First proper mono attempt - Orion Nebula - comments and critique appreciated.

[SiD the Turtle]

Hi folks,

Recently I've moved to a mono camera setup and managed to get a couple of nights imaging earlier this year. Finally I've had a little bit of time to have a stab at processing them, so I've included the final result and my steps below and happy to receive critique and advice as this is the first image I'm quite chuffed with sharing, but I can see a number of areas for improvement still. I'm also very much a beginner at PixInsight and it's thousands of options.

Final Image

Image Capture

• HEQ-5 Pro with belt mod
• William Optics ZS73-II with field flattener
• Skywatcher Evoguide ED50 guidescope with ASI120MC-S guide cam
• ZWO ASI1600MM-Pro
• ZWO LRGB and Ha stock filters
• ZWO electronic filter wheel electronic auto focuser
• Astroberry/Kstars/EKOS/PHD2
• 26 red, 15 green, 15 blue, 30 L, 7 Ha (didn't use the L in the final image, see below), all 120s at -20c. So excluding the L, about 2h of subs. Had one night capturing the first 15, then a second night that I got some more red, L and the Ha before the clouds rolled in.
• Dithering after every frame and autofocusing on each filter change
• Darks, flats using paper over the lens and pointed at daylight method, no bias (understand no value in bias for this camera)
• Bortle 5 officially, feels darker as I'm on the edge of a small down, but for Orion pointing over the haze of the town centre + moonlight

Image Processing

Using Pixinsight alongside Light Vortex Astronomy tutorials:

1. Weighted batch pre-processing of subs to align and create masters for each channel, used default options
2. Dynamic crop edges caused by dithering
3. Dynamic background extraction on all channels to get right of some of the light pollution and a gradient (though luminance was a failure as it found loads of bad points, but I pressed on and hoped for the best)
4. Linear fit
5. Chanel combination on RGB, background neutralisation (didn't seem to make much difference) and then colour calibration and SCNR to remove a lot of the background gradient
6. LRGB Combine, but this created a muddied image so trashed it
7. Added Ha data to RGB using PixelMath
8. Lots of stretching using Histogram Transform

Raw images (all auto-stretched)

Red

Green

Blue

L

Ha

Master Dark - note the vignette, this is with the camera detached from the image train, lens cap on, in a boot sock and in a closed cupboard. This doesn't look right.

Master Flat for blue - looks okay as I think as it's picking up the dust motes and vignette, but does it look 'wispy'?

RGB combined and L stretched prior to LRGB combine

LRGB combine (thrown as I couldn't get this looking better even with further stretching)

Including Ha - the raw image, then with DBE, then stretched

May have made the final stretch too dark.

Final image, rotated and cropped again.

Notes

• I haven't quite mastered masks so I haven't done anything with noise (when I tried a lot of the detail disappeared).
• Similarly I haven't tackled the stars, bloat etc.
• The middle of the nebula was hard to avoid blowing out when trying to bring out detail elsewhere.

Thanks for making it to the end!

 

[barbulo]

Nice image @SiD the Turtle
I’ve recently moved to mono as well but haven’t dared this target yet. I consider it a tough one because of its high dynamic range. 
Good job!

[Xsubmariner]

A cracking image for your first venture into mono AP. Adding more frames over time will improve the signal and will deliver a great image.

[Skipper Billy]

You are certainly on the right track.

In think your blacks are a little 'clipped' and maybe thats deliberate to try to reduce the brightness of the core??

Its a hard target to get right - ideally you need a full set of very short exposures for the core - and a set of long exposures for the whispy parts the blend them together in processing.

Well done !!!

 

[wongataa]

Looks good.  The core of M42 is blown out though.  You can see the clipping in the channel images you posted.  In the future it would be good if you had shorter exposure images for the core area so you can get detail there and not have it blown out. I also found getting detail in the bright core of M42 without it blowing out tricky.

[The Lazy Astronomer]

That's looking really nice 👍

As has been said, short exposures for the core are needed for M42 - blend them together with HDRComposition.

L is where all of your detail is, so generally you'd look to shoot more L than RGB, I usually do 3:1:1:1 LRGB. Because L is a less restrictive filter, you might also want to consider a lower exposure time - with B5 skies (same as mine) it's likely that your optimum minimum exposure time for L will be <10s, so plenty of scope to reduce exposure time without a loss of recorded detail.

The L channel also seems to still be affected by a light pollution gradient, which is limiting what autostretch can do in terms of pulling out detail, hence your L doesn't look as good as your RGB, so maybe revisit DBE here.

For some of the more complicated things like deconvolution and noise reduction, I think the EZ processing suite is a good tool to start with; it takes some of the manual complication out of the processes while you're learning.

And finally, maybe consider looking into the fairly recently released GHS script for stretching, it can help prevent blowing out highlights when trying to dig out faint detail. 

[Hicks]

Great image, but the post itself is just as good. It's nice to see the raw frames that made up your final processed image. It's always a challenge to know if your own final result could be improved more by better post-processing skills, or better data and seeing what others get out of their data is a great help.

I haven't tried M42 since getting a set of colour and narrowband filters, but I remember it been a challenge with no filters just to get the stars in the core to show without no longer having the whisps. As others have mentioned, long+short exposures and some post processing combination will help with that.

[SiD the Turtle]

Thanks all for the constructive advice, lots to read up on and research, but sounds like I've got some good ideas from you all for trying to improve the image (plus waiting for another clear and free night to get some shorter subs).

On the topic of clipping- what are you looking for in the histogram to see clipping? Are we talking data right at either end of the spectrum and also maxing out the top of the graph?

Edited April 24, 2022 by SiD the Turtle

[Skipper Billy]

1 hour ago, SiD the Turtle said:

On the topic of clipping- what are you looking for in the histogram to see clipping? Are we talking data right at either end of the spectrum and also maxing out the top of the graph?

The left hand example shows clipped blacks, the right hand clipped highlights.