M31 OSC + NB accents

[dx_ron]

My first time using the new NB accents. Shot 5 hours of OSC at my club's Bortle 4 site (high end of 4), then added 3 hours at home (high B6) with the Antlia AlpT filter. Preprocessed and registered to a common frame in Siril.

AT65EDQ, Risingcam IMX571C, CEM40, KStars/Ekos/PHD2

I would appreciate critiques on color and style. Too gaudy? Too blue? Not enough blue? I hunted for scientific measures of the colors of M31 - found one paper showing 80% of the disk is... white. So, no, 'scientific' coloration is right out.

M31_OSC+NB_v1_500k.jpg (489.7 KiB) Viewed 1737 times

[Mike in Rancho]

Hard to say. As you mention, there are so many versions of M31 out there, all differently balanced. There might be more blue than usual here? I'm trying to think if I've seen as many blue stars across the face of M31 like that.

Generally, absent some good reason not to, I think I've been getting good balance with the star sampling followed by Max RGB for any green evaluation. That seems work well unless something has gone awry with the acquisition or stack. Then just a matter of the saturations, which for M31 can be hard (I'll often make bright sat pretty low due to the core), especially depending on what happened in HDR.

I'm not sure I would buy that 80% of the galactic disk is "white." In what context? It ought to be an array of colors just like the MW is. Though perhaps due to distance and collimation of all of those photons the central disk might average out as white?

[fmeireso]

It looks fairly good to me. As Mike said, maybe just a tad of too many blue stars, but otherwise a nice picture of the galaxie.

I would push the HDR more, to get more detailed structures in the core,but that is personal. I tend to do this with everything, but i guess not everbody likes that.

[hixx]

It seems to me the stars only suffer from too much blue. Mostly there seeems to be missing yellows. (Did You suse a CLS / UHC / LPS filter to shoot this?. This is a great Luminance stack, but M31 could use more blue in the outer arms, where you O / B stars are located.You will not see this if you didn't shoot an RGB stack with Clear or UV/IR only filter, though.
Shooting with Light pollution Filter not only cuts the blues in spiral arms (and in refection nebula as well, BTW, but mainly it cuts yellows, so You won`t see a yelllow center and the color Module will try to "balance" this by incorrectly retain the blue resulting in those blue star fringes.

To bring out the spiral arm blues, just add a color stack in Compose R, G and B channels and Your existing stack as LUM (all need to be registered the same, though) - use the better site on countermoon to shoot this- you will not need much exposure time. Then add the Antila stack as NBA to boost H-a regions.
Color module may leave more blue then, as there will also be "balancing" Yellow (results in less blue reduction) to bring out the outer arms. You can mitigate the star blues if needed by:
1) In Color, try the Highlight repair, set it to some 5 pixels or so
2) Using the Shrink module's Tighten preset (maybe dial in some more color Taming. Experiment with the iterations to throttle the overall effect.
3) Push back stars a bit using Super Structure / DimSmall or Isolate

clear skies,
Jochen

[dx_ron]

Thanks for the comments. I definitely noticed the over-representation of blue local stars when I set the color balance for M31 to something that I find "pleasing". Any effort so far to shift toward red makes the galaxy kind of flatten out - though that is probably scientifically accurate as well.

Hard to say. As you mention, there are so many versions of M31 out there, all differently balanced. There might be more blue than usual here? I'm trying to think if I've seen as many blue stars across the face of M31 like that.

Generally, absent some good reason not to, I think I've been getting good balance with the star sampling followed by Max RGB for any green evaluation. That seems work well unless something has gone awry with the acquisition or stack. Then just a matter of the saturations, which for M31 can be hard (I'll often make bright sat pretty low due to the core), especially depending on what happened in HDR.

I'm not sure I would buy that 80% of the galactic disk is "white." In what context? It ought to be an array of colors just like the MW is. Though perhaps due to distance and collimation of all of those photons the central disk might average out as white?

A star mask (even while masking out the galaxies) give a very green starting point.

starmask-to-full.jpg (443.63 KiB) Viewed 1666 times

Yes, I am referring to the fact that we cannot resolve the individual stars in M31, and the averaged light from much of the galaxy is basically white. I recall even seeing suggestions (maybe for other software) to use a galaxy central region as the "white" reference.

by fmeireso » Sun Sep 04, 2022 5:07 am
It looks fairly good to me. As Mike said, maybe just a tad of too many blue stars, but otherwise a nice picture of the galaxie.

I would push the HDR more, to get more detailed structures in the core,but that is personal. I tend to do this with everything, but i guess not everbody likes that.

I did a couple of things differently in the version below (and above). First, I pushed the gamma parameter in Autodev to 1.4 right at the start, mostly because I want to have a version more suitable for showing non-astrophotographer friends and family and also potentially for printing. Brightening up the whole image gives the galaxy a bit more 'pop'. Second, I did try pushing HDR harder. The first version used the Optimize preset, so the core was left fairly bright. This time I combined the settings for Optimize with those from the Reveal preset. It's OK, though I think the outer dust lanes have actually been pushed so far down as to nearly disappear. I'll probably go back to just Optimize settings.

