I have posted a series of threads on my efforts to improve my flat frames to eliminate artifacts in my imaging. This data is mostly in the following post:
https://www.cloudyni...ring-artifacts/
Most of this thread discussed edge of filter light leakage issue with the ZWO 36mm filters and all the work that was put into resolving that issue. I then purchased a set of 36mm Astrodon filters and did a comparison to the ZWO filters and that data was covered in this post.
https://www.cloudyni...ontrolled-test/
The conclusion of these two posts was that by either masking the edges of the ZWO filters or going to much better Astrodon filters that are fully coated to the edges you can eliminate the serious edge of image "vignetting-like" patterns in flats that is uncorrectable with out serious post processing of data. By masking the ZWO filter or using full-coated, better quality (and more expensive) Astrodon filters, I now can get FLATs that are good enough so that the artifacts can be removed in post processing.
In other words, although the flats still had ring patterns, I have no uncorrectable issue with these flats calibrating my light frames. With the ZWO filters with all the edge-of-filter masking or with the AD filters that were fully coated, my calibrated light frames show no edge color rings that I could not correct "easily" with Pixinsight. Are they perfect? Well DynamicBackgroundExtraction (DBE in Pixinsight) is a wonderful tool. To my eye, I had good results with my ZWO filters and my AD filters - they were good enough. What remained in the flats were very weak (~1-2% intensity variation) concentric ring patters only visible with an aggressive screen stretch. Without stretching the flat the flat was -- well -- FLAT. You cant seen anything in my current flats without strongly stretching them.
However, I wanted to see if I could fully eliminate them. I was convinced that the ring patterns - as weak as they were - were still the effect of the remaining reflections in my optical path. Flocking my focuser draw tube was my first attempt to determine where they were coming from and whether I could fully eliminate them. The results of this work was shown in this most recent post:
https://www.cloudyni...cuser-drawtube/
Did I need to do this - probably not. Good enough is good enough. However, that said, the previous reflection and filter issues show that reflections are non-linear effects. And even if weak - too weak to really notice - they are still there. I don't think it will hurt anything to fix them. The less I have of these artifacts, the less I have to do with DBE. At least that is the theory. This post will show the further step I used to eliminated them. I applogize up front for the long post and the number of breaks. There is a lot of data here so bare with me.
Even after flocking my focuser draw tube, reflections in the optical train of my system were clearly visible: This was through the OTA, focal reducer, filter wheel and LUM filter with the flocked focuser