Issues With Pixinsight Local Normalization Using WBPP

[Juno16]

Hi Everyone,

I just read a thread on CN concerning issues with using Local Normalization when using WBPP with certain datasets (not sure if the issue still exists with the new WBPP script update yesterday - 2.4.5).

The last two images that I processed in WBPP using LN (with defaults) produced wild color gradients seen in the stretched masterlight. The same issue is described in this thread. https://www.cloudynights.com/topic/8269 ... rocessing/

I had this problem for the second image in a row and ended up dropping back several versions of PI to try to troubleshoot. I ended going back to v1.8.9 and just deselecting LN in WBPP to get good results.

Just want to pass this along in case anyone else was having issues.

[starfield]

Hey Jim, I wonder if it's a problem with the reference frames that got selected. Try the new version, it lets you decide which frames to use. Also might want to try using PSF SNR as weighting and see if that helps. FYI, Adam Block just released some new videos explaining the new weighting schemes.

[Juno16]

Hey Jim, I wonder if it's a problem with the reference frames that got selected. Try the new version, it lets you decide which frames to use. Also might want to try using PSF SNR as weighting and see if that helps. FYI, Adam Block just released some new videos explaining the new weighting schemes.

Hi Steve,

Thanks for responding and the suggestions!

I have been wearing out my pc running WBPP with different settings. I appreciate the suggestions.

I tried some things, but I just realized that the Reference Frame Generation was set to Best single frame. That must have been a leftover from the previous version.

I set it to "Integration of Best Frames" and set the Normalization Scale to 1024 as opposed to the default 512 (as suggested by Juan Conejero of PI staff) just as a test. The resulting Masterlight looked very good.

I am running a test now with the same settings except for the Normalizaion Scale set to the default of 512 to see if the fix was (as you mentioned) setting the Reference Frame Generation to "Integration of best frames".

I will post the results.

Thanks Steve!

[Juno16]

Just finished WBPP. Went back to a Normalization scale of 512. Looks like the new setting in WBPP in 2.4.5 with the selection for Reference frame generation "Integration of best frames" did the trick. Excellent results!

[STEVE333]

Just finished WBPP. Went back to a Normalization scale of 512. Looks like the new setting in WBPP in 2.4.5 with the selection for Reference frame generation "Integration of best frames" did the trick. Excellent results!

Capture.PNG

Capture.PNG

Thanks for sharing your efforts on this Jim. I've never tried Local Normalization and will give it a try on my next data.

Steve