Astrophotography Processing Challenge October 2024

[AstroBee]

CALL FOR DATA!
We are in need of processing challenge data. Please consider submitting data for this fun challenge.
If you have data that you are interested in sharing, please reach out to me via PM so we can discuss it.

For October this year we have some narrowband data that was kindly submitted by Carole (carastro) of NGC 281, The Pacman Nebula.
Carole mentions that the SII stars appear a bit bloated compared to the OIII & Ha data so be mindful of this during your go at processing.

Carole's comments:
This was taken in Bortle 6 back in 2021 at a friend's house bordering SE London/Kent UK.

Skywatcher ED120 and FR TS 0.79
Atik460EX Baader filters
Ha 22 x 600 (3h 40m)
Oiii 16 x 300 binned (1h 20m)
Sii 16 x 300 binned (1h 20m)

Total imaging time 6hours 20 mins
---
Carole requested that I re-register the files and crop accordingly. I did do an image registration on the files and found very little need to crop so check those corners just to be sure I did not miss anything.
Here is the link to the download files and as usualy, I've provided XISF, FITs & TiFF formats. No need to download all three sets as they are identical except for file format.
https://drive.google.com/drive/folders/ ... sp=sharing
Also in that drive is a PNG image of Carole's edited version. Some folks may wish to go into the edit with a clean mind and come up with their own version, others may want to peek at how she did it as a reference.

As always, have fun! It's a challenge, not a contest, so everyone is a winner!

[Graeme1858]

Excellent timing, we have cloud forecast for weeks!

Thanks Carole and thanks Greg.

[Jockinireland]

So here is my go at this. Thanks for sharing the data Carole and Greg for the Admin.

Processing was in pixinsight with minor tweaks in photoshop.

On mono images

Cropped
Graxpert
BlurXterminator (correct only)
NoiseXterminator

Combined as SHO using ChannelCombination
Stretched using Blanshans HTstretch_Unlinked
Correct magenta stars
removed stars using Blanshans Screenstars
Clonestamp to remove the light stripes at right of image
NarrowbandNormalisation
adjustments to contrast and saturation using GHS
Sharpened a little using MLT
Contrast adjustment with HDRMT
Noisexterminator

Export to Pihotoshop for :
Balance background using levels (I just find this easier than in PI)
Minor tweaks to color and saturation in Camera Raw Filter
Selective high pass routing to pick out some detail

Back to Pixinsight
replace stars using Blanshans Screenstars
Star reduction with Blanshans StarReduction (protect small stars)
Rotate and final crop
Save as jpg

[Graeme1858]

Here's my go.

I did an ABE, a Masked Stretch and a Starnet2 on each of the individual files and stretched the starless versions with Curves to bring out some detail. Think I might have over stretched them a bit! Then I recombined each with their stars with Pixel Maths and combined the three files with LRGB Combine (Luminance unticked)

Then I did a Starnet2 on the Combined version, Stretched the nebula, pumped up the Saturation and shrunk the stars. Put the Starless and Stars back together again with Pixel Maths, cropped it and did a bit of an Unsharp Mask.

Other than the over stretching, I'm quite pleased with it!

Nice data Carole.

[Graeme1858]

Combined as SHO using ChannelCombination
Stretched using Blanshans HTstretch_Unlinked

Hello David

Interesting that you Combine and then Stretch. Whenever I have combined without first stretching I get weird oddly coloured results.

I'm wondering if the clue to the answer is the Unlinked stretch that you use?

[Jockinireland]

Combined as SHO using ChannelCombination
Stretched using Blanshans HTstretch_Unlinked

Hello David

Interesting that you Combine and then Stretch. Whenever I have combined without first stretching I get weird oddly coloured results.

I'm wondering if the clue to the answer is the Unlinked stretch that you use?

Hi Graeme,

I find it very difficult to stretch before combining as because that can muck up the balance of the channels.

Blanshan provides 3 different options for stretching - linked, unlinked and GHS. The unlinked is specified by him for using on combined NB images prior to using his NarrowBandNormalisation. Doing it this way keeps the channels at the same balance as they come out of the camera which then allows NBN to work its magic. The image straight after Unlinked Stretch is, of course,very green and with horrible magenta stars but application of the correct magenta stars script, followed by NBN and then masked curves allows me to get exactly the colours I want.

This is the method which Blanshan describes and I find it works for me. My only niggle is that it means I cant use GHS for my initial stretch and I have found that in terms of stretch and contrast that I get better results with GHS. Then again, you cant use GHS before removing stars with StarXterminator. So my current method is as described above where I use GHS on the starless image to adjust contrast and saturation. it works, but not as well as using it for the initial stretch from linear.

Note that if I'm working on a RGB image which has been colour calibrated (SPCC) and background corrected, I will use Blanshans "Linked Stretch" to take from linear.

Hope that helps.

David.

[Graeme1858]

Your Blanshan Method clearly works well, as seen in your submission David. I will be looking into it.

Thanks.

[carastro]

Hi Guys, I only just got a chance to look in and see how this is going.

Nice results so far. Jock, you and I often seem to produce similar results, though I only use Photoshop.
I must confess I stretch each narrowband channel separately, particularly as Ha is always overwhelming and the other channels (when I do it) don't get a "look in" if I process them together.

Carole

[Jockinireland]

Hi Guys, I only just got a chance to look in and see how this is going.

Nice results so far. Jock, you and I often seem to produce similar results, though I only use Photoshop.
I must confess I stretch each narrowband channel separately, particularly as Ha is always overwhelming and the other channels (when I do it) don't get a "look in" if I process them together.

Carole

Hi Carole, yes we do seem do often end up with similar results.

I think that Blanshans NarrowbandNormalisation process in Pixinsight is designed to balance (normalise) the 3 channels in much the same way as you'd be doing manually when you stretch the separate channels in Photoshop.

Rgds

David.