Jellyfish Nebula IC443

User images created with StarTools.
Startrek
Posts: 351
Joined: Mon Dec 30, 2019 3:49 am

Re: Jellyfish Nebula IC443

Post by Startrek »

Stefan,
I missed this one
Beautiful image with so much fine detail
Colour is tremendous
Obviously you loaded your OSC Dualband Data set through Compose ?
What Luminance colour did you use in Compose ?
What Color Style did you use and did you use a preset in the matrix with tweaks in bias and saturation ?
I’m having difficulty in producing those rich brown / tan colours in the nebulosity similar to images processed in PI. ( Not that I want to copy or follow PI ) Mine end up light tan even when I bump up saturation.
I know all objects emit different levels in the Color spectrum but lately for emission nebula using Dualband I’m stuck on a similar result each time for each object.
Any tricks up your sleeve ?
Sorry to be asking too many questions
In regard to focusers and coma correctors , I used the Baader Mk3 for many years but it just didn’t eliminate coma in the corners even at f6 , also tilt was a bit of an issue with such a short barrel into my GSO dual speed focusers.
One of our local supplier / telescope manufacturers recommended I buy the TS Optics GPU coma corrector. It has a 109mm long barrel with a 25mm brass sleeve specifically designed for focusers like GSO which allows it to clamp up tight and square using those horrible thumb screws. First time I used it , Wow , tight stars , no coma and no tilt.
They’re not cheap but they do the job , an excellent optical accessory

Clear Skies
Martin
Stefan B
Posts: 422
Joined: Sat Oct 31, 2020 8:59 pm

Re: Jellyfish Nebula IC443

Post by Stefan B »

Hi Martin,

thanks for your feedback! :mrgreen:

You're right, I loaded the duo NB data set into R,G and B in Compose and used the "L + Synthetic L from R(2xG)B, R(GB)(GB) (Bi-Color from OSC/DSLR)" compositing scheme. The style/matrix was H(HO)O and I think I used "Artistic, detail aware" for Style. As already said I decreased saturation from the default 200%. If I changed the bias slider from the default I can't remember, could look it up in the log, just let me know. I brought back the saturation in the SuperStructure module. This has a huge impact in my eyes. I didn't use the default value there but reduced the saturation a bit. But still it increased the saturation in the bigger structures (as intended).

Do you use PS? There you could do some saturation tweaks in the Lab color space. I did that for the first image in this thread. These were subtle alterations. The second rendition with 1.9 was ST only, no PS involved. So you see it basically works without PS ;-)

I agree on the GPU/coma corrector stuff. Someday I will upgrade to an GPU CC and hopefully the star shapes will get better! :-)

Thanks again
Stefan

PS. Would it be helpful for you if I upload the stack and corresponding log? So you could exactly see how I ended up with the color where I did?
Startrek
Posts: 351
Joined: Mon Dec 30, 2019 3:49 am

Re: Jellyfish Nebula IC443

Post by Startrek »

Hi Stefan,
Thanks for the feedback, pretty much how I’ve approached OSC Dualband processing
No need to upload your log files , thanks for the offer
One of our local experienced forum members kindly reprocessed my recent Running Chicken Nebula in Startools and finished the Color in PS. I’ve never used PS
I suspect its some kind of false Hubble Pallete as the result was quite spectacular but he removed the stars which I’m not a fan of.
I don’t think Startools has the capability to produce this kind of Color palette ?? Or maybe with true Narrowband using a Mono camera where Ha , Oiii and Sii signal is far greater than with OSC

Your thoughts and comment on the Startools/ PS version

Cheers
Martin

Sorry for deviating from your original posted image
Attachments
B8E5C052-B67E-4528-A5CA-197711E067FA.jpeg
B8E5C052-B67E-4528-A5CA-197711E067FA.jpeg (159.62 KiB) Viewed 2112 times
Mike in Rancho
Posts: 1142
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: Jellyfish Nebula IC443

Post by Mike in Rancho »

Interesting, Martin.

I agree it is a striking and spectacular appearance, except for the starless which in most cases I find problematic.

A few issues are raised I think for potential discussion --

One indeed is the choice of palette, and I believe it has been discussed before at various times. ST has several presets, quite a few in the case of tricolor NB, all probably based on colorwheel-opposites theory. But you still don't really have full continuous hue control for choosing the base color mapping of bicolor or tricolor. I believe it was mentioned as a potential future upgrade to have that capability?

A second is just what changes occurred here using PS. Since we don't know what was done, or how potentially selectively manipulative it was, we can't quite evaluate the integrity of the changes. However, also as discussed in the past, once the color mapping has been set (meaning the channels have been combined/changed), further changes to channels/colors would result in what I think Ivo called "scrambled eggs." At the very minimum for having the two or three NB hues actually still be true to their emissions origin.

