All sensors will exhibit some visible artifacts if you look closely enough. The vertical pattern visible in Blueman's master flat is typical of KAF-8300 sensors, but more importantly it's nothing to worry about. The variability between the light dark bands is likely on the order of 0.1% of the signal level. The key is that the artifact will be consistent at any given temperature and exposure time (or signal level for flats) and will therefore calibrate out using the standard image calibration techniques, which were developed specifically to deal with the real-world issues of digital image sensors.
I've attached a roughly equivalent master flat made from 30 flats and 30 matching flat darks from a QSI 683 at -10C. You can also download the FITS file from this link:
You must rememeber that the chip is a whole lot of columns of pixels and they are all separated by spaces so there is always a pattern to an imaging chip. Some show this more than others. But this is pattern noise that must be removed if you wish to have a properly calibrated image.
There can be other types too. Some chips will have polishing marks that show up as arcs or lines. Plus the chip does not have an absolutely equal response, so there will be brighter areas and darker ones.
Quote: Both my FLI and QSI show similar patterns in the flats. Isn't that actually caused by imperfections on the chip itself?
more likely these patterns are the result of chip structures or fabrication issues that affect the QE of certain columns differently than others. it doesn't make much sense to talk of spaces between the capacitors as being part of the final image. two columns may be separated by some distance on the chip but the resultant pixels for those columns are right next to one another in the final image.