Processing Globular Clusters

[midwayexpress]

This is 3 hours total on M92. Overall it came out pretty nice with tons of small galaxies in the background field but I was wondering if anyone had tips for processing globular clusters. Specifically I'm having a lot of trouble with cap green not fixing the core of several stars near the core of the cluster resulting in grey stars. See the 100% crop version. The very core of the object itself also wants to turn pink on me.

Full frame

http://www.astrobin.com/184216/


Here is the 100% crop version

http://www.astrobin.com/184307/

You can download the stacked file here
http://1drv.ms/1MjIpdh

[akeru]

Hi. Globular clusters are very dificult for me too. I use the following trick when I process images with bright stars. In colour and noise module, I apply an inverse mask to fat stars. It prevents green cores in fat stars and works for me.

Regards.

[admin]

Hey guys,

I'm currently unable to look at the data, but I was wondering what sort of color balancing procedure you guys employ?
The default color balancing of globular clusters may be off, as their star population is often skewed (and hence not a good white reference). Some tweaking is therefore warranted (for example you could lassoo out the glob and sample only the rest of the star field for white balance). In general "Cap Green" should only serve as a last resort to suppress green hues.

Will have a look at the data as soon as I can (next week).

[midwayexpress]

I have been manually color balancing because as you suggested it is much harder to do with a globular cluster star population. Actually since I have switched to the Atik 490 from 420 I have found that it is much harder to correctly use stars for color balance because the 490 is sooo much more sensitive to green.

I have discovered why the stars have green centers but I'm not sure how to fix it. If you look at the stack from DSS what you see is that all of the stars in the red channel have black centers. At first I thought this must be some bizarre focus issue but the light frames themselves look fine. It appears that this issue is caused by debaying but I tried both bilinear and ahd with no success.

[midwayexpress]

Ok so I partially resolved the issue. It seems that the cosmetic hot pixel remover in DeepSkyStacker was treating the core of the stars in the red channel only as hot pixels. When I turned that off it resolved the star core issue. It does create a problem where I get some hot pixel noise but I suppose the fix is to change the sigma (I use 3) that DSS is using.

That said I'm still struggling with the amount of green that I get with the Atik 490 vs. 420 and the increased difficulty of color balancing as a result. I really wish StarTools had a way of plate solving the image and then selecting white balance based on a database of stars in that image.

Here is the new fixed data to play with.
http://1drv.ms/1JwJ6zX