true colour in globular clusters

Questions and answers about processing in StarTools and how to accomplish certain tasks.
Rkonrad
Posts: 247
Joined: Sun Feb 07, 2016 2:55 am

true colour in globular clusters

Post by Rkonrad »

Hi all

Nice to return to the forum after a long time. I've recently imaged 47 Tucanae - globular cluster and found that scientific colour was not able to give the stars the correct colours. In a case where there isn't an evenly balanced variety of stars, how would I achieve the proper white balance? Here's the image I processed (very lightly) and I'm sure globular clusters aren't pink. Cheers!

Richard

https://www.flickr.com/gp/rkonrad/o6wpTgNza9
fmeireso
Posts: 365
Joined: Mon Sep 28, 2020 8:46 pm
Location: Belgium

Re: true colour in globular clusters

Post by fmeireso »

That is a very weird result, i must admit. I imaged a couple of Messier globular clusters and i never had an issue with color getting out of Startools.


Did you use any filters perhaps. What cam did you use and scope?
Rkonrad
Posts: 247
Joined: Sun Feb 07, 2016 2:55 am

Re: true colour in globular clusters

Post by Rkonrad »

I had the exact same problem with m13 some time ago. The scope is a 12.5 RC with RGB filters - taken in Australia under nice dark skies. My guess is that the field of view does not contain the wide variety of stars needed for a proper white balance. I can forward the data to anyone who would like it though the processing was quick and simple. Stacked with APP with the settings made optimal for startools. Thanks.
Mike in Rancho
Posts: 1120
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: true colour in globular clusters

Post by Mike in Rancho »

Hi Richard!

Interesting. :think: I actually wouldn't mind looking at the original stacks and the capture particulars if you want to provide it, though I probably can't take a peek until tonight after work. Others might be able to give a try earlier.

In the normal course of things, my globs tend to include a wide perimeter. Thus, I simply create an auto star mask, then touch it up (using drag circle) to remove the glob stars from the mask, since they are not representative of "average." Sampling that then creates a good starting point for channel balance. Here, the entire square (533MM?) is the glob though, so that trick won't work.

I did download a jpg from Flickr, but (have to check with Ivo?) that may already be too altered to give me a reasonable platform for figuring things out. That said, I opened and chose reverse stretch even though I don't think it still looked the same, and could not get a very good balance. So I drew a small mask on just a no-star piece of background and sampled that. It neutralized the background (Max RGB showed a lot of white) but the stars were still way off. I also did an Undo and used Wipe, figuring that might eliminate any casts. But really no matter what, getting the image anywhere close to "normal globular" colors required a large degree of increased green (or reduce red+blue) plus green cap to (mostly) yellow. Along with a lot of highlight repair as I saw some dispersion effects.

Weird. Again maybe the fact that this is a finished and balanced image prevents re-balancing, in the same way we are told not to pre-white-balance our stacking?

Do you have any other broadband images taken with this exact setup of scope, camera, and filters? If that balanced out properly, you might check the log for the color channel bias settings in Color used on that other image and then duplicate it. :confusion-shrug:

Compared to OSC, when I use my mono cam with filters and of course after Compose, Wipe, Color sampling often results in quite minimal channel bias numbers.

Three registered stacks were created by APP, correct, not pre-combined into a single RGB file?

If the original stacks don't balance, we might have to start digging into the filter pass graphs maybe. :?
Rkonrad
Posts: 247
Joined: Sun Feb 07, 2016 2:55 am

Re: true colour in globular clusters

Post by Rkonrad »

You did a lot of work Mike - thanks! I've used this scope a lot as it is one of the "free" scopes at Itelescope. Here are few other rgb images I took with the same setup. https://www.flickr.com/gp/rkonrad/06RT3ZPee4 (2 pictures). Did you want to see all the light frames or just the stacked one? The light frames are calibrated by Itelescope.
Do you have any other broadband images taken with this exact setup of scope, camera, and filters? If that balanced out properly, you might check the log for the color channel bias settings in Color used on that other image and then duplicate it. :confusion-shrug:
Thats a good idea - never thought of this.

Thanks again

Richard
Rkonrad
Posts: 247
Joined: Sun Feb 07, 2016 2:55 am

Re: true colour in globular clusters

Post by Rkonrad »

So I tried copying the information from the log of prior rgb images, applied and got this https://www.flickr.com/gp/rkonrad/74Pw0t51je.
Now too blue. I would love to hear Ivo's take on this problem if he has time.

Thanks! Richard
dx_ron
Posts: 227
Joined: Fri Apr 16, 2021 3:55 pm

Re: true colour in globular clusters

Post by dx_ron »

Nice thing about the well-studied globs is that they are, well, well-studied. There is an visible HR diagram at https://lweb.cfa.harvard.edu/~mashby/jo ... /47Tuc.pdf
Mike in Rancho
Posts: 1120
Joined: Sun Jun 20, 2021 10:05 pm
Location: Alta Loma, CA

Re: true colour in globular clusters

Post by Mike in Rancho »

Well that seems a much better starting point, Richard. Anything get better if you massage it from there?

Reducing blue might also reduce some of the magenta-like appearance of the redder stars. :confusion-shrug:

Here's an ESA-Hubble rendition, though the FOV is so tight I suppose it's not very helpful. And good grief this thing is way down there. https://esahubble.org/images/heic1510a/

What was used for stacking and registering? I wonder if there's a slight channel misalignment. :think:

To match the ESA Hubble balance, you'd want a fair bit of white, some standard blue, and it seems almost a golden-red. Query if doable.

The thing with globs is to try to avoid them being too much of a candy blue and yellow, like a bowl of m&m's or skittles. :lol:
Rkonrad
Posts: 247
Joined: Sun Feb 07, 2016 2:55 am

Re: true colour in globular clusters

Post by Rkonrad »

Sorry, I've been away for a week or so and won't be back for several days. I will post a Dropbox of my stacks and I'll also restack them with a different programme though my m13 looks the same (used with a different scope and osc ). Looks like Ivo may be away or busy but I'd love his view on this as well.

Richard
Last edited by Rkonrad on Sat Nov 25, 2023 6:35 pm, edited 1 time in total.
Rkonrad
Posts: 247
Joined: Sun Feb 07, 2016 2:55 am

Re: true colour in globular clusters

Post by Rkonrad »

Here are my stacks for Tucanae 47.

https://drive.google.com/drive/folders/ ... drive_link

Thanks for all your help!
Post Reply