Globular Cluster colors . . .

[Terry Danks]

I see many globular cluster images that seem to have nothing but bluish stars.
These are ancient assemblages of mainly red dwarf stars though, with relatively few blue stragglers here and there.
It seems to me they should be more a ruddy red colour rather than predominantly blue.

To apply colour correction, I have done background neutralisation in the normal fashion first and then colour correction by defining a preview that includes the entire cluster . . . much like a spiral galaxy. This seems incorrect though as globulars should not integrate to white.

The G2V method seems beyond me in that identifying a G2V star requires access to some really detailed charts.

So, my globulars look blue too.

Any suggestions on getting colours of these clusters more resembling their true physical colours?

[RickS]

You could try using eXcalibrator (http://bf-astro.com/excalibrator/excalibrator.htm) to calculate a manual white balance.

[aworonow]

Man, you've struck a nerve! I totally agree that almost all GC images are way too blue. Any HR diagram tells you these things largely have old, red stars. So what to do. One thing I have had some success with is selecting just the core of the GC and letting PI do it's color balance with that "preview." Did you try disabling the "structure detection," which apparently presume the average color of the stars is white? Finally, I used the saturation tool to bring up the red channel.

Of course, this does not precisely calibrate the colors, but at least they are not way too blue!

Alex

[Terry Danks]

Thanks for the link to excalibrator, Rick. I am trying to figure it out now.

Alex: I am glad I am not entirely alone. This has bothered me for a while. Many people, perversely IMO, actually seem to positively PREFER their globulars be blue!
Adjusting the red channel skews the neutral background and that bothers me as much as the blue stars.
I will try your suggestion of selecting only the core in the white balance preview for the CC tool. And deselecting structure detection as well.

The goal for globular processing, as I see it anyway, should be a neutral background, an overwhelmingly dull ruddy hue for the cluster with any blue stragglers actually present, showing their true colours as well,

[vicent_peris]

Hi,

Please take a look at the table in page 22 of this paper: https://www.noao.edu/ggclib/ms.pdf.

This paper synthesizes the spectra of entire globular cluster with the goal of recognizing them around other galaxies, where you only see fuzzy points. If you take a look at the B-V color index, they are pretty blue! Note that the B-V index of the Sun is 0.65 magnitudes. In that table, only two globular clusters have a higher (redder) color index than 0.65. And many of them are close to the color index of Vega (0).

So, according to this paper, those bluish colors are absolutely correct.

Best regards,
Vicent.

[Terry Danks]

Hi Vicent:
Lengthy post here, perhaps not appropriate other than my globulars are still blue in PI and I don't think they should be.
Many thanks for your reply. Sorry to be so late in getting to it.
I note the table you refer to on page 22 of the paper by Schiavon, et. al.
However, column 8 in that paper refers not to colour index (B-V), as you state, but rather to colour excess or reddening E(B-V). While (B-V) is indeed a measure of intrinsic colour, E(B-V) is a measure of reddening due to gas and dust along the line of sight. So stating the integrated excess of a cluster says little about it's integrated colour, only how much redder it appears than it really should.

An article on globulars in the Encyclopaedia of Astronomy and Astrophysics shows a colour image of Omega Cen. on p.1. http://www.astro.caltech.edu/~george/ay20/eaa-globcl.pdf. That image comes close to my admittedly preconceived notion as to what a colour image of a globular cluster should actually look like.
An APOD of this same cluster again shows a lot more red and very little blue . . .
https://apod.nasa.gov/apod/image/1604/OmegaCen_Colombari_1833.jpg

Now, as you point out, the colour index of the sun is 0.65. The colour index of the vast majority of individual stars in most globular clusters lies to the red of that figure. Here is a CMD for omega Cen . . .
https://www.google.ca/search?sa=G&hl=en-CA&q=color+magnitude+diagram&tbm=isch&imgil=zY83tBUJnXZzRM%253A%253B3xOOnjjC7ZyF0M%253Bhttp%25253A%25252F%25252Fspiff.rit.edu%25252Fclasses%25252Fphys301%25252Flectures%25252Fmw_size%25252Fmw_size.html&source=iu&pf=m&tbs=simg:CAES0gEJBwad--BpUaQaxgELEKjU2AQaAggKDAsQsIynCBpiCmAIAxIopAieCJ0IogiEA58Iowi_1CNwTiAPZKYA43z3UNMM92z3cPdgp4z20NBow1Adgw0BZ-OU3Q359edTtg47waVu0TSqh1ac-da9ex02xtOn9sj1KaiRtSzmT1sYsIAQMCxCOrv4IGgoKCAgBEgTU1ur8DAsQne3BCRo0ChgKBHBsb3TapYj2AwwKCi9tLzA0cTM0N3kKGAoGbnVtYmVy2qWI9gMKCggvbS8wNWZ3Ygw&fir=zY83tBUJnXZzRM%253A%252C3xOOnjjC7ZyF0M%252C_&usg=__9woYQ1Xwm28uBXIpaBrXrbNImD4%3D&biw=948&bih=457&ved=0ahUKEwjY14TK4o7UAhXkJ8AKHVg6DxgQyjcITQ&ei=8MEoWdifA-TPgAbY9LzAAQ#imgrc=OlOwwaLq9WzcFM:

Good grief! That is a horribly long URL!  I hope it comes through properly.

Blue stars, blue stragglers and the stars of the HB are neither especially bright, as compared to the far more numerous red giants, nor very numerous. Their contribution can only contribute a rather small amount to a globular's integrated colour. It's difficult to take a CMD of a globular and have its predominant star colour be bluish.

