Image with severe gradient

[gianpri]

HI,
I have a MasterLight with serious gradient problems. I used the new GC process but with disappointing results. From the comparison with DBE the difference is evident. In GC I tried to vary the parameters (with/without: automatic convergence, simplified model, structure protection) but without success.
I'm sure I'm doing something wrong!
I attach the link to the MasterLight.

masterLight_BIN-1_4164x2795_EXPOSURE-300.00s_FILTER-L-Pro_RGB_autocrop.xisf

drive.google.com

 

[nico1038]

HI,
I have a MasterLight with serious gradient problems. I used the new GC process but with disappointing results. From the comparison with DBE the difference is evident. In GC I tried to vary the parameters (with/without: automatic convergence, simplified model, structure protection) but without success.
I'm sure I'm doing something wrong!
I attach the link to the MasterLight.

masterLight_BIN-1_4164x2795_EXPOSURE-300.00s_FILTER-L-Pro_RGB_autocrop.xisf

drive.google.com

Hello @gianpri

Your googledrive link is not public

 

[gianpri]

Sorry

masterLight_BIN-1_4164x2795_EXPOSURE-300.00s_FILTER-L-Pro_RGB_autocrop.xisf

drive.google.com

 

[nico1038]

I agree it is a difficult image.
The best I got is with the following setting (note that I disabled both "Simplified Model" and "Structure Protection" because I couldn't get them to work well)

With these settings I think the background is ok but the nebulae is clearly over corrected.
I'm curious to see what others can get.

 

[vicent_peris]

Hi,

This is indeed a very difficult case for GradientCorrection. You need very exotic parameter values to get a reasonable result. For this image, MARS works very well. I'm attaching the image corrected with MARS user-contributed data.

Please be patient, we are working hard to get MARS publicly available as soon as possible.


Best regards,
Vicent.

 

[fredvanner]

Looking at your screenshot, I wonder if the "GC" image has not had the STF re-applied after GC (so is being displayed with the pre-GC calculated STF), which does not give a "like for like" compare with the DBE image. With these settings:

I get:

which compares reasonably well with DBE.
The combination of a strong left-to-right gradient (looks like moonglow) with a separate near-radial gradient in the bottom left hand corner (looks like LP), certainly makes this a difficult one to fix.
BTW, I ran SPCC first (to make sure the colour balance was roughly right). The WB plots suggest that this image may have been captured with a dual band filter, which will limit the colour calibration.

 

[gianpri]

Thanks Nico for trying.
Thanks Vincent, excellent result. We await the release of MARS with interest.
Fred, I only applied STF after GC, as I did with DBE.
You're right: the strong gradient is due to the presence of the full moon and the IL of my site on the outskirts of Milan Italy. The shot is with L-eNhance filter.
Thanks for the tips with GC.
John

 

[gianpri]

In the end something succeeded. I used DBE!