StarTools

Last week we had 2 consecutive clear moonless nights which is extremely rare in Summer. Captured NGC 3324 Gabriela Mistral Nebula which is located on one of the outer arms of Carina

Location : Sydney , Australia
Skies : Bortle 8 heavy light polluted skies
Conditions : Average to Good Seeing
New Moon
Very light SE wind
No Dew

Backyard semi permanent set up
Telescope : 8" f5 Klaus Helmerich Carbon Fibre Newt
Mount : Skywatcher EQ6-R Pro
Imaging Camera : ZWO 2600MM temp -10C , Gain 100
Coma Corrector : TS Optics GPU
Filter Wheel : ZWO 7 x 2"
Filters : Antlia 3nm Ha , Oiii and Sii
Guide Scope : Orion 60mm with helical focuser
Guide Camera : ZWO 120MM
45 x 3min subs Ha
40 x 3 min subs Oiii
44 x 3 min subs Sii
Darks from Library
Bias from Library
Flats for Ha, Oiii , Sii
Tracking and Goto : Eqmod and Stellarium
Aquisition : APT
Guiding : PHD2 ( 0.45 to 0.55 arc sec total rms avg )
Analysed , Calibrated and Stacked in ASTAP
Decided to process in an older version of Startools version 1.7 to see what sort of result I could achieve , I haven’t used 1.7 in 2 years. ( Data Sets loaded into Compose , Color SHO: HST preset with tweaks in all colours )

Total integration 6 hours

I decided to leave the Auto stretch as default and try to include as much nebulosity and detail as possible with the hours I had on this object.
The actual noise floor only required a 2 pixel de noise
Fairly pleased with the result

Astrobin link below for full resolution…..

https://www.astrobin.com/ttfhzt/

Comments welcome
Thanks for looking

Clear Skies
Martin

NB: I decided to reprocess to give the image more contrast and detail
I think it’s marginally better than the original ?

Beautiful shot, Martin!

And the retro 1.7 processing looks great too.

Seems you may be revisiting all your greatest duoband/OSC hits, now in SHO/mono and I suppose the CF Newt as well? The third channel - if there's data to be found - really does add the touch that takes the overall image to the next level.

Are the crops just crops of the same final image, or separately processed at higher resolution (i.e. less bin)?

Hi Mike,
Thanks for your comment
Much appreciated
After struggling for a long time with HDR and SV Decon in v1.8 , I thought to myself why not give 1.7 a whirl and see how it performs with Mono data sets.
Whilst painfully slow compared to 1.8 , 1.7 did a great job on this image.
My images are usually Binned to 71% ( I hate pixel peeping and seeing blocky small stars ) thus leaving an image size of around 4,200 x 2,800 after cropping edges. Some of the modules in 1.7 like Sharpen , Superstructure and Decon can take 2 or 3 minutes to complete.
SV Decon using samples in 1.8 is way too powerful ( even at 5 iterations) and leaves the star core pixels at max white point ( coalesced) which looks abnormal . Even using Synthetic has a similar look. At least with 1.7 Decon the stars retain their natural captured look from the AutoDev stretch even though the algorithm has its limitations ( there is a point of diminishing return )
I’ve tried everything in 1.8 SV Decon to alleviate the “whited out” coalesce star core issue. Don’t get me wrong the Deconvolution in 1.8 is unbelievable but those damn whites out Star cores detracts from the result.
Next is HDR , I haven’t used 1.8 HDR is a long time , it’s way to powerful and leaves the nebulosity clouds and structures with a shattered glass look or dark spots look
Although to be fair I have used the Tame preset with some success on certain objects.
Now HDR 1.7 is brilliant, always had good results and has excellent control over its function.

