"Falling into a rut..."
05/17/2024. Sometimes "adopting a consistent workflow" amounts to "falling into a rut." I was not happy with the sharpness of some recent solar images. I noted (for the tenth time) that it's probably better under my usual seeing to increase gain and decrease exposure time. I boosted gain to bring exposures down to around 3-4ms from 7-8. The immediate result was a somewhat lusterless raw file but an improved processed image. Coping with added noise and reduced dynamic range beats working to make the most of smeared details. (This wisdom of this strategy will depend somewhat on the nature of the day's seeing: slow and large, fast and large, fast and small, perfect...)
Then I reviewed my post-processing routines (usually AutoStakkart>ImPPG>Photoshop but sometimes with a preliminary noise-reduction/cropping step in Photoshop). Today, starting with the raw capture (via FireCapture, 16-bit, SER format, unity gain or close to it, 1000 frames, gamma disabled), I tried new and novel methods in post.
AutoStakkart4! output up top / Pixinsight refined image below
Messing with ImPPG's adaptive USM options (along with the usual 50-70 passes of L-R decon, ringing suppressed) produced some improvement. Then I tried Photoshop's SmartSharpen and skipped ImPPG entirely (450%, 10 pixels, "remove lens blur") and got even better results (at least this time). Photoshop's Image / Adjustments / "Shadows/Highlights" was useful, too (play with that one; I'm not entirely clear about what all its sliders actually do). Substituting Pixinsight's excellent and highly adjustable deconvolution routines for ImPPG and Photoshop might have produced the best result yet (see image above). I suspect ImPPG is as capable as any of these workflows, but care must be taken because excesses lie but a tiny slider nudge away.
Yes, as a matter of fact, while I've been discarding 99% of my SER capture files (they're huge), I have kept a ton of Tiffs straight out of Autostakkart!3 and 4 with which to experiment with post processing. Expect to be subjected to further verbosity as various treatments prove out.
Just for the record, some of these adjustments and considerations follow from another reading of this page:
https://solarnutcase.livejournal.com/18572.html
And yes, I do see that in addition to bringing up solar details, I am picking up Newton's rings on the upper left side. Those should be easy to fix by taking a screwdriver to the tilt-adapter ring.
5/20/2024: A few more using Smart Sharpen and Unsharp Masking without an ImPPG step. In fairness, the seeing was especially steady.
:: top ::