[BasilG]

Indeed, it should be more like "until diffraction really starts to limit sharpness". And the sharpening in the PZ MTF workflow does mask the effect  (the amount) a bit.

Lenstip shows diffraction softening a bit more clearly because of that:

3880_roz_centr.jpg

 

"Diffraction kicks in at f/5.6, causing sharpness to drop to "very good" levels by f/11".

 

Whatever. I brought this up before. What the reviewers mean to say is that by f/X, diffraction has reduced sharpness by one "school mark", usually excellent to very good. I know it's a really minor point, but I wonder why, if this is what is meant, something else is actually stated (specifically, that diffraction effects become visible at f/X - when clearly diffraction effects are visible in the test data several f-stops before).

 

This may seem minor, but in fact, a lens for which diffraction "kicks in" only at f/11 is usually not very good. Having the center diffraction-limited at f/5.6 is a good sign (f/4 even better). So there's actually something to be learned from the data that is currently not paid much attention to. Anyway, I'll stop now.