3 images in progress

[Stefan B]

Hi Ron,

thanks for the interesting quotes! Will have to read into that.

It also makes for a nice close up as I recently saw: https://www.astrobin.com/h6trai/

Regards
Stefan

[dx_ron]

Learned something that maybe should have been more apparent to me. The comment above (intended to be about the odd orange color of the Garnet Star) that "the other stars are blue" pointed to an artifact of cranking up the Layer mask fuzz to better blend in the Garnet Star. That was causing loss of color diversity in the small OSC stars, but I was too focused on the one problem star. So I went back and used Layer twice - first for most of the stars with the Garnet Star excluded and very little mask fuzz, then with only the Garnet Star.

Elephant_trunk_AlpT+OSC_gaudy.jpg (456.41 KiB) Viewed 19022 times

There's some problem stars when you look close, caused by trying the "bloat the OSC stars" strategy, so I need to go back and touch up my OSC processing. But hopefully you get the point.

[decay]

Hi Ron,

so different masks for different star sizes? That sounds reasonable. Until now I didn’t try to layer RGB stars into my duoband images. Most times the halos haven't been too annoying (probably because of saturated stars ) and other times I just desaturated the stars using the Color module and a star mask. Or I used the blue/brown-orange mapping.

But I like your result, so I think I will try that too. And then I will remember this thread

Thanks for sharing!

Best regards, Dietmar.

BTW: have you decided on the “final” presentation of your bubble? ST only processing or pre-stretching with Siril? Using the dark side of the Force? Yepp, seems to be a matter of conscience somehow …

[Mike in Rancho]

Yes much better color and spectrum variability to the RGB stars now than the last one, though the base image is different too.

I've run across similar matters in the past, though at the moment I can't remember what procedures I was trying to pull off. But yeah both two-step Layer due to some big-big stars that had to be handled separately, and indeed being mindful of mask fuzz. Too much can snuff out a lot of whatever you are trying to blend in. If something seems to require extra fuzz anyway, possibly a mask grow could help push out the boundaries some.

[dx_ron]

Hi Ron,

so different masks for different star sizes? That sounds reasonable. Until now I didn’t try to layer RGB stars into my duoband images. Most times the halos haven't been too annoying (probably because of saturated stars ) and other times I just desaturated the stars using the Color module and a star mask. Or I used the blue/brown-orange mapping.

But I like your result, so I think I will try that too. And then I will remember this thread

Thanks for sharing!

Best regards, Dietmar.

BTW: have you decided on the “final” presentation of your bubble? ST only processing or pre-stretching with Siril? Using the dark side of the Force? Yepp, seems to be a matter of conscience somehow …

Thanks! I think the layers of Layer is mostly for images with particular problem stars. Growing the mask is worth trying. I don't know if things fuzz in both directions. I admit I had just kind of imagined mask fuzz working mostly out from the mask, but it obviously works in from the unmasked regions. Does it go both ways, though?

btw - both Bubbles are pure StarTools. They only differ in the Strength slider in the NBAccent module. There's a pure-Siril Crescent rendition that I'm likely to use as my final image for that target.

[dx_ron]

So more tweaking. One early-stage tweak was to bump up the Red and down the Blue contributions to the synth-L in Compose. I re-processed the OSC stars image to be normal, rather than way over-cooked. The Layers tweak was to actually use it 3 times. First for most of the star colors. That mask excluded the Garnet Star, and I used Sensitivity 0 to catch as many small stars as possible and Threshold somewhere between 50-70% (forget exactly) to catch as much of the full profile as possible of the larger stars. Then tweaked some of the individual star masks as needed (and I forgot to save the tweaked version of the mask - oh well). Cleared the mask, then used Lasso to make an intentionally irregular shape around the Garnet Star and all that 'halo' (it's clearly not traditional filter halo - not sure why it has a diamond shape, but maybe it's something to do with the overall nebulosity there). Anyway, I first loaded the OSC stars file back into Layers and used Lightness of fg with something like 50px of mask fuzz. Saved and finally went back again for Color of fg with the same giant mask fuzz.

ET_identity_500sat_triple-layered_1600_500k.jpg (482.46 KiB) Viewed 18985 times

Then I went to gimp to try to replicate Freddie's nice sky blue shades - without much success. This is from just bumping the Cyan saturation, I think:

ET_identity_500sat_triple-layered-gimp_1600_500k.jpg (496.7 KiB) Viewed 18985 times

[Mike in Rancho]

Cool. I think the first new one (pre Gimp) is fine, unsure if the cyan really needs to be brighter.

If a lot of custom work goes into a mask and I forget to save it out as a jpg, there's always the the converter if you cut n paste from the log. Sometimes the png can be big but can always be resaved from mask as a jpg.

https://www.startools.org/modules/mask/ ... artoolslog

Looking at my ET from a couple months back, while the FOV is much wider, all your Oiii seems to be in the right place. There's just only so much there.

viewtopic.php?p=14750#p14750