Finally, there was a question and a bit of theorizing I posted a couple weeks back, but I think it got lost, regarding something potentially similar but related to "SHO/HOO Normalization" as popular pixel math in PI. As I understand it, basically the weaker emissions channels are normalized in the same way you would for stacking, referenced to the dominant channel, and this results in overall stronger colors than we get in ST with our relative channel balancing or throttling. There may be potential pitfalls - raising of noise requiring extra alterations being one, and processing after star removal being another - since the weak channel stars could blow up. That aside, I wondered if some of such differences are a result of other NB balancing being able to raise the weak channels luminance, as opposed to only relatively balancing just in the chrominance. And thus, what is the "correct" way to balance narrowband channels of very differing strengths - which they commonly are.

Having say, both Ha and OIII raised to simultaneously show equivalently (or nearly so) strongly does look great, but is it a reasonable representation or too artsy? :confusion-shrug:

I suppose a post-Color SS-Saturate of large structures could head somewhat in this direction, but may still be limited by the luminance having been set in Compose with weak non-dominant channels when creating the Synth L.

:think:
dx_ron
Posts: 246
Joined: Fri Apr 16, 2021 3:55 pm

Re: Jellyfish Nebula IC443

Post by dx_ron »

Your rendition is interesting enough that I went back to my own Jellyfishie and tried H(HO)O mapping. I didn't get nearly as rich a coloring as you did. For my "normal" rendition I added OSC stars, but I think that would look weird with this color palette. Bottom line, though, is still that I need more time on this target.
Jellyfish_AlpT_only_H-HO-O.jpg
Jellyfish_AlpT_only_H-HO-O.jpg (691.05 KiB) Viewed 2069 times
Stefan B
Posts: 422
Joined: Sat Oct 31, 2020 8:59 pm

Re: Jellyfish Nebula IC443

Post by Stefan B »

Hi Martin,

yes, looks really interesting. Since I don't know what has been done in PS, I can't tell if it's still documentary. Even if I would know, probably my expertise wouldn't be enough. But I guess you can't get there with ST alone. Perhaps Ivo can tell more on this.

But I think you can get a more saturated look of your images within ST. Maybe you want to share a duo NB dataset and I would try.

Same for @dx_ron: Being a bit heavy handed in SuperStructure with the Saturate preset as a starting point might boost the appearance quite a bit. If you want to share your stack we could have a look.

By the way, I like RGB stars in an image with the H(HO)O color palette. You need a good star mask for this and often I am too lazy to do that appropriately but the result look like this:

Image

Image

Haven't done this for quite some time since meanwhile I prefer combining duo NB and RGB via NBAccent which gives a more natural look and overall really great results.

Regards
Stefan
Startrek
Posts: 351
Joined: Mon Dec 30, 2019 3:49 am

Re: Jellyfish Nebula IC443

Post by Startrek »

Hi Stefan,
It’s a hot topic ( false or manipulated HB color palette ) at the moment in our local forum. Everyone is wanting to learn how to use PS selective Color
Here’s my image reprocessed in PS now with stars included
Looking at it again it’s way over saturated, even the member who processed it agrees, but that golden colour against the blue with dark nebulosity in between provides a real contrast. It just needs to toned way down
I’m going to reprocess my image again in ST and try a few different things. I found Entropy works ok for OSC Dual band data to provide some great contrast and colour tone
I assume Ivo jumps on line from time to time and I know his thoughts on False or Manipulative Color which has no scientific value at all.
An interesting topic to say the least
We really should start an independent post and thread
Cheers
Martin
Attachments
9B7A6E72-092B-46D4-A8C7-026EA0CEC93F.jpeg
9B7A6E72-092B-46D4-A8C7-026EA0CEC93F.jpeg (183.62 KiB) Viewed 2041 times
Stefan B
Posts: 422
Joined: Sat Oct 31, 2020 8:59 pm

Re: Jellyfish Nebula IC443

Post by Stefan B »

Hi Martin,

I get these issues being controversely discussed. I guess what should guide us (or what guides me) is the question: where is the gas, in case of duo NB obviously Ha and OIII. In my eyes you can pretty well locate the two in your PS rendition. If you render your Ha red, brown or orange doesn't matter to me as long as I know what it is. And as long as color wasn't introduced in a painting by the numbers style with arbitrary selective masking or the like.

I know that Ivo has a much more stringent and technical attitude to this, also concerning non linear stretching of single colors.

If being very true to the data (like ST usually does I think) and knowing lots of stuff about your equipment (sensor QE for every wavelength etc.) you might do some analysis on where is more of Ha and OIII etc. (Whatever more means...more photons of a certain wavelengths depending on the wavelength specific sensor QE or molecules of a certain gas or molecules in particular volume etc.) But for me these images aren't scientific data. I have an idea of what that means since I've been working in academia and am currently working in industrial science. But since basically nobody knows the wavelength specific QE of the sensor etc with which a particular image was made when looking at a random image, it's fine with me when someone takes some freedom in processing NB data and shifting colors/hues etc.

