Resolving the core of Omega Centauri

[Atmos]

I've always loved globular clusters, it's why I did my masters thesis on them :screwy:
This particular image has always been one of my all time favourites not only because it's the largest globular cluster but on a personal note the original data set comes in at a low ball 1.37" FWHM.

I decided to put BlurXTerminator to the test and attempt doing a deconvolution on a star cluster; something I have attempted a few times in the past and failed at miserably. Heavy deconvolution on nebulosity is easy because you carefully mask the stars but doing anything more than light and careful deconvolution on stars for me as always been a terrible idea.

Full resolution link (https://cdn.astrobin.com/images/57732/2023/18b68d90-b7e8-44c4-9e3c-2ec98775df5e.jpg)

So, to give this some context. I've compared this against the Chart-32 team, bit of a gold standard of ground based imaging. It sits right along side their rendition of Omega Cent in sub arcsecond seeing. It was all well and good that all of these smaller stars were resolving but I was very skeptical as to whether anything was being "created". To be surprise... nope. Not a star out of place between the two images both even having virtually identical absolute resolution.

What I will say is that there is still no substituting seeing and that shines from an astrometric standpoint. Although all of the stars showed up in both images, the Chart-32 image correctly displays their relative brightness differences in the smaller stars where as mine shows them all being about the same brightness with very little to no variation.

You'll like this Mike as it's also been captured at 0.4"/pixel ;) It's time to go a smaller image scale!

[glend]

Colin, 1.37" FWHM! Amazing, I can only dream. But surely not from Melbourne?

[Atmos]

From my back yard. Last night I would have comfortably been getting 1.5-1.8” in 120s exposures. That’s from checking a couple in the morning and a few when it started, who knows what happens in the 5+ hours in between :lol:

[Ryderscope]

An excellent application of new tools on some old data Colin. I like your comparison of the results helping to verify the application of new tools such as BlurXterminator.

[strongmanmike]

Bah! :scared2: :tasdevil:......:sad:

BTW is that 1.37" the best star in a sub, the best star in a stack or an average across a few stars in a sub or a stack or the final image, or something else? I'm new to Maxim and measuring FWHM's, so just interested to know how you determine your quoted FWHM's?

Mike

Great Omega core BTW ;) :thumbsup:

[Atmos]

Thanks Rodney. I remember back when the Topax module was getting flack when it started getting used for astro images as it appeared to be creating detail where it didn’t exist. Thankfully BlurXTerminator doesn’t suffer from this.



I use Maxim for approximations while I’m out imaging but when I’m doing measurements I use PixInsight I randomly select 30-50 stars and then remove any from the list that don’t have the right star shape (Moffat) or roundness. By the time I’ve done that I likely have 20 or so left over and I just average those. It gives a few bits of information but I just look at the star size in pixels and multiply it by the image scale.
The sharpest individual sub was around 1.15” from memory but the stack came out at 1.37”. The dominating aberration on-axis I found with the Mewlon was astigmatism which started showing past ~1.5”. Haven’t noticed it in the BeamTech so that’s good :thumbsup:

[strongmanmike]

Cool, thanks for that Col, wow, so in the 1.15" sub, was that the average of multiple Moffat stars and was that from PI calculations or from you looking at the pixels and multiplying by the image scale? How do Maxim results compare with Pix insight, do they concur well on the same files, whether individual or stacks?

Hope you don't mind all my questions, as a noob to all this I (and I imagine others) really apprete your feedback. Im just trying to get my head around how to make meaningfull and accurate FWHM measurements :thumbsup:

Mike

[alpal]

Good questions Mike -
it sounds like we're all using different measuring sticks?

Also - with your focal length of 1120mm with an image scale 0.84"/pix -
the closer you get to sub 1 arc second conditions the more error there will be in your FWHM measurements.

cheers
Allan

[Atmos]

Yeah, it was an average of a few stars in PixInsight. I don’t trust single stars as much and last night was a good example of that for me in Maxim. Although the average star was 1.5-1.7” I had a range of 1.2” to 2.3” in the dozen or so stars I scanned over.

One tool on PixInsight is called FWHMEccentricity which is great for checking your FWHM as long as you have a bit of an idea of what your Moffat shape is. Under good seeing I know to use a Moffat 2.5 but under poor seeing I need a Moffat 4 to get reasonably accurate results.

Moffat, Lorentzian and Gaussian all refer to different statistical distributions and are used to describe the different shape of stellar profiles, I’ve got no knowledge past that though :lol:

I haven’t done much of a comparison between PixInsight and Maxim but I could easily do that tonight. Maxim requires you to do something that our brains are really good at, statistics! We’re not great at getting accurate numerical values but if we look at a bunch of numbers in order we’re pretty good at identifying outliers.



A lot of what you say is the reason why I went for 0.5”/pixel. 0.4”/pixel is a bit on the low side of what I wanted but if I ever was lucky enough to get 1” seeing I’d be able to resolve it; undersampled but enough.

Ideally we’d all be using the same measuring stick when measuring seeing but I don’t know how Maxim calculated compared to PixInsight. I’ll give it a go tonight.

[atalas]

Excellent resolution Col :thumbsup: I'd like to see a no AI version as well I think....although I expect a great result with this good seeing data set using careful sharpening too.

[Atmos]

I did some quick comparisons on a random 120s Ha exposure last night.
Maxim: 1.65” (checked a few and made a guess on the average)
PixInsight FWHMEccentricity: 1.72” with a Moffat 2.5
PixInsight manually: 1.7”
CCDInspector: 1.83”

I’ll upload a non AI version tonight. No sharpening at all as I cannot do it any justice at all :lol:

Non-AI version. (https://cdn.astrobin.com/images/57732/2023/e796f353-bfac-48f3-b160-dc067d280efa.jpg) Bumped up the saturation a bit to show just how many blue stragglers there are!

[strongmanmike]

Interesting results Colin, good on you for posting them :thumbsup: I'm not surprised different analysis gives different results. I suspect there are so many variables in this task, across different systems and OTA's, image scales and pixel size etc not to mention depending on what software you use and the parameters you give each, that an absolute measure of the real FWHM or seeing is next to impossible to be confident of. Having said that, your resolution is clearly very good regardless and for me the improved seeing up at my new observatory, over my previous location, was pretty obvious immediately, just by seeing the subs coming down (https://pbase.com/strongmanmike2002/image/173195561/original) on the screen each session and way before I tried Maxim. They are fun tools to have I guess but in the end our eyes are pretty good judges :)

Thanks for all your input...go Mewlons and Beam Techs! :P

Mike

[Startrek]

Colin,
Great result on a spectacular glob
Incredible resolution
Thanks for posting
My aim over the coming months is to image a few globs under various conditions ( hopefully at both sites ) and apply Startools Spatially Variant PSF Deconvolution to how it stacks up against BlurXterminator.
So far it’s performed beyond my expectations

Cheers
Martin

[Atmos]

Glad to bring up this whole conversation, it’s an interesting and important topic. I’d say that there isn’t much difference between Maxim and PixInsight measurements.
Two nights ago seeing was around 1.6”, last night was 3.5” :lol: the grass and everything was bone dry this morning, no dew at all! Dew in my backyard much be a dictator of good seeing.



Do it! I love globs! NGC 6752 is a great one. NGC 362 is one I want to have another crack at, it’s smallish and forgotten due to its proximity to NGC 104. M4 is a good one, not too small or dum but there is a lot of dust to contend with so colour calibration is a pain with the reddening.