NGC6946: How can I improve the post processing on this image? - Forums

Topic starter

Aug 21, 2025 · Edited: Aug 23, 2025

This is about 60h of data taken with a Seestar S50: OSC camera, 50mm aperture, 250mm focal length. I got 85h kept the best 60h and drizzled 2x.
Post processed in PI. SPCC, BlurX, NoiseX, StarX, Statistical stretch.
There was quite a bit of IFN/tidals in the data but I did want to emphasis that too much. Wanted to focus more on the stars and the Galaxy color. Worked mostly with curve then on the color.
How can I improve the post-processing?

Concise Engaging

kuechlew

Aug 21, 2025

Great image and kudos for investing so much time collecting a lot of data. Must be painfull to integrate all those thousands of individual images, allthough most likely a capable Program like Astro Pixel Processer will handle it well. Not sure if I would throw away 25h of data, so maybe you can be somewhat more conservative in this respect.

When it comes to stretching I prefer Generalized Hyperbolic Stretch over anything. It gives you a lot of controll and the amount of data collected certainly deserves to get the fainter stuff represented well in your image.

Clear skies

Wolfgang

Helpful Respectful Supportive

Olivier Lefebvre

Topic starter

Aug 21, 2025

Thanks for your message Wolfgang, appreciate it.
I stack using PiSiril with a custom script that batches the data and then do stacks of stack. It is scalable and quite quick (about 2h to drizzle and stack 16k subs).
I cull quite a bit based on FWHM to improve the outcome of drizzle and since tracking of S50 is not great. But I'll try to keep more of the subs next time to see if that improves the outcome.

I hear you regarding the GHS, the one click statistical stretch is tempting but it is probably sub-optimal. Will get more into GHS to improve, thanks!
Clear skies to you as well,
Olivier

Helpful Respectful

Mikołaj Wadowski

Aug 21, 2025

The overall intensity is spot on in my opinion. The colors are okay but are starting to shift to purple, especially near the core.

I think there are two main ways to improve the image. The first is easing off blurx strength. The image is very oversharpened, making the structures and dust lanes turn into little worms. Second is denoise - the image is perfectly smooth and gets blotchy in the outer arms of the galaxy. I would back off the denoise settings and try to work on the blotch, perhaps by changing the denoiser (try deepsnr) or correcting it with MMT.

A more minor point of improvement (though also important) would be neutralizing the background. It has a pretty obvious blue cast, correcting this is trivial.

Well Written Helpful Insightful Concise

Peter LeFebvre

Aug 22, 2025

There is a video on the PixInsight channel on M101 in HaLRGB. Even though your data is OSC many aspects of this series apply, especially the sections where all the channels are combined. It goes into detail on SPCC for this type object, how to manage the spiral arms using luminance masks, and HDRMT, MMT, and Local Histogram Equalization to fine tune as final touches. What you’ve done already looks great to me but these are the sorts of techniques that get it to the next level.

Well Written Helpful Insightful Supportive

Olivier Lefebvre

Topic starter

Aug 22, 2025

Thanks Mikolaj for your message.
Yes, I am not 100% happy with the colors either, I was going more for a blue/red than magenta/purple. Will keep working on this. Noted for the over sharpened, over de-noised comment. I will try to ease on this for the next revision.
The blue cast in the background was intentional, but maybe not such a good idea. I will make it more neutral for next revision to see if that improves things.
Thanks for your thoughtful suggestions!

Well Written Respectful

Olivier Lefebvre

Topic starter

Aug 22, 2025

Thanks Peter for your comment,
I will look into these more advanced techniques. I have watched a few of Adam Block videos on this but I have yet to fully utilize them on an image. I will check the M101 video too.
Thanks for taking the time, much appreciated!
CS,
Olivier

Well Written Respectful

Olivier Lefebvre

Topic starter

Aug 23, 2025

Thank you all again for your comments and suggestion for improvements.

Here is a revised image of the same dataset.

I did less sharpening and de-noising, used GHS to bring-out the fainter details more . I also reworks the colors.

I think it is probably better overall, but I am still struggling with star halos (purple and blue) specially for the stars in front of the galaxy that leave unnatural color. I manually masked them and de-saturate them, but not sure if this is the best way to deal with this or if there is a way to avoid these halo all together.

There are also what looks like de-convolution artifacts in the outer arms/dust of the galaxy. Are there known techniques to prevent this? I think I’d want to do a de-convolution only within the galaxy and probably a convolution to add a bit of blur to the outer dust/IFN.

I might have pushed the brightness too much in the galaxy too. I wish I had more contrast within the galaxy.

How can I improve this revision?

📷 NGC6946_crop_v3.png