Description: Added 2 more nights, total time 19.5 hours. I expect I'll learn more about processing dust like this but think I did pretty good for a first time. I really like the contrast, think it has a reasonably good background, and the Iris pops very nicely. Stars are bloated, but that's more about image collection I think. I was not exceptionally selective, am learning about differential flexture, and expect I'll be back again for a better go at it.. maybe next year!
Uploaded: ...
Description: Been trying to process Seahorse Nebula, B150, and decided to come back to Iris and see if I could learn something more about dust on an object where I know there's something decent to work with. Oh boy is it cool! This is a "learn as you go" image and there are problems with noise management. I'll give it another go sometime but don't feel like going back as far as it would require right now.
Uploaded: ...
Description: Messing around with processing. This go is pretty good but there's some speckeling in the background. I tracked it down to the luminance, so something with denoise before LRGBC.
Uploaded: ...
Description: Playing around more with the dust. Better management of the noise this time but there's some speckling that seems to be introduced by the L in LRGBC.
Uploaded: ...
Description: Ok, lots of learning this morning. Much better result. Shows I have some DBE issues on the right side of the frame, which is a dark edge I think introduced by my field flattener. The dust is much cleaner in this image and the stars much crisper.
Uploaded: ...
Description: Improvement in background at cost of star edges. The learning continues!
Uploaded: ...
Description: Bit better processing of the dust and stars, not quite as much pull on the dust though. But it overall looks better I think.
Uploaded: ...
Description: Dialing in on dust for this shot. Didn't do much with the Iris itself. Next go I'll deconvolve it a bit. Shooting OSC I don't keep any processed masters around, something I'll have to reconsider as it's a bit of a pain to get the star mask right to manage ringing for this object.
Uploaded: ...
Description: Inspired by the processing from yesterday I came up with something completely new. It probably could be refined to use different scaling functions, but basically I extracted residual using layers of 4, 5, 6, 7, and 8 with ATrousWaveletTransformation, passed it through PixelMath to rescale, then added the 5 resulting images together in PixelMath with rescale. Do another LRGBC but with a new inverted L mask then it's denoise (which was a bit special), sharpen, manipulate colors, and done. I think I'll write a script to do the bulk of it since it's very repeatable.
Uploaded: ...
Description: Last version was a good idea but bad implementation. Pulled back and using Gaussian scaling functions with different kernel sizes, not modifying the residual layers. This keeps it at the large vs small structure space while still allowing some contrast enhancement in recombining the various residual layers. The result is a tighter cloud. Also took a good bit of care not to blow out stars or introduce artifacts such as can be added with starnet.
Uploaded: ...
Description: Wrote a script to automate extraction of layers and recombination to cleanly (and without effort) enhance the dust. I'm pretty happy with it. This version also is coming back to the roots which was dark but not crazy. I feel this is the right balance of contrast and brightness enhancement since revision B. Sorry for all the in-between but it's a journey! Maybe this'll be the last revision. Right...
Uploaded: ...
Description: Saturated stars in this version and was more careful with noise reduction. Used Affinity Photo for a final slight sharpening with high pass (Iris only) and clarity.
Uploaded: ...
Description: Been spending more time processing dusty things and figured out a better way to bring up the dust. Not exactly what I want but progress!
Uploaded: ...
Description: Figuring out dust processing better. I think this one is a bit unbalanced and I toned down the right side a bit. Need to see if there's something wonky with background extraction maybe.
Uploaded: ...
Comments