Yes after using the 2600MC for just over 3 years It’s hard not to revisit the same objects to see what the 2600MM can do. Honestly in regard to Narrowband imaging , the 2600MM with dedicated filters is leagues ahead of the 2600MC with Dualband filter. I image from a big City ( Bortle 8 ) and rural country side ( Bortle 3 ) and the 2600MC whilst an excellent camera is not in the same league as the 2600MM in relation to signal fidelity, noise and fine detail.

The crops are just of the same image ( using the Crop module )

I most probably use both 1.7 and 1.8 moving forward as 1.7 is an excellent version
I haven’t taken the plunge to 1.9 , but I suppose I should at some stage

Thanks again

Clear Skies
Martin

Hi Martin,

After 1.9, going back to 1.8 is similarly SLOW!

Would be interesting to hear your takes on it though. SVD has been changed up a fair bit, and there's been some relief to the creation of flat white disks for coalesced star cores, which I found troubling myself. But there's a lot going on there that I maybe haven't quite figure out yet, including the stellar profiles that come out of AutoDev/OptiDev.

Have you tried increase dynamic range at all? Possible that can alleviate too much power being put into the stars, though it then dims everything.

I like new HDR myself and can usually get it to do what I want by adjusting the parameters. The only issue I sometimes run into, compared to say 1.7, is that the affected field sometimes seems too limited, even at the default of 50. And going beyond 50 requires more horsepower than my computer can throw at it. I know it was said that newHDR's default range was still more than 1.7, but yeah sometimes it doesn't seem to act the same and things get left off the table.

Interesting that you are binning 71 for a 4000-wide image at your focal length. For APS-C I've been processing at around 3000-wide lately, up from my previous 2000. Generally I don't think I've been seeing squared stars, even at only 580mm.

The 2600MC is no slouch, and while I never took that step (jumping right from DSLR), full 2600 mono is almost like cheating, even in heavy LP.

Thanks Mike,
I’m sure SV Decon since inception has been a works in progress and can only improve for the next version 1.10 ( if Ivo has the time and motivation)
In regard to coalesced white star cores , I've tried everything in 1.8 to alleviate them including adjusting dynamic range, can’t really solve the problem to be honest.

I guess trying 1.9 is my next challenge

In relation to Binning at 71% , when I referred to blocky stars , I mean those iddy biddy tiny stars when you zoom in to say 800% . I guess I’m a control freak as those stars don’t even have enough illuminated pixels to form any centroid or concentric shape.

I don’t need to improve SNR by binning at say 50% like I did with my 2600MC as the data from the 2600MM is so clean and low noise ( using my Antlia 3nm NB filters the noise is almost negligible,using the denoise module only requires around 1 or 2 pixels only, and this is under Bortle 8 and Moon glare )

Yes the 2600MC is an excellent OSC camera and is a great companion for the 2600MM

I don’t know about cheating , Mono acquisition , stacking and processing is a big deal , it’s definitely not easy

Clear Skies
Martin

Have your tried the new DSO Core preset in 1.9 HDR? I can't find any specific documentation on it; but from its parameter settings I am guessing that it may be targeting the result 1.7 used to give.

Thanks for your suggestion and advice
As I mentioned I’ve had success with 1.8 HDR Tame which is similar to DSO core in 1.7 HDR
However , I’ve tried to emulate 1.7 Optimise Soft ( and even Optimise Hard) in 1.8 HDR and it ends up unsatisfactory, it doesn’t accentuate specific structures and target key areas like 1.7 does
The only drawback from 1.7HDR is that you do end up with slightly bigger Stars , which its affect can be reversed in Shrink.Other than that 1.7 HDR is a brilliant module

I don’t know why but it seems way too powerful even throttling the settings back
The default setting of 1.8 HDR leaves my image looking like shattered glass with dark smudges

I did post a thread on this issue a year or so ago and Ivo was very accommodating but I never really ended up happy with 1.8HDR ( except for Tame )

Thanks
Martin

How difficult would it be to add a separate alternative HDR module with the 1.7 version functions in the current 1.9 Beta?