M3 globular cluster first attempt

[Menno555]

This is my first real attempt with capturing a globular cluster. It's almost another discipline with capturing and processing compared with galaxies and nebulae
Also a bit connected with this topic: viewtopic.php?f=18&t=3877
I know short(er) exposures are working best for objects like clusters. But, longer exposures with lower gain and offset would reveal more also of the surroundings of the object. It's more of a "balance game" then to not blow the core out.

Bortle 7/8
Meade LX200 8" f/10 ACF OTA
Ioptron CEM25EC mount (no guiding)
Optolong L-Pro filter
Zwo ASI071MC Pro camera

Captured with SharpCap Pro @ -10 Celsius
103 x 30 sec / Gain 90 / Offset 20 / White balance R50 B50

Stacked in DeepSkyStacker

Processed in Siril and Photoshop
Siril: Photometric color calibration, Background Extraction and Histogram
Photoshop: Camera Raw Filter (blacks, saturation, clarity), Astro Flat Pro filter, levels, curves, reduced image size

[oopfan]

Hi Menno,

You are right, there is a lot of dynamic range in that globular cluster. The solution is HDR.

You've got two things going against you:
1. The Full Well Depth (FWD) of your camera. In your case, it is 46,000 electrons which is actually quite good.
2. The low dynamic range of most monitors.

Affinity Photo, a mere $25 cost, allows you to create an HDR image. So you need to create two images (TIFF) from two stacks: 30 seconds and 60 seconds. Click File -> New HDR Image. Add the two TIFF's (you can add more). Affinity will take those and compress the dynamic range down to what a monitor can represent.

I've only tried this on nebulae, not star clusters, but the same rules apply. Here is a video that I made of M81/M82/IFN from Insight Observatory's data. The 300s stack captured the core of the galaxies without saturation but the IFN was weak. The 600s stack showed off the IFN but blew out the galaxies. Affinity compressed those into a single image:

https://rumble.com/vcqant-brians-late-n ... d-flu.html

Brian

[Menno555]

Thanks Brain!! You really are some kind of Well Of Wisdom (so to speak ).
But really great tip! Photoshop has this too so I am going to try for sure with a new set but then with 60 or 90 second exposures.
Also busy with gathering data of M106 (which really is my favorite galaxy so far ) and this methode can be very useful for the bright core and very faint arms!

My monitor is a decent one. It has HDR but the lowest version 10 (some call it the "fake HDR") and Windows doesn't like that.
It uses 10bit though, so that's a plus. Plus it's hardware calibrated but that's more contrast and color related.

Menno

[oopfan]

Menno,

It's good to have a great monitor, but a lot of people do not. Something to think of, in the future your gallery should offer two images, one for HDR monitors and another for lesser monitors.

Brian

[timh]

Hi Menno,

Inspired by you and by Oopfan's comment I hope that you don't mind that I have been a copy cat ---plus the fact that M3 is conveniently placed at the moment.

Tried out the idea above of using a high dynamic range composition on the target--

So Bortle 6, just after waxing half moon with some high cloud,

8 " SW PDS200 Newtonian , Baader flattener, F5.0 on a CEM 70 mount , PHD2 guiding with an ASI 120 mm (not that good about 1.5 arc sec RMS due to conditions)

3 stacks - 10 minutes of 10s exposures at gain 124, 10 minutes of 20s exposures and about 12 minutes of 40s exposures combined together as a high dynamic range composition in PixInsight (the programme effectively uses concentric masks to vary the exposure in the final composition) then with a high dynamic range stretch to further flatten out the brightness at the centre.

By the time it gets translated from a 64 bit original to a small JPEG an a monitor it becomes quite a 'flattened' view but it does show some detail to the core and as on your rather nice original image the colours and blue stragglers are obvious. Not a great image --- poor conditions - but a first try at using high dynamic range composition.




TimH

[oopfan]

Tim,

I hope you don't mind. I grabbed your image and cropped it down to 18 arc-minutes across which is what the catalog says is the cluster's diameter. The original crop included too much space that detracted from the subject. The background was a tad bit cold (blue) so I warmed it up with a small shift towards yellow. I compared star magnitudes with a catalog. You are easily seeing down to 17th magnitude. That is really remarkable considering the total integration time of only 35 minutes. I can do that with my scope but I need a lot more integration time. It's amazing what can be done given the proper aperture!

Brian

[oopfan]

Gents,

That bright star in the lower-left hand part of the image is magnitude 9.5. As I said, I can identify stars down to magnitude 17. However, there are even fainter stars that aren't in the catalog. They are barely detectable but I see them. For the sake of argument, let's say they are at magnitude 19.5. So the difference between magnitude 9.5 and 19.5 is 10 magnitudes. That is a contrast ratio of 10000 : 1. (Calculate 10 ^ ((19.5 - 9.5) / 2.5) = 10000.) How many stops is that? It is 13.3 stops. (Calculate log(10000) / log(2) = 13.3.) So the difference between the first stack at 10s and the third stack at 40s is only 2 stops. I think you need more stacks. I am not suggesting 13 stacks, but at least a couple more.

Brian

[timh]

Interesting comments Brian,

I am glad that you had a look. I still haven't downloaded that freeware planetarium program that you pointed me to. It was a pretty awful night with high cloud when I took those frames and am hoping to repeat on a better night and over a wider range just to see what happens. These CMOS cameras are impressive - I was amazed last year when I first started using it in earnest and so easily saw the (lensed) double quasar in Ursa major at a mind-numbing 8B ly and mag 16-17 or whatever.

OK I think that I follow your logic. So I just opened pixInsight to look at the statistics of the core of the image to include the magnitude 9.5 star within a few of the 10s frames. That confirmed that, as the brightest point, the star was not saturating (about 20%) at 10s and unity gain. Logically enough it was getting up to about 90% in a frame I happened to pick from the 40s set.

I think that means that there is no point in me going any shorter than 10s - and 10s may have been shorter than was really necessary? So any room to dig deeper has to be in the other direction. So maybe 20, 40, 80 - possibly 160 ?! Of course the histogram will be well over to the right and the remaining dynamic range small at high exposures but the picture will be filled in by all the lower exposures.

Tim

[Menno555]

Well, I tried the HDR Merge mentioned by Brian and it indeed opens some new possibilities.
I am busy with M106 in RGB and Ha/OIII (secondary arms are showing up too to my surprise ) and the HDR merge of this is giving some interesting results already.
It's going to be a challenge to get the secondary arms, the main arms and the bright center to show and this might help a lot to accomplish it

Menno

[Menno555]

I totally forgot about this
I did try the HDR Merge option as suggest by Brian and that works really well. Tried it with M106 but that is giving positioning problems.
But M3 worked wonders. I used the original 30 second exposures and did add a set of 5 x 300 sec and 6 x 240 sec exposures.
Basic processing in Siril and merged the 3 TIF files in Photoshop. Big advantage is that with the merge, you can use the Camera RAW filter in 32bit.
The result is indeed stars visible in the center and more stars around it.

Menno

M3 HDR Merge 3 TIF files

M3a.jpg (841.16 KiB) Viewed 2043 times