Local Normalisation: unable to find a local normalization reference frame

Dane

Dane

Well-known member
Hi,

with a stack I always get the error message “unable to find a local normalization reference frame”, and I don't know what to change. I have increased the grid size from 4 to 10 in the local normalization settings, which unfortunately does not help.

2025-03-10 13_06_35-WBPP Execution Monitor.png


What could be the cause and how can I fix it?
 
You probably have selected reuse the LN frame as a reference, in the lights tab. And as this one is not there, it does not find it. It should be enough to uncheck that checkbox.
 
Thanks for the hint, but the same problem still exists.


1741625874378.png


I uncheck the box "Reuse last reference frames"

1741625911176.png


Or did you mean a different setting?
 
Just to rule one known issue out of the equation, is the directory in which you are processing subject to any form of file replication I.e. TimeMachine, One Drive etc?
 
I use Backblaze to backup my drives to a online service. Could this be a problem?
 
It can be a problem if that online backup disk is used as a working disk, PI does not like that.

If your working disk is local, it can be internal or external, there should be no problem. Do keep in mind that PI also does not like the use of strange/foreign characters in file names or spaces at the end of folder names.
 
Ok, a new attempt. I have removed all spaces from the directory names. And I have stopped the backup service. I have completely refilled WBPP with the data, and a new run is currently in progress.

The strange thing is that I always use the same way to create the directories, and the backup service is always running, and so far there has never been a problem with local normalization. :unsure:
 
Ok, the problem still exists, unfortunately the points were probably not the cause.

1741638055879.png
 
Dane said:
Ok! I have attached the log file here.
Click to expand...
The failure can be caused by a few things, very large stellar profiles, bad calibration, and more. Can you upload to a cloud host one of the images that is failing? For instance, F:/Fotos/Astro/NGC4565_Nadelgalaxie/Stack_2025/registered/Light_BIN-1_6072x4042_EXPOSURE-300.00s_FILTER-NoFilter_RGB/Light_NGC 4565_300.0s_Bin1_2400MC_gain158_20250305-021544_-10.0C_0072_c_d_r.xisf
 
I keep this in mind with the cosmic correction setting. Do you have a recommendation for this setting, so that im not going to high?

Its really strange. I do some other stacks with data from another object, i didnt have this problem there.

When i understand it right, the normalization process is for flatten the image.

This is the result after WBPP, and there is a gradient. Maybe this could be a problem for the normalization process?

2025-03-11 13_19_14-PixInsight.jpg
 
I don't think that is a "gradient". The radial pattern suggests imperfect flat correction. This most often happens if the flats don't match the lights, or if dark calibration (of either the flats or the lights) is incorrect.
 
I also think that it comes from the flats, but can usually be smoothed out with gradient removal without any problems. But could something like this influence the local normalization process?

It's really strange, I've stacked a lot of things, but I've never had this problem before.
 
Hi @Dane

This is obviously a very nice image but it is true that it does not seem very well calibrated with some hot pixels remaining a a bad flat field correction. But I find it strange that these problems explain the error with LocalNormalization?

First I think you should update to Pix last version (1.9.3)
Then can you try to reset every setting to factory default in WBPP and try again (and share the log if it still fails).
 
Dane said:
I keep this in mind with the cosmic correction setting. Do you have a recommendation for this setting, so that im not going to high?

Its really strange. I do some other stacks with data from another object, i didnt have this problem there.

When i understand it right, the normalization process is for flatten the image.

This is the result after WBPP, and there is a gradient. Maybe this could be a problem for the normalization process?

View attachment 26537
Click to expand...
I only make note of cosmetic correction because there are still quite a few hot pixels in the image you provided. There's no way to suggest a specific setting; you just need to lower the sigma value until they are gone. Also, ensure that your master dark is of good quality if you're using the automatic CC built into WBPP calibration, because that's what's used to isolate hot pixels.
 
Thanks, I have also seen the dots. I have reduced the Cosmic Correction from 10 to 4. unfortunately without success. Maybe i should try to set it to 1.

I am currently downloading the latest PI version.
 
Dane said:
Thanks, I have also seen the dots. I have reduced the Cosmic Correction from 10 to 4. unfortunately without success. Maybe i should try to set it to 1.

I am currently downloading the latest PI version.
Click to expand...
This could indicate a problem with your master dark. You should also try using the regular CosmeticCorrection process, which applies a different statistical approach.
 
You must log in or register to reply here.
Back
Top