Originally posted by boriscleto According to DxO, yes. But the results are always suspect with DxO. The lil'Bigma rates higher on the K10D than the K20D...
I completely agree - use caution regarding the precise DxO numbering.
Also, the references in the earlier post regarding DxO were related to overall quality - not the sharpness setting. When a lens is rated at 6 mp at its sharpest setting, that could mean 5.55 mp on the K5iis and 6.45 mp on the K3 - but show up as simply 6 mp due to rounding. Also, keep in mind, that is the rating for one arbitrary focal length and one aperture setting. Many of the better Pentax lenses look relatively weak when looking these numerical ratings because their "best" setting is not terribly high, but a wide variety of other settings are strong (focal length and aperture combinations). And then there is contrast - a Pentax strong suite not measured by DxO.
---------- Post added 01-01-2016 at 02:51 PM ----------
Originally posted by blende8 Here is an image from a lenstest of mine.
At the number 4 (red arrow) it appears that the orientation of the bars is changing. Here the frequency is so high that the sensor cannot cope with it anymore correctly and artificial, "spurious" patterns appear. I think this has to do with the Nyquist theorem.
When you blow things up to 400x, it can be any number of factors involved in the muddled mess you have posted. However, it looks like simple CA for the most part. It almost certainly has nothing to do with the Nyquist theorem, or any other theory related to the idea of sensor out-resolving the lens. It is possible that the sensor itself has some noise issues... but that would be separate from the Nyquist Limit.