Staring @ the Sun, 136

:: home ::
          <<  128  129  130  131  132  133  134  135  136  137  138  139  140  141  142  143 
                Solar Resources:  SDO | Solar Monitor | Spaceweather | Real-time X-ray | NASA | H-a Monitors | NOAA | SRCH
 

Solar quilting.


06/30/2024. I sipped coffee and made some calls while waiting for the filter to warm up this morning, then snagged a couple of 2000-frame clips and a bunch of shorter ones with which to experiment. Here's a four-frame panorama of the Sun's active regions. The busy part of this frame is the best 500 frames from 2,000; the far left and right portions are the best 200 of about 1,000. The only processing was a little added contrast for each panel in Adobe Raw, Photomerge, then the Solar_Suite_1 action followed by a light USM pass and a little tweak to the histogram via curves. It's sharp enough to merit some storage. The click-take is 1,600 pixels with very slight compression (the original is 6,600).

 

4x
Oh, hell yes, you should make it big.

 

I also had Photoshop's "photomerge" do a nine-panel version which it did without drama. That worked out well enough to be encouraging. I'm working my way toward something like a 15k x 15k full-disk solar portrait at sub-arcsecond resolution. A good flat, a consistent sky, and an interesting solar "landscape" should produce a killer photo. Here's a quick look at a quick 9-panel workup:

 

9-up

 

07/01/2024. The success of "solar_suite_1" depends very strongly on the scale of the image it is fed. I should have expected that, but it was a nasty surprise. I developed it using data drizzled 3x, so must remember to scale images given to it accordingly (upsample undrizzled data by 300% for instance), or just routinely let AutoStakkart!4 do a 3x drizzle on all ASI178MM bin2 data and be done with it.

I tried an 8-panel mosaic with some success. I was dodging clouds and used only the best 20 frames from 100 frame clips, but that was enough, with some help, to produce a nice-ish full-disk. Probably not going to show that one off (a second effort worked better). I was looking for a scale that would suffice for detail and not be too onerous to stitch. This might work. The setup was the snout with internal compressor on a single 20mm extension. My flat was through a single layer of plastic of the centered, defocussed Sun.

I got a better result by processing each panel individually and manually assembling the 8-panel mosaic, but that is by no means a practical approach for more ambitious full-disk images -- instructive but a dead-end as a workflow. A couple of defocussed soft-light flats brought out prominences and evened out the inherent panel-to-panel variation.

 

8x
Clicking encouraged.

 

Working on it.

I took a couple of 1,000 frame clips using the same optical setup and used the best 250 frames for a two-panel panorama. Photomerge also failed to stitch that output (!), but a 2-panel mosaic (left/right) is no problem to sew by hand:

 

july1
Yes, you really should click it.

 

So, yeah, that's not bad as far as detail goes, but it's probably a little too tight to make an 8-panel mosaic. Twelve might work, but to keep it easy, go back to 6. Would a 12-panel mosaic be detailed enough to be worth the trouble? My first thought was not, but that image up above might make a believer out of me. It might be just about right. But before getting too excited, I have to get Photomerge (or some stitching s/w) to work with these images.

 

 

:: top ::

 

 

                   © 2024, David Cortner