Now all my globulars appear predominantly bluish. an example is my recently processed M5 here: http://danks.netfirms.com/m5.htm. I've done nothing to "make" them look like this. it's how they come out after a background neutralization and then a colour correction using "target image" in the colour calibration tool with structure detection on.
While a "pretty picture," I do not feel it conforms to reality and I remain perplexed as to why.

[vicent_peris]

Hi Terry,

Sorry, I wrongly read the table and yes, that's not B-V color.

Please could you upload your globular cluster image? I'm out of home and don't have access to any of my globular cluster photos. I would like to try it with the new PhotometricColorCalibration tool. I can test the image using a spiral galaxy as white reference.


Thanks.
Best regards,
Vicent.

[Terry Danks]

Hi Vicent:

I'd be glad to upload my recent images of GCs.  But to where? What file size and format would you like?

I have since communicated with a prominent GC researcher on this very matter and I quote him here. I do not feel it appropriate to identify him in this forum as I have not permission to do so but will communicate his identity to you privately, should you desire it. Here is what he had to say when I expressed my opinion that GCs are frequently depicted as too blue:
"I agree that many pretty pictures of GCs are rendered too blue.  they should be basically a pale yellow, with individual blue HB stars looking either white or blue, and the individual bright red giants looking orange or even in extreme cases actually red."


Perhaps it would be best that I upload calibrated, linear R, G and B masters to dropbox? I have attempted to "force" the clusters to fit my preconceived notions as to better resemble how I feel they should appear. To no satisfaction, I might add as I find myself too actually preferring the unreality of blue galactic globular clusters!  But I know it isn't "correct" that I should do so!

The three masters and a full size jpg of my combine are at this link: https://www.dropbox.com/sh/8d97lkyzm20mbqf/AACcjrVd3al2zlyBfiAagGYVa?dl=0

Whoops! but I forget you are not in a position to download large files at the moment. My finished image of M5 can be accessed at this link: http://danks.netfirms.com/m5.htm

I repeat that the stars turn blue when I use the colour calibration tool. Doesn't seem to overly matter what I choose as a white reference, blue stars are the result. For reference, the parameters for M5 are as follows, integrated (B-V)=0.72, E(B-V)=0.03 and [Fe/H]= negative 1.29dex. I do not have the HB morphology parameter at hand but at that metallicity I would expect a healthy number of  blue HB stars.

[vicent_peris]

Hi Terry,

Yes, please upload the RGB masters from this image, I want to run PCC on it.

Anyway, B-V 0.72 is not so red. That's the color index of a Sc spiral galaxy. It's a bit redder than the Sun though. It all depends on what you choose as white reference...

Best regards,
Vicent.

 

[Terry Danks]

Vicent:
The three masters are at https://www.dropbox.com/sh/8d97lkyzm20mbqf/AACcjrVd3al2zlyBfiAagGYVa?dl=0

Please tell me if you have trouble accessing them, or wish me to upload them elsewhere.
Thanks . .

Terry

[vicent_peris]

Hi Terry,

Thanks. I calibrated the color with PhotometricColorCalibration using an average spiral galaxy as white reference. Hope you'll like it:



The RGB weights were 1:0.888:0.785. Of course, there are a lot of blue stars, but also a lot of yellow ones.


Best regards,
Vicent.

[vicent_peris]

And this is the result of calibrating with a G2V star with PCC:



Obviously, the color is biased towards the red. Each one has a different color rendition. The G2V version puts the balance on a "wihte" star according to our eyes. The spiral galaxy version put the colors of those stars in context, in the sense that a spiral galaxy has a significative contribution from all the different stellar populations; this means that, placing the balance point on those stellar populations, the globular cluster is not so red. I like more the spiral galaxy version since it shows properties related to the objects and not to our eye.

Best regards,
Vicent.

[Terry Danks]

Stupendous, Vicent! Yes, this is it!

But now I have two questions.
1/ Where do I find charts that will allow me to identify a G2V star in a particular field? How did you do that! This has always led me to a dead end in G2V calibration. 
2/ How were you able to use a spiral as a white reference when there are no spirals in the image?
3/ Is this PCC a new tool not yet available to all PI users? AFAIK, my PI is current, v.01.08.04.1195(x64)

[Juan Conejero]

Hi Terry,

3/ Is this PCC a new tool not yet available to all PI users? AFAIK, my PI is current, v.01.08.04.1195(x64)

It is a new tool that will be available in PixInsight 1.8.5, to be released in a few weeks, along with many new tools and lots of improvements. See this forum thread for more information and examples.

1/ Where do I find charts that will allow me to identify a G2V star in a particular field? How did you do that! This has always led me to a dead end in G2V calibration

The PCC tool performs plate solving and photometric analysis automatically. You don't have to select stars or other objects. PCC works with all of the stars in the image, not just G2V stars. It can work with thousands of stars.

2/ How were you able to use a spiral as a white reference when there are no spirals in the image?

PCC uses the integrated color of a spiral galaxy (actually, the average of several spiral galaxy types) as its default white reference. White references have been computed by Vicent from photometric data, using a methodology that we'll describe when we release the PCC tool. You don't need to have samples of the white reference in your image, just stars included in astrometric and photometric catalogs.

[Terry Danks]

Whoa!!! This PCC tool sounds just fantastic! I will wait . . . as patiently as possible! 

Thanks Vicent and Juan very much for the information imparted in this thread.