going forward 😉

Thank you very much for sharing this issue and please let us know if all is okay now.

Cheers,

Mabula

Posted by: @mabula-admin

Hi David @mackiedlm & @drano & @wvreeven & @vincent-mod,

Good news!

I have located the source problem that causes the rings and it explains 100% why Pixinsight did not create the rings. And there is a simple solution to get it right in Astro Pixel Processor.

The issue is that Astro Pixel Processor by default creates 16bit masters and Pixinsight creates 32bits masters. Your calibration masters (the MasterDark & MasterDarkFlat) really suffer when you create 16bits masters. Severe posterization is the result... this is due to the sensor characteristics clearly. So you really need to create 32bits masters in APP by setting:

2) Calibrate, enable create 32bits masters

Then all should go well and give a proper result without rings. I checked myself and the result is now good 🙂 !

Compare the histograms of the 16bits MasterDark:

to the histogram of the 32bits MasterDark:

You can even see the difference visually.

This is the first time for me, that I clearly see that 32bits masters are needed over 16bits masters. So I have changed APP's current code to make 32bits Masters by default going forward 😉

Thank you very much for sharing this issue and please let us know if all is okay now.

Cheers,

Mabula

That is just brilliant Mabula - thanks so much. Now you have explained it I remember that I have seen that type of histogram but was not smart enough to understand what it was saying.

I'll go and run a new integration now and report back.

Thanks again.

Hi David @mackiedlm & @drano & @wvreeven & @vincent-mod,

This is what i get with the 32bits masters, the integration looks very good now. The rings are completely gone.

(Result does look a bit better then the Pixinsight WBPP result as well if you ask me 🙂 ).

Cheers,

Mabula

@mabula-admin - that worked perfectly.

Output from 16 bit masters - normal and strong stretch

32 bit masters - normal and strong stretch

The rings have totally gone. Still a nasty gradient but thats not APP's issue and it is easily fixed

Thank you for your input - this is indeed great news.

Rgds

David