Omega Centauri and stacking

[BC]

I tried taking 10 x 30sec shots of Omega Centauri last night, 10 darks and 10 bias and put them all together in DSS. While the background noise cleared up nicely, the basic globular blew out badly with no detail. I reduced the size of the stack until, really the best central detail is with a single 30sec shot. How do I stack without losing the central detail? I have attached both the single shot and a stacked one. This stacked one is just 2 lights, 2 darks and 2 bias.

Canon 500D
Equinox ED 80
ISO 1600
30sec

Thanks,
Bruce

[silv]

not suggesting it's true. just leaving a comment so I can follow your thread.

the sweet spot of your focal path was not where the camera's sensor looked. hence, the stars within the cluster have more coma than they'd have if you'd captured them in the middle of your sensor/the focal point.

I reckon, for DSS the Omega centauri is very difficult if the stars are not absolutely accurate. the process takes an educated guess and averages what it finds. And then there's the noise signal to filter out... oiyoiyoi ... I think.
As I said, only leaving a comment to easily follow your thread.
:)

[BC]

Thanks silv,

Interesting points. I don't have a field flattener which I'm sure makes it more difficult and yes, I focussed on a star with live view, and it was in the middle of the view. I couldn't see OC in the view but it showed up in the image and I thought, "well ok". Clearly I should have repositioned.

Bruce

[jenchris]

I see what you mean.
I sometimes wonder if the upper atmosphere being a bit uneven causes this - sometimes a lunar shot that is stacked can have some awful aberations when stacked that aren't visible as a single image.
I know globular clusters I take often blow out when I stack them even if the guiding has proven to be pretty good. More so when the jetstream is fast

[silv]

I don't know whether that is a factor. I'm a noob, like you. Hope, someone else sheds light into your question.

You could try something in DSS with the existing data:
set the star detection threshold (Advanced tab) higher to (possibly) tidy up that bright blob in the middle of OC.
or/and reduce the number of images chosen by DSS to 60 to 70% (main registration window)
maybe that helps and you can enjoy more what you already have.

or just take one single light frame against your darks/bias?
please, someone else - enlighten us. :)

[BC]

Yes silv, we are thinking the same thing. The last thing I tried last night was the single light with a dark and bias, but the result was similar.

I can see how stacking helps a great deal with nebulosity, but I began to wonder whether it was actually good once you have a basic star point. I assume it must be because we see some wonderful images which are stacked.

Bruce

[silv]

"the single light with a dark and bias, but the result was similar" <--- very good to know. care to upload that?
so all you need is someone who knows about proper cluster DSS settings and its limitations.
:lol:
edit: and someone who reads your first post thoroughly.. I'm so sorry! :o

[rolls05]

Would knocking the iso back to 800 help? Cant wait to try that glob myself. I'm still trying to get a nice milkyway, so I can't talk.:( To my eyes (uneducated as yet) they look ok for a first effort.

[whzzz28]

BC,
Do you have a copy of the raw stacked file?
So when you stacked in DSS did you save the file as a TIF or FITS file?

If so, possibly upload it to a filehost (like rapidshare.com or something) and put a link here. We can take a look at it a bit more and play with it.

A quick look at it though leads me to believe it is more of an image processing issue rather than a data issue.

Your focus is a little bit out and there is a tiny little bit of elongation in your stars, but nothing too bad.
The glob not being in the center of view isn't too much of an issue. Ideally you do want it to be in the center but the issues you are seeing aren't because it is off-center.

Also Omega Centauri often blows out in the core. The amount of stars in that small space is massive. This often causes a white wash in the center, and can be processed out somewhat with different tools.

[BC]

silv, I will recreate the single stack later, just a bit busy right now.

Nathan, I'm not sure how to do the big upload to other sites. The original .tif has been played with too much anyway by now. I'm encouraged that it might be a processing issue, because really I'm pretty amateur at that.

Thanks,
Bruce

[jenchris]

I had a look at my images of Omega Centauri and found my sharpest image was a single frame too.
This is it - it is a little overworked but I think it needs to be a bit eyepopping

[Forgey]

This is my omega centauri

Taken on the 11th June 2012 with a Canon 1100D modded, Saxon ED100
Total of 50x15sec exposures @ ISO-800, Stacked in DeepSkyStacker and processed in Photoshop CS5.

sorry I can't remember the DSS setting I used.

[ourkind]

Wow :eyepop: Paula that's beautiful!!!

Unfortunately I do not have much to add as I cannot speak from any experience, like Silv I'd like to follow this thread :)

[BC]

jenchris's example is the sort of core I'm after. It really looks quite sharp and it's interesting that it is a single shot. Like my single shot, the background isn't as black as the stacked examples. Forgey's example has the really nice background and while it has a lot more detail in the core than my attempt, the processing makes the core just a little less detailed than jenchris' single core (not trying to be personally critical here, just trying to understand a good approach).

How would we go about retaining the single shot core and the stacked background?

Thanks all for your contributions,
Bruce

[rolls05]

I'll bet some one mentions masking. Other than that ,I got nothin, Loving the thread.

[whzzz28]

Masking isn't really a good idea for globs as you will get funky illuminated stars.

There are other tools that will help remove/reduce it, for instance the Local Histogram Equalisation (LHE) tool in PixInsight helps a lot with OC.

Also getting the right exposure length will help too. I did 120sec subs on my CCD and the results are a little too much, but LHE gets rid of a lot of the core brightness.

I still think there is an underlying processing problem with the OP's image. I don't mean that disrespectfully either.

Just remember: when trying to get detail out of an image, bring it out with Curves. Use the histogram (levels) tool to only change the black level and don't try and get the detail out with one pass of curves. Do a little at a time.
If you stretch based only on Levels (histogram) and bring the mid-tones point down then you will blow out your stars, and given this is a glob made of nothing but stars, this is important.

[cometcatcher]

Stacking doesn't overexpose the core, post processing does. If one frame doesn't overexpose, neither should 1000 frames stacked.

To prevent core burnout, don't do any highlight stretching. Only stretch mid-tones.

[Rex]

Hery Bruce, you may have this already sorted but when you saved it from DSS, there are two tick boxes one that says " apply changes to saved image" or something like that, the other says " embed changes but do not apply" make sure that you have the Embed but do not apply ticked as DSS really does strange things sometimes when you apply them. That's my experience anyway. Hope that helps. Your saved image will be a lot darker and quite often you can't even see anything until you stretch it a bit.

[BC]

Excellent comments both Kevin and Rex. I thought I was getting a blown core before I did any processing, so I wondered about Kevin's comments until I applied Rex's suggestion about the embedded option. I had been running with the default which certainly was killing the centre right from the begining. So now I have been able to stack my 10 x 30 secs (small in reality I know), and the result (as Rex suggests) is quite dark to begin with. I am a little uncertain whether I am "stretching the mid-tones" when I move the centre triangle in the histogram, but the the core is looking really quite good, however it seems to be at the expense of the background which is getting some dark green noise.

Anyway, thanks guys for the tips. When I get something I'm not embarrassed to put up I will do so.

Bruce

[cometcatcher]

For green noise give the Hasta La Vista, Green filter a try. Personally I love this filter.

http://blog.deepskycolors.com/archivo/2010/04/26/hasta-La-Vista-green.html