pale/missing colours in dark areas

[decay]

Hi there!

I am currently reprocessing an image I took some years ago - now with StarTools, which I discovered in the meantime. This is really a stunning piece of software , and I would like to take advantage of the great and advanced possibilities.

Now I am facing the problem of missing colors in the faint and dark areas of M42 in this image. I tried a lot of different workflows and settings, but the problem is almost any time nearly the same. Perhaps someone could take a look and give me a hint - I am sure, I am only missing something stupid. I read the documentation and some threads in this forum, but I still had no luck.

- Newton 200/1000
- Canon EOS 2000D - unmod.
- Baader MPCC

- unguided

- about 50 light frames (each 32 secs, ISO 1600)
- no flats

(It was one of my first shots in AP and at this time my skills were not advanced enough to take flats. I will surely take a new shot with proper flats next year. But I would like to understand where the problem lies in this case and improve my understanding of ST.)

- stacked with DSS with the recommended settings (hopefully)

Here is a link to the FITS Output from DSS, if someone would like to take a look:
https://c.web.de/@334960167135216273/md ... RmvFL9CwPQ

This is the outcome of a simple and quick workflow:

- open file, Type of Dataset: Linear, from OSC/DSLR
- first bin down to 25% as the image is not that sharp and subsequent processing will be much faster
- AutoDev, Keep (yes, as expected there is strong vignetting visible due to missing flats)
- Crop out small borders for security reasons - despite no stacking artefacts are visible for me
- Wipe: the strong vignetting is not my issue in this case, so default settings and Gradient Aggressiveness down to 10%.
- AutoDev, Redo global stretch
- Ignore Fine Detail: 5px
- Color: The module comes up with reasonable balance, I guess. Perhaps to much saturation, but that's a just matter of taste, isn't it :
- _But at this point_ I am missing colour saturation in the outer darker and fainter areas, e.g. at the top and right of the image. It's almost only grey. OK: Bright Saturation down to 1.5 and Dark Saturation set to "Full" - but unfortunately still nearly no reasonable colour in faint areas
- Noise Reduction with Grain Size set to 2.0 px.

The resulting TIFF image may be found here:
https://c.web.de/@334960167135216273/W_ ... pBZRwDJXiw

The log may be found here:
https://c.web.de/@334960167135216273/3E ... qvAZq-BY3w

I processed the image quick and dirty with Fitswork, in order to point out the issue:
- shrink down to 25%
- Histogram stretch
- colour correction
- colour saturation

2021-04-10 17_51_59-Fitswork.jpg (134.49 KiB) Viewed 7308 times

(please compare with TIFF image result from StarTools, see link above)

cutout of the image processed with StarTools:

2021-04-10 17_52_41-StarTools.jpg (14.76 KiB) Viewed 7308 times

cutout of image processed with Fitswork:

2021-04-10 17_52_25-Fitswork.jpg (21.75 KiB) Viewed 7308 times

I have no idea what I am doing wrong. Or is there a problem with the dataset? Any hints or guidance would be much appreciated!

Thanks for any help, Dietmar.

[Burly]

links not working

[decay]

Hello Burly, thanks for trying and for your message. I checked the links, but they do work for me . Do yo get any error message?

2021-04-12 08_08_47-Online-Speicher.png (2.76 KiB) Viewed 7275 times

Sorry, it's only German , "Herunterladen" means "Download". If it really does not work for you, I will try a different file hoster.

Thanks a lot, Dietmar.

[Burly]

Ah must be my firewall , mobile data seems to be able to access will have another look .

[decay]

That would be very kind of you. Thank u. No need to hurry, image data is 2 years old

[admin]

Hi,

The dataset has already been color balanced at the very least (which is the bulk of the problem), but I did not have any particular trouble replicating the FITSWork rendering...

StarTools_559.jpg (253.01 KiB) Viewed 7241 times

On my color calibrated screen I overdid it with the dark saturation, but I did that so there is no doubt that the color is there if you want it (and JPEG also meses with the red channel quite badly).

Because the data was already color balanced prior, I set all multipliers to 1.0 (which mimics the FITSWork color balance).

This image was the result of a Crop, Bin to 25%, Wipe with Uncalibrated 1 preset, AutoDev with Ignore Fine Detail < set to 3.2, and then straight into the Color module with settings as seen above (Legacy preset to mimic the way legacy software like FITSWork destaurates highlights, Dark Saturation set To Full, Bias multipliers all set to 1.0).

Does that help?

[decay]

Hi Ivo, thanks a lot. At first I will double ckeck the settings for colour balancing in DSS. This would mean, that a lot of or at least some my other images have been wrongly stacked as well . I will give a short feedback in any case.
After that I will try the settings you described. My goal is (surely!) not to imitate the Fitswork output, but i would like to have some colour in the fainter areas. I tried many times, but I never reached your result.

Dietmar.

[hixx]

Hi Dietmar,
Dark Saturation would be the parameter to control the fainter area's saturation...
By default it is set very modest to mitigate Color noise of the background, this is why you got your first result

However there are more parameters to play with:
- try the "Matrix" correction of your camera model, it usually provides better hue and saturation.
- try the various presets of the SuperStructure module which allow you to enhance the large nebula in brightness and saturation without affecting and push back stars and background

Be sure to check the "Inoffizielle Deutsche Anleitung" available on download. It includes a lot of tips.

regards,
jochen

[decay]

Hi Jochen,

thanks for your reply!

I already pushed Dark Saturation to "Full", but it effected only slightly brighter areas, not the dark areas shown in the cutouts.

Concerning the other mentioned parameters:
- I already tried the matrix correction, but my camera model EOS 2000D is not listed. I tried models, from which I thought they would have a similar sensor and yes, there was _much_ more colour, so that I had to reduce the "Saturation amount" - but again, after that, no reasonable colour in faint areas.
- I am aware of the "Super Structure" module and I used it a lot of times to push back noise in the background field - really, this module does an incredible job . But in this case I intentionally omitted this module because I noticed the missing colours in areas from which I know, that there is colour information available. I would like to understand the cause of this problem and I felt like doing "cosmetics" with the "Super Structure" module - and not solving the underlying problem.

The example processing Ivo did now gives me the opportunity to analyse what went wrong in my processing. I will do that and then post the results here. As mentioned, it's not to "rescue" this particular image - I surely will take a better one with reasonable flats. But I want to understand the problem.

Thank you very much pointing out the "Inoffizielle Deutsche Anleitung". I saw it, but I thought it only would contain the same information available as well in the English resources - just translated. Of course, now I will take a look!

Best regards, Dietmar.

[decay]

Hi Ivo,

I checked the settings in DSS - and of course - you are right. I failed to set "No White Balance Processing" in the RAW File Settings and therefore the dataset already was balanced. Shame over me. For sure, I read the "Recommended Deep Sky Stacker settings" several times - no idea what went wrong missing that click .

But I am still not completely convinced that the colour balancing is the main problem. With the help of your example processing I now managed to get a similar result like yours. Now I will check step by step what makes the difference. I'm quite curious about that and I will report back as soon as I figured out something.

Best regards, Dietmar.