I've been curious about the reports of focus shifting with the Ultron f/2 since I first thought to buy one. The M-P typ 240 finally provides an ideal platform to do some real testing that isn't ferociously time consuming. So I did a set of test exposures with the Ultron 28: at a focus distance of approximately 7 feet ...
Originally Posted by HiredArm
Set 1: focus with the RF, make exposures at that focus setting at f/2 to f/11 aperture settings.
Set 2: focus with the EVF/LCD (focus peaking off), and readjust for each aperture from f/2 to f/11
Set 3: focus with the EVF/LCD (focus peaking on), and readjust for each aperture from f/2 to f/11
I repeated the full set of test exposures four times. What I found is that
a) There is variation in the focus between what the RF and the EVF tell me is "in focus", and that the variation is not always better for the EVF than it is for the RF. This is probably down to the vagaries of my eyes' ability; I don't have eyes that perform perfectly either. ;-)
b) The focus variation is mostly visible at 2:1 magnification. Drop down to 1:1 and it is very difficult to see in most cases, excluding obvious missed-focus cases.
c) Focus peaking nets a slight improvement in focusing speed but little improvement in focusing accuracy. It works better in the field hand-held than it does for tripod-mounted critical focusing efforts.
I don't see much on this example of the lens that says it focus shifts by much, although the overall sharpness between f/2 and f/4 improves so much by f/4 you'd think it's a different lens.
Regardless, the Ultron 28 has been a source of consternation to me since I bought it. Images made with it show it to be a good performer, but it doesn't inspire me much for some reason. The Color Skopar 28mm f/3.5 is more inspiring and produces results I like more, despite the need to do corrections for color shifting.
I'll use it for a while yet. "Equipment is transitory ..." et cetera. ;-)