M33 Outer Dust Lanes vs. Sky Background - Forums

Topic starter

Sep 21, 2020

Dear community,

yesterday I acquired the second image with my new EQ6-R mount and chose M33 as target. I even got up at 2:50 in the morning to do a manual meridian flip, when the scope was only 10 minutes away from hitting the tripod. In the end, I think it was worth it and I'm really happy with the result I got, especially the detail and star shapes (except in the extreme corners).

However, during processing I noticed that the outer dust lanes appear darker than the sky background. That is something I couldn't find in other M33 images on Astrobin, at least not in the maybe 25 images I looked at. Is that something you would consider to be real or might it be a processing artifact. I mean, actually it makes sense as the dust would of course absorb the background light. I didn't really touch the image a lot as the data was already pretty good. For example, I didn't use any masks to adjust certain aspects of the image. So, I tend to think it is real, as the darker patches all match with the slightly brownish dust patches.

What is your opinion? Did someone else notice the same?

Helpful Insightful Respectful Engaging

Steve Ludwig

Topic starter

Sep 21, 2020

Small addition: I just checked the background model image that was subtracted in the background extraction process. It looks normal, as you would expect it. It contains only the light pollution with vignetting towards the corners. No further structures in there. So, I can confirm that it is not a background extraction artifact.

Well Written Helpful Concise

Michael Timm

Sep 21, 2020 · Edited: Sep 21, 2020

Hi Steve, this effect is well known to me. It happens in two situations in my experience: When using the AutomatedBackgroundExtraction Process (which is always a bad idea) instead of the dynamic one. Or when choosing background points containing stars or other foreground information in DynamicBackgroundExtraction. Could one of these points be the cause? CS Michael

Edit: Just saw in your description that you used ABE. This is for sure the cause. You should use this process only for a first check if the image is okay, not for your final processing.

Helpful

Steve Ludwig

Topic starter

Sep 21, 2020

Michael Timm:
Hi Steve, this effect is well known to me. It happens in two situations in my experience: When using the AutomatedBackgroundExtraction Process (which is always a bad idea) instead of the dynamic one. Or when choosing background points containing stars or other foreground information in DynamicBackgroundExtraction. Could one of these points be the cause? CS MichaelEdit: Just saw in your description that you used ABE. This is for sure the cause. You should use this process only for a first check if the image is okay, not for your final processing.

Thank you for the hint. I normally use DBE as well, but lately I didn't had the time to do the time consuming manual sample placement. I will give it a try later today.

Michael Timm

Sep 21, 2020

Good luck and CS to you!

Leonardo Landi

Sep 21, 2020

ABE artifact for sure. Have a look at Visible Dark's tutorial about easy DBE. The result is quite good!

Douwe79

Sep 23, 2020

I never use ABE for exactly this reason

Steve Ludwig

Topic starter

Sep 23, 2020 · Edited: Sep 23, 2020

Hi,

you were all right, it was an artifact from the "automatic background extraction" process. I did another revision, carefully using the "dynamic background extraction process" which led to a much better result. I also tweak the color balance a bit.

Many thanks for your help identifying the root cause!

Best Regards

Steve

Well Written Respectful Concise Supportive

Michael Timm

Sep 23, 2020

Steve, glad you could solve the problem. The image looks good!

Well Written Respectful

Douwe79

Sep 24, 2020

Indeed, looking good!

Wes Schwarz

Oct 7, 2020

Steve Ludwig:
Hi,you were all right, it was an artifact from the "automatic background extraction" process. I did another revision, carefully using the "dynamic background extraction process" which led to a much better result. I also tweak the color balance a bit.

Many thanks for your help identifying the root cause!

Best Regards

Steve

i have to disagree. The originsl image looked very natural. You have now just darkened the image so the surrounding structures cannot be seen. The first images had very slight background extraction artifacts in the darkest areas. The rest is natural dust in this heavily ifn laden region of the sky. IMO blackening the image out as you have done is a common amateur adjustment, I think it is common because it also makes it easy to hide other artifacts such as vignetting, ampglow, and noise from short integrations. Your first image was almost right on the money, if anything it just needed a more subtle pass with abe or dbe.

Helpful Concise

Steve Ludwig

Topic starter

Oct 7, 2020

Wes Schwarz:
Steve Ludwig:
Hi,you were all right, it was an artifact from the "automatic background extraction" process. I did another revision, carefully using the "dynamic background extraction process" which led to a much better result. I also tweak the color balance a bit.

Many thanks for your help identifying the root cause!

Best Regards

Steve

i have to disagree. The originsl image looked very natural. You have now just darkened the image so the surrounding structures cannot be seen. The first images had very slight background extraction artifacts in the darkest areas. The rest is natural dust in this heavily ifn laden region of the sky. IMO blackening the image out as you have done is a common amateur adjustment, I think it is common because it also makes it easy to hide other artifacts such as vignetting, ampglow, and noise from short integrations. Your first image was almost right on the money, if anything it just needed a more subtle pass with abe or dbe.

Hi Wes,

I didn't just darken it. I started to process it from the beginning, (carefully) using the DBE module this time. Feel free to brighten the image a bit to see the difference. The are no gradients towards the edges of the image anymore. I think I can also confirm that the things I saw in the first image revision really came from the ABE process. As soon as I placed the samples manually in the dark regions only (without including some stars like in the ABE process), I got rid of the gradients. However, in the end it sometimes is very difficult to "know" where to place the DBE samples. Dependent on the choice, the image will look different afterwards. So, I think in the end its more about personal taste.

Helpful Insightful Respectful

Leonardo Landi

Dec 27, 2021

ABE artifact. Wrong function degree that's for sure. ABE is an exceptional tool, but you have to choose the right function degree.
0 if you have an uniform gradient
1 for a linear gradient
2 for circular gradients.