Resampling sharp Foveon images

Yeah - and this is one of the problems with discussions like this. Different people will evaluate the images differently. Some will not see any (existing) differences at all and some might even invent differences that are not there. Thats why the best thing you can do when claiming you dont see any degradation is to publish the images. Then anyone can draw their conclusions.

Yes - this is very important to understand. With practice you get better. This is also true for seeing image quality.

BTW - I still think you shall do the rotate back and forth experiment. It is much more easy to compare images that shall look the same. And faults accumulate, so two distortions should be more visible than one. Rotate one degree and minus one degree. And be very sure that the original is VERY sharp.

--
/Roland
X3F tools:
http://www.proxel.se/x3f.html
https://github.com/rolkar/x3f