Concentric Banding in Stacked Images

[MarkS]

After a lot of detective work and false leads, I've finally traced the cause of a problem I've found in all my recent processing - concentric circular bands in the centre of the image and other banding striations elsewhere in the image.  This is why it has taken so long to process my September images.

Here's an example of the concentric circular banding (I've deliberately oversaturated and overstretched the image to make it more visible):


If you still can't see it, this should indicate where to look:


Sony A7S images are very noise-free and (without a DSLR mirror box) are relatively gradient free that I can stretch the images far more than the Canon.  With the Canon, the background patterning (aka Canon banding) used to be the main limitation. The enhanced data stretching I can now perform is revealing a processing artefact caused by the flat frame calibration in IRIS.  The flat frame itself is perfect but IRIS does its calculations in integers.  So it divides the light frame by the flat frame then truncates the result to integer before saving as a temporary file.  This truncation is causing a very subtle posterisation of the background sky which then shows up after stacking the resulting images followed by background subtraction and extreme stretching.

So I've now reached the end of the road with IRIS.  I now need to take the hit of converting my image calibration sequence fully to PixInsight.   In the long run I'm sure that will be a good thing.

Mark

[JohnP]

Mark - only you could have found that problem... Will be looking forward to your results using PI. John

[Fay]

i would not have noticed it if you had not pointed it out, Mark