So, the first lens I tried on my new (to me) M8 is the Zeiss C Sonnar 50mm f1.5. It is an absolutely wonderful lens and there are plenty of blogs and sites on the web that will attest to that, so I won't dwell on it's qualities. I have only heard of one site that mentions the one significant issue this lens has: focus shift. This means that at wide apertures the plane of best focus is in front of the target indicated by the rangefinder patch.
I have done some tests using a template from Bob Atkins and I came up with some conclusions. I have shot dozens of frames doing focus bracketing every centimeter but for the sake of brevity here are the final results for the tests at 1 meter distance:
First at f 1.5, the first picture is focused on the center (un-numbered) vertical line, second one is focused on the line 5 cm behind it.
Since the depth of field at this aperture and distance is only 3 cm, if you focus exactly on your target, it will be entirely outside the lens' depth of field!
The situation at f2 is not much better. The first picture is focused on the center (un-numbered) vertical line, the second one is focused on the line 4 cm behind it. At f2 the DOF is 4 cm, so your subject will still be out of focus.
At f2.8 and up the focus accuracy improves dramatically to the point where you can focus normally using the rangefinder patch. I won't show any samples for that.
I've also done the same tests at 4m and the results are similar. In the pictures below taken at f1.5 and f2 the focus point was the small bolt circled in red. Again you can see it is outside the DOF.
So, what is one to do? You could come up with an index card of distances to back-focus at every aperture and distance and use that for every shot, but that would sap the living joy out of using this otherwise excellent lens. You could avoid shooting below f2.8, which would be an unspeakable waste.
Fortunately I happen to really like the way this lens renders at f2, so I kept thinking there must be a better way, and I believe it is this: I noticed that the amount of back-focus required was approximately such that the double image in the RF patch is superimposed offset by about the same amount as the width of the 50mm frame line.
One way to do this is: focus on your subject then turn the ring slightly toward infinity until the patch defocuses by on frame line. As you get better, or with moving subjects, you can just focus offset directly.
This works quite well at f1.5, for f2 use a little less than one frame line width. Most importantly, this rule of thumb works quite well at both close (1m) and medium (3-5)m distances. I haven't noticed any issues at distances beyond that.
Here are the four shots of the wire fence focused exactly that way, at f1.5 and f2, focused on the bolt in the pictures in red and focused offset by one frameline width without the red circle:
F1.5
F2
....and F2.8 for comparison, focused on bolt:
Really though, the best way to fix this might be to use the Summilux 50mm ASPH, and it may come to that in the future, but for now I like this lens enough to give it a second chance.
In case you were wondering, my M8's previous owner (since new) assured me that the rangefinder was untouched from the factory and the focus was spot-on on his Noctilux, so I have no reason to believe it's an issue with the camera's rangefinder accuracy and not the lens.
No comments:
Post a Comment