Weird banding artefacts in mosaic mode only

Greetings to the APP team and forum members,

first time poster but have been using the program consistently for two years and I absolutely love it. I am currently at a stage where I am really enjoying the quality that the software brings out of my images and allows me to plan imaging sessions accordingly so my heartfelt congratulation to the team.

I have unfortunately reached an issue that I have not found recorded either in the APP forum or off the internet related to the camera I use which has the imx 183 sensor notorious for banding issues.

These banding artefacts look like transparent vertical trails and produce an embossed effect occurring at a regular interval over the mosaic in parallel lines. At first glance its layout seems consistent with the type of banding you get on those cameras. On component panels of the mosaic they appear almost at a 90 degrees angle since the camera is positioned vertically on my rig. What is annoying is that you can clearly see them at 100 percent view and it spoils the features of the image. I have seen the grid pattern typical of CMOS sensors before, but nothing like that on a processed image. What is peculiar is that the banding is on top of the stars and features of the image unlike the regular banding you’d find in the background of an image.

What is more baffling is that these trails are not present on any of the mosaic components before I assemble them in a completed mosaic.

This is frustrating as the mosaic is absolutely seemless so far as I can see even although the overlap is minimal.

For context I am doing mono RGB images from an imx 183 mono sensor for my stacked panel components with an average of 30 hours per stack. Subs for Red and Green channels are 300s and the blue one only 60s to avoid too much bloating. Flats and Darks are used but no bias nor flat darks. Dither is 5 pixels. Each panel has a wide variety of exposure time on them but again these artefacts are not present on any of the individual panels.

I have tried a variety of different settings in the star analysis, registration and normalize tabs. I have not tampered with the default settings of the calibration tab.

I noticed the issue with version 2.0.0-beta29 version which I then upgraded to the latest 38 version with similar results. This is on a win 11 machine eight core ryzen 7 with the app using 15 threads, 64gigs of ram and the work directory is on a half-full SATA SSD which is not the drive where the OS is installed (I also tried to process the mosaic on the C: M2 SSD drive which made no difference). I have also tried disabling GL4 and relying on the CPU only for all tasks to no avail.

Any advice would be much appreciated. I know troubleshooting issues takes a lot of your time so I am grateful for any insight you might offer.

Thomas