I was lucky enough to get a string of SIX clear new-moon nights in late July to revisit one of my earliest DSO targets - the Cygnus Loop!  The first time I imaged this was in 2020 using a SkyGuider Pro star tracker, stock 70D DSLR, 300mm lens, and L-Enhance filter.  At that time, it was one the longest project I'd attempted (5 nights, ~220 subs @ 240 sec).  The result was OK (my limited processing skills back then didn't help), but I was not happy with my effort.  It was not long after that session that I decided to invest in my current setup, but I had not returned to image this target since my equipment upgrades (and my move from APT to N.I.N.A. for acquisition).

Total exposure time per panel was not equal because I suspected poor seeing a couple nights and tried to supplement with more time, but in the end I used all frames (the two right hand panels have ~5hr/filter/panel while the left hand is ~3.5hr/filter/panel).  Mosaic processing was done in Pixinsight (MosaicByCoordinates + PhotometricMosaic scripts).  I think the seams and gradients were much better than my last mosaic project using GradientMergeMosaic.

I tried to use Deconvolution on the stitched linear image (both with and without stars) but the many bright filaments kept producing ringing artifacts, so I decided to skip it altogether.  Not sure if that is a common issue on this target or not?  After playing with a couple different approaches (ChannelCombination before vs after stretching, with and without star-removal, with and without Local HistrogramEqualization) I settled on a "simple" approach.  Combined channels in linear state after some background noise reduction and using a 20/80 blend  of Ha/Oiii for green, neutralized background and used PhotometricColorCalibration, followed by an unlinked STR stretch.  After that, just some star reduction and a few curves modifications... nothing more!