NGC7635+M52, re-processed with 2 year gap

Anything that doesn't fit into any of the other forums
Post Reply
User avatar
turfpit
Posts: 1359
Joined: Mon Feb 13, 2017 8:13 pm
Location: UK
Contact:

NGC7635+M52, re-processed with 2 year gap

Post by turfpit »

I was looking at some old data and decided to reprocess. Some detail of the differences in the side-by-side are lost due to JPG compression.

compare.JPG
compare.JPG (212.19 KiB) Viewed 161 times

The 2018 effort was calibrated & processed in Siril, then 'tortured' (too many curve stretches) in GIMP. The 2021 effort was calibrated and processed in Astro Pixel Processor, Siril was used for green noise removal and processing completed with a 'light touch' in GIMP. The additional 2 years of experience in imaging has given me a better idea of what end result to aim for.

For the 2018 version:
  • background was too black
  • star colours washed out
  • brighter stars had a blue effect at the left hand side
  • Bubble Nebula was orange
  • evidence of a 'green caste'
For the 2021 version:
  • background now richer and less black
  • stars have richer colours
  • bright stars are 'tighter'
  • brighter stars not over-saturated
  • Bubble Nebula red, which is correct for an Ha region
  • green caste removed
  • the 'strained' look of the image has gone
The Subtractive Chromatic Noise Reduction (SCNR) technique is used to deal with green noisy pixels (a by-product of having 2 green pixels for every blue and red pixel in the camera sensor matrix). This is a 'one click' operation in Siril.

The Digital Development Process pre-sets in Astro Pixel Processor have contributed to the image improvement as the uncontrolled curve stretching by me has been removed. The DDP available in dedicated astro-image processing software is of great benefit.

If I was to re-process using Siril I think I could improve over my original due to better understanding of the functionality available in the software.

Capture was with an Altair 66ED refractor and Altair 183C CMOS camera.
The original data: lights 60x60s, darks 50, flats 100, bias 50.

With experience some of the capture settings could be improved. Gain was 150 and black_level set at 100. If I revisit these objects, I will try gain=400 and black_level=50 plus increase (at least double) the number of light frames.

The exercise demonstrates to me that it is worth revisiting old data to measure how skills are evolving.

Dave
User avatar
oopfan
Posts: 830
Joined: Sat Jul 08, 2017 2:37 pm
Location: New York
Contact:

Re: NGC7635+M52, re-processed with 2 year gap

Post by oopfan »

Thanks, Dave.

Your statement rang a bell:
The Subtractive Chromatic Noise Reduction (SCNR) technique is used to deal with green noisy pixels (a by-product of having 2 green pixels for every blue and red pixel in the camera sensor matrix). This is a 'one click' operation in Siril.
I have virtually no experience with OSC, except for my first camera, the ASI120MC, when I got started in AP. I do remember the double-green pixels. I need to familiarize myself with the demosaic process, but it seems reasonable to me that the extra green pixel may be used to provide luminance. If that is the case, then I wonder why there is such a heavy green cast? It makes me think that the demosaic algorithm, in some cases, isn't being supplied the proper inputs to efficiently operate. Maybe there is more to it than simply supplying the pattern, for example "RGGB".

Brian
User avatar
turfpit
Posts: 1359
Joined: Mon Feb 13, 2017 8:13 pm
Location: UK
Contact:

Re: NGC7635+M52, re-processed with 2 year gap

Post by turfpit »

Brian

Gary Palmer did mention 'green bias' for one shot colour and DSLR cameras in a Deep Sky workshop I attended. I don't truly understand the mechanics but it is easily dealt with in Siril. PixInsight has SCNR removal as straightforward process. In APP I think the tool HSL Selective Colour can be used but I haven't managed to decrypt that yet.

Dave
nexusjeep
Posts: 252
Joined: Thu Oct 18, 2018 3:01 pm
Location: Gloucestershire

Re: NGC7635+M52, re-processed with 2 year gap

Post by nexusjeep »

In APP you can also correct it using the star colour calibration tool as there a sliders to adjust green or magenta casts in the image there is a video were Mabula shows how to use the star colour correction tool and the start of this is removing a green cast from a wide field DSLR stack.

I think the green cast is inherent from the RGGB arrangement if I do not neutralise the background in APP when processing a stack then the image is always green toned until I correct it later in one of the APP tools or initially select the neutralise background option when integrating. You also have the HVLG plugin for photoshop that can be used to remove it after integration the plugin also works with Affinity Photo.

Cheers
Nick
Post Reply