Jochen:
The main stack is OSC with only a uv/ir-cut filter.
I am reluctant to start using Highlight Repair, but I will play with it a bit. The 'final' images shown are all after SS Isolate. I am one of those weird people who is OK with fairly dense and prominent star fields (to a point).

So - back to color adjustment from the very green starting point above, clicking on green areas leaves pink

starmask-to-click-green.jpg (487.61 KiB) Viewed 1666 times

It becomes a matter of how far to throttle the red channel. If I pay most attention to the local foreground stars, I get to here:

starmask-to-click-green-less-red.jpg (497.33 KiB) Viewed 1666 times

That could be viewed as "realistic", though not as satisfying as the blue-stars version.

[alacant]

Hi
Nice shot.
Went through wipe, color, fringe killer, flux and autodev to keep more faint stuff. Working with the original stack, this should be really good.
I'm not sure about the purple stuff in the galaxy arms. Otherwise, I think as you've seen, with the proliferation of false colour stuff, these days, anything goes!
Cheers and HTH,
Steve

[Mike in Rancho]

Hi
Nice shot.
Went through wipe, color, fringe killer, flux and autodev to keep more faint stuff. Working with the original stack, this should be really good.
I'm not sure about the purple stuff in the galaxy arms. Otherwise, I think as you've seen, with the proliferation of false colour stuff, these days, anything goes!
Cheers and HTH,
Steve

Well, anyone can do as they please, of course, but I don't think I would agree with the "anything goes" philosophy. Though sure, we are all going to personally draw different lines in the sand for the transition from photography to art. And some mine not draw any line at all!

Two or three channel narrowband (in which I include duoband OSC) is an entirely legitimate way to present an astro subject. Sure, it is a matter of taste in picking the hues you want to assign to those emission bands. Those hues may or may not end up in a similar RGB region that they would as per their "natural" place in the spectrum. They also may or may not be opposed sufficiently (i.e. color wheel) in order to visually differentiate the emissions, and we can also globally throttle dominant emissions in order to relatively reveal the others. All of that seems reasonable, particularly if explained to the viewer, and would not be any kind of "free for all."

Unfortunately, it seems that YouTube, the internet, and even CN, are replete now with techniques and pixel math that may make pretty pictures, but do not have a supportable logical explanation behind them. That's just art and graphic painting, more or less based on astro acquisition. And if art is your end goal, then yes, I suppose anything goes.

To me, broadband is even more constrained, has nothing to do with how false color narrowband may be processed, and we should endeavor get a white balance that is appropriate. That may be easier said than done, and I think has been one of dx Ron's discussion points here. We may need one of Ivo's "color" tutorials?

Saturation I think is a matter of, again, both taste and desire to reveal what can be revealed, tempered by watching out for bleeding-over areas if it is pumped too high.

To me (and maybe just me?), saturation of chrominance is logically equivalent to stretching of the luminance detail. I do find it humorous that some will stretch the heck out of an image (overdone for the data IMHO), but then only lightly saturate and tell us all how "natural" that is.

Anyway...was that a rant? Do I have to put "close rant" now?

[dx_ron]

Hi
Nice shot.
Went through wipe, color, fringe killer, flux and autodev to keep more faint stuff. Working with the original stack, this should be really good.
I'm not sure about the purple stuff in the galaxy arms. Otherwise, I think as you've seen, with the proliferation of false colour stuff, these days, anything goes!
Cheers and HTH,
Steve

Hi Steve,
Thanks for the comments. (was my original Wipe so poorly done that it was improved upon by running the highly compressed jpg through Wipe again? )

With many broadband targets (nearby galaxies, globular clusters and suchlike) there should be an objectively correct color balance. Whether a color spectrum we get from scientific instruments can be matched after filtering through our telescope + camera response is debatable, but we can try.

Sometimes, though, the "correct" colors are also a bit boring - which I think is the case with M31. One of my problems is that, while not color-blind, in real life I am pretty much "color-agnostic". So my wife chooses color schemes around the house and helps me pick out clothes for important occasions!

I am setting M31 aside for a bit, waiting to see if I get a chance to get back down to the club site to add more data before the end of the month before I decide on a final version.

I should go find that Wishlist entry for the ability to keep the current color channel weights in place when you re-enter the color module and add my vote. It would be nice to be able to decide something is "maybe just a bit too blue" and go back to tweak it without having to start over with sampling.

[alacant]

Hi everyone
What we see, is an interesting one. Colour in particular.

Arheim's masterful work 'Art and visual perception' (alianza 1954, ISBN 978-8420678740) deals in depth with how we see stuff and how our mind deals with what we see. It's almost as heavy as trying to process an image in pi. By way of st example, I'm sure Ivo won't mind reference to his equally masterful 'un-glow' algorithm which uses the idea of actual and perceived, very well.

Cheers and HTH

rhav.jpg (20.39 KiB) Viewed 1517 times

**EDIT:
Apologies to @dx-ron. In no way did I intend to suggest your 'wipe' (or anything else) was bad or wrong. It was pure luck that reversing the stretch on a .jpg worked as well as it did. Again, all credit to ivo and his coding. Zero to myself!
Cheers