Hi, Ivo
I'm currently wrestling with a Crescent Nebula shot, and minimizing all those stars after they have been stretched into visibility is difficult.
It occurs to me that rather than stretch the star field into full visibility and then deal with them downstream, a better approach might be to stretch them less in the first place. And yes, I did use an ROI on the crescent, and it did reduce stars a bit, but not nearly enough.
How about enabling masks in Dev/Autodev?
Like this...
Autodev->bin->wipe->Autodev
Generate star mask
Autodev or Dev again, minimal re-stretch on stars in mask
Invert Mask
Autodev or Dev again, ROI re-stretch (over a masked area) on nebulosity.
Thoughts?
- Bob
Dev/Autodev mask?
Re: Dev/Autodev mask?
Hi Bob,
Though you need to be quite careful and probably won't like the results very much (stretching the stars differently to the rest of the image can look odd or create discontinuities and artefacts), you can effectively use masks with the output from any module. After stretching the image, launch the Layer module, Click Undo > Foreground, which will put the image-before-it-was-processed-by-the-module into the foreground layer. Then use the Blend Amount to blend the two images to your liking - you can launch the Mask editor from the Layer module to generate or fix up your mask (e.g. generate a star mask). In this particular case, using a suitable Max Fuzz parameter, combined with a specific brightness mask ("Where composite is light & dark, use bg") may help somewhat.
There are only two generally accepted ways of Star prevalence; by means of multi-scale structure decomposition (as used by Life w/Isolate preset) or by morphological filter (Shrink).
If this doesn't help, would you be able to give an example and/or perhaps upload the dataset?
Though you need to be quite careful and probably won't like the results very much (stretching the stars differently to the rest of the image can look odd or create discontinuities and artefacts), you can effectively use masks with the output from any module. After stretching the image, launch the Layer module, Click Undo > Foreground, which will put the image-before-it-was-processed-by-the-module into the foreground layer. Then use the Blend Amount to blend the two images to your liking - you can launch the Mask editor from the Layer module to generate or fix up your mask (e.g. generate a star mask). In this particular case, using a suitable Max Fuzz parameter, combined with a specific brightness mask ("Where composite is light & dark, use bg") may help somewhat.
There are only two generally accepted ways of Star prevalence; by means of multi-scale structure decomposition (as used by Life w/Isolate preset) or by morphological filter (Shrink).
If this doesn't help, would you be able to give an example and/or perhaps upload the dataset?
Ivo Jager
StarTools creator and astronomy enthusiast
StarTools creator and astronomy enthusiast
-
devonshire
- Posts: 60
- Joined: Mon Sep 24, 2018 12:30 am
Re: Dev/Autodev mask?
Thanks, Ivo!
I'll give this a shot and see what happens.
This is just a case of nebulosity getting overwhelmed by a very busy star field. I did try standalone starnet++, but complete removal looks odd and leaves artifacts behind (infill). I also tried shrink/tighten, but those also introduced a bit of mottle where the infill went. In a perfect world, I'd like to end up with fewer, fainter stars accompanying the nebulosity.
Then the light went on that the only reason had all those troublesome stars...was because I stretched them into existence. If the nebulosity and starfield could be stretched separately, I expect that would do what I'm looking for.
Anyway... thanks again - back after I've tried what you suggested.
I'll give this a shot and see what happens.
This is just a case of nebulosity getting overwhelmed by a very busy star field. I did try standalone starnet++, but complete removal looks odd and leaves artifacts behind (infill). I also tried shrink/tighten, but those also introduced a bit of mottle where the infill went. In a perfect world, I'd like to end up with fewer, fainter stars accompanying the nebulosity.
Then the light went on that the only reason had all those troublesome stars...was because I stretched them into existence. If the nebulosity and starfield could be stretched separately, I expect that would do what I'm looking for.
Anyway... thanks again - back after I've tried what you suggested.