I have a more stringent view on broadband data since there is some kind of a true color to it, i.e. star clusters of hot blue stars shouldn't be rendered red, even if this would look more dramatic. This would be misleading. And the viewer would get a wrong idea about the nature of these stars. But with NB data there are more degrees of freedom I guess.

Don't know if I was able to make myself clear...as already noted: Sometimes not so easy for a non native speaker ;)

Regards
Stefan
Mike in Rancho
Posts: 1142
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: Jellyfish Nebula IC443

Post by Mike in Rancho »

Nice image, Martin. It's very appealing and shows color differentiation quite well. The question is, of course, what was done to achieve that and did it selectively manipulate in an improper way. :think:

Stefan I think the only freedom with NB is that you can set your own hues for each NB emissions. Thus, NB is inherently always false color, but that does not mean faux or fake color. If that makes sense?

There are quite a few out there, particularly in the "other software" camps, who want to claim that NB being false color opens the door to a free-for-all. I don't find that to be a logical conclusion.

There are multiple sub-issues being covered here, so I don't know if it needs a new thread, or perhaps several new threads, for deeper discussion? Maybe. I did start one as I mentioned above on the "normalizing" of the emissions channels: viewtopic.php?f=4&t=2792

That's a popular method in PI right now. With caveats, it can also result in boldness and strong differentiation, easily perceivable, of the different filters. I don't know if the PS method is doing something similar, channel-wise, or is more of a selective alteration after-the-fact.

Saturation level itself doesn't bother me, for BB or NB, as I find that a matter of imager's choice, no different really from amount of stretch. Avoid bleeding and artifacting, of course. But there's nothing terribly natural about long exposure astrophotography, and what the human eye could see live through an eyepiece shouldn't be relevant. Unless you are intentionally trying to replicate what a visual observer would see, like for writing an S&T article I suppose.

But I can understand saturation levels very much ends up a matter of subjective preferences, and so may have no right or wrong. :confusion-shrug:
User avatar
admin
Site Admin
Posts: 3367
Joined: Thu Dec 02, 2010 10:51 pm
Location: Melbourne
Contact:

Re: Jellyfish Nebula IC443

Post by admin »

Startrek wrote: Tue Feb 28, 2023 12:36 pm Hi Stefan,
It’s a hot topic ( false or manipulated HB color palette ) at the moment in our local forum. Everyone is wanting to learn how to use PS selective Color
Here’s my image reprocessed in PS now with stars included
Looking at it again it’s way over saturated, even the member who processed it agrees, but that golden colour against the blue with dark nebulosity in between provides a real contrast. It just needs to toned way down
I’m going to reprocess my image again in ST and try a few different things. I found Entropy works ok for OSC Dual band data to provide some great contrast and colour tone
I assume Ivo jumps on line from time to time and I know his thoughts on False or Manipulative Color which has no scientific value at all.
An interesting topic to say the least
We really should start an independent post and thread
Cheers
Martin

From the horse's mouth. Yes, I called you a horse Adobe - deal with it! :P
"For example, you can use selective color correction to dramatically decrease the cyan in the green component of an image while leaving the cyan in the blue component unaltered."
and
"The adjustment is based on how close a color is to one of the options in the Colors menu. For example, 50% magenta is midway between white and pure magenta and receives a proportionate mix of corrections defined for the two colors."
That is all, of course, problematic. Suppose we have a straight SHO:RGB mapping. Now consider our image has;
  • an O-III (blue) dominant area with a smattering of Ha signal (e.g. blue, but somewhat cyan)
  • a Ha (green) dominant area with a smattering of O-III signal (e.g. green, but somewhat cyan)
Then this implies that this tool "dramatically decreases" the O-III in the Ha-dominant area, while leaving the O-III in the O-III dominant area intact.

Of course, emissions and chemical signatures don't magically appear or disappear, depending on what (and how much) other emissions or chemical signatures are present in an area. It is meddling with hues for no good reason. :( Besides all this, I have a feeling that, in order to apply this tool, you would need to do this on the non-linear signal to begin with, compounding the nonsensical nature of it all...

Pretty? Aesthetics are of course debatable. Reflecting reality? Definitely not. Of course, it may be that I am misunderstanding the procedure being applied.

The Color module should give you a lot of leeway however to achieve some rather more dramatic renditions if you so please, without compromising on the accurate image-wide depiction of relative emissions strengths.

@Mike in Rancho My apologies for completely missing your "Narrowband balancing, throttling, boosting" post. :oops: Not an excuse, but I've been a bit under the pump lately (no end in sight just yet). I will read and respond now.
Ivo Jager
StarTools creator and astronomy enthusiast
Post Reply