mapgraphs Offline Upload & Sell: Off
|
It's difficult to draw much in the way of meaningful conclusions when comparing a 46mp (8256x5504) sensor with a 23.6mp (5952x3976) sensor, without scaling the higher resolution image down to the lower rez image. Ideally one would want to compare 1:1. The higher rez sensor is also going to be less forgiving with small mis-focus issues. I've seen a lot of Sony samples claiming that this or that lens didn't work on one Sony body or another which were off simply because the point of focus was just beyond infinity - I can recognize this sometimes because I've done the same thing. It is very easy to do, and when done the edges and corners go south - they look bad, no matter what the f/ stop is.
Also one might want to keep in mind that Leica cuts off 24pix on each side and 12pix top and bottom, 5952x3976 on the M (240) and does a similar thing on other models, 5952x3992 on the M10, 5952x3968 on the M-262 etc. (the SL is 6000x4000). So, cross comparisons should probably take these into consideration as well. First fut off the offending 24pix on each side... ; - )
Sony comparisons are problematic because it may be cover glass thickness out of spec, not simply the thickness itself. If some of Roger's asides in his tear-down of the Z7 are a clue, the whole sensor assembly, stack and mount registration may come into play - something Nikon appears to have deliberately designed to avoid (the take-away from Roger's article, imho). We know there are mount registration variations between (apparently all) Sony bodies that affect native lenses ability to hit infinity so, there may be some clues in that too. Sony initially released the a7 and A7R with plastic mounts so it would seem they started out without much concern about tolerances variations... (imho). In contrast, I don't think there is going to be much variation in Leica's manufacturing tolerances.
At any rate, more samples are better. Hopefully my Kipon M to Z arrives soon. In the mean time, the 24-70 s is a gem! ; - )
|