Banding in Images Processed in PixInsight

I recently experienced strange banding in stacked images from PixInsight.  I checked the initial images as well as the Flats, Bias & Darks and did not see anything that would account for this in my final stacked images.  I checked the Forums and ran across something called “Canon Banding Reduction” or CBR.  This sounded like it might explain, although my banding was vertical and CBR sounding like it was horizontal.  I ran the script on my images as well as rotated 90 degrees and it did not help.  

In looking through the files generated, they seem to show up in the Registered files from WBPP.  I re-ran WBPP with no Bias, Darks or Flats and the banding was still in the registerd files, not in my images as far as I can see.

I then tried unchecking the Linear Defects Correction, and the registered files showed NO banding.  I then re-ran WBPP with my Bias, Flats and Darks with Linear Defects Correction un- checked and the problem seems to go away entirely in my Narrow Band Nebula only image and in my RGB image, from the night before, reduced greatly.  So, there be be some very fine anomalies in my image.

Is there any explanation that you can think of that would explain why unchecking the Linear Defects would work?  Almost seems like unchecking this box may make be prone for some sort of defects…

Camera was ZWO ASI 2600MM Pro and I was using a RedCat 51 as well as an ASI Air and the mount was ZWO AM5.  The images where under-sampled and I had dithered and applied Drizzle Integration.

The attached image shows one Registered file with the version on the left having Linear Defects Checked and image on right unchecked.  The stacked images enhance the banding.

Any thoughts as to why checking seems to provide worse results in this image?  Any thoughts or suggestions?

Thank you,

Jim B