SH2-73 Reflection Nebula in Hercules

[old_eyes]

My first process on 4 hours data on SH2-73 in Hercules on the border with Serpens. Pier 5 @Roboscopes in Spain (Tak Epsilon 180, ASI 2400MC Pro, Paramount MX unguided).

I am happy to have pulled out this faint reflection nebula, but it was a bit of a struggle as there are a number of small galaxies in the field of view that tended to blow out or look very weird if not carefully masked during the stretch.

Anyway - here we are. The bright star to the top right is 44 (pi) Serpens.

 

[Sunshine]

Wow! your images are so clean and contrasty, the star colours are just beautiful! how do you control noise so well? that's what I meant by clean.

Do you have a website where I can see more of your work? 

Edited March 9, 2023 by Sunshine

[old_eyes]

The data is good as I am using a remote observatory in the mountains of southern Spain (www.roboscopes.com) rather than my cloudy, murky backyard in North Wales. However, even here only about half the subs were usable due to high thin cloud (that plays merry hell with faint nebulae).

All processing was done in Pixinsight. This is one-shot-colour so we are RGB all the way through. If I remember, the processing went something like this:

1. Basic calibration and stacking using WBPP
2. DynamicBackgroundExtraction to get rid of gradients
3. BlurExterminator to control stars - I was using a 300sec exposure so stars got a bit big
4. SpectrophotometricColorCalibration to adjust colour balance using a preview of part of the image I felt was genuine sky as the background reference
5. Starnet2 to remove stars and create a star mask
6. NoiseExterminator on the nebula image to reduce colour noise in the background. I am trialling NoiseExterminator at the moment. EZ-Denoise gives about the same result to my eyes, but NoiseExterminator is faster
7. MaskedStretch to see what I am looking at for further stretches
8. RangeSelection to create a mask that would protect the galaxies in the field. As the outer layers of the galaxies were the same brightness as parts of the nebula, I imported the range mask into a paint program, and painted out any nebula and the halo of the bright star 44-Serpens so I would only protect the galaxies. Then imported mask back into Pixinsight and applied to the nebula image.
9. GeneralizedHyperbolicStretch to reach a final stretch I was happy with
10. HistogramTransformation to get a black point I was happy with
11. ArcsinhStretch on the star mask to stretch and boost colour
12. PixelMath to combine the two images using the max(starless, star_mask) function to choose the brighter of the two pixels at each position.

Seems complicated when I write it down, but it all flows quite naturally. The only place where I had to go back and redo a couple of times was 8 & 9. I spent a while figuring out how to do the stretch without blowing out the galaxies. Couple of false starts there.

[old_eyes]

Oh - I forgot. There was a small amount of LocalHistogramEqualization added between 9 & 10 to improve the contrast of the wisps and threads in the nebula.

[old_eyes]

29 minutes ago, Sunshine said:

Do you have a website where I can see more of your work? 

Not yet. I do not compare with the true masters, but I am making progress and perhaps one day 🙂. Other images are scattered around this site. Both from Roboscopes and my home observatory. One day I will sort them out, but at the moment - too much data!

[MartinB]

I love this strange mix of a seemingly mainly mono image with the dramatic blue reflection and the small coloured stars.  Fabulous!