Hiep P writes:
Thank you for the thorough review on the Otus 55. It is an outstanding write-up.
I would have pre-ordered the lens but because of this very precise reason you brought up, no EVF option on DSLRs, that I have decided not to. I am currently shooting with the Sony A99 and my two Contax 55/1.2 and 85/1.2. Since these two are even less corrected for aberrations, precise focus is paramount to get a good image. So I understand your frustration. For me, Live View is a must for any critical work (certainly so since my main focus is on landscape and urbanscape). And only with EVF, such implementation is seamless (no awkward framing with the LCD).
[diglloyd: please pre-order using Zeiss lenses ordering page]
What I am contemplating right now is whether to get the Nikon version or the Canon version. With Nikon, I have to deal with the opposite focus direction and just do a Leitax conversion to A mount. With Canon, I can retain the ”right” direction, use a Metabones EF-E adapter to mount on the upcoming FF NEX.
After reading Roger’s articles on lens skew from using adapter and your articles on lens-camera alignment issue on Canon and Nikon DSLR, it gets even more complicated. Definitely going native would allow me to send the camera in for a proper alignment (or so we hope). David at Leitax would certainly make a very precise mount replacement. But without a MTF tester, I am not sure I can get a similar quality as the original lens. Going for the Otus means that one want the ultimate performance. There is no point to pay this much and to haul with much weight and size to have anything less than that (I can certainly get that with my smaller, lighter 55/1.2).
Since you’re savvy on this matter, I would like to hear your opinion. Thank you.
DIGLLOYD: Precise focus is critical to extract the best from a very high performance lens. This is where a DSLR optical viewfinder is a dismal failure and one reason I strongly desire an EVF.
In general I have had good luck with lens adapter over the years, but most of that has historically been with lower resolution cameras. With high-res cameras (24 and particularly 36 megapixels), small issues show up much more readily.
The issues are complex and I will detail just the major ones here:
- Lens mount to sensor parallelism is not guaranteed. Thus even without a lens there are issues starting with the camera body itself, see the link noted above. It is one reason that MTF lens tests are suspect: how was the camera measured for perfect flange to sensor parallelism?
- Lens mounts can and do bend and warp if not used with care. Ditto for lens adapters. Amounts invisible to the eye become optical problems (e.g., 10 or 20 microns). Consider what it might take for a lens adapter vendor to guarantee 10 micron or tighter planar tolerances on a lens adapter (“my $39 lens adapter works great”—yeah sure it does).
- Large and heavy lenses like the Otus 55/1.4 APO-Distagon should not be used with an adapter without supporting the lens somehow (e.g. with a hand under it). Lens mounts experience a lever-arm torque which can ultimately bend the lens mount, a very expensive repair. Not to mention “sag” while shooting. [Kudos to Leica for detailing weight limits in their R to M adapter brochure. And the Leica adapter is as robust an adapter as you will find.]
- It is the rare lens that shows perfect symmetry left to right, and the likelihood of retaining symmetry as the elements move through the focusing range is slim to none for both manual and autofocus lenses. The Otus 55/1.4 is held to very high tolerances (to some extent the price contemplates a higher rejection rate), but this cannot be considered a guarantee of symmetry throughout its focusing range. Even Leica S lenses show symmetry issues.
- The 55/1.4 APO-Distagon revealed a small problem with my Nikon D800E flange-to-sensor alignment back in the spring; Nikon gave my D800E special attention at my request. It was probably already “in spec” however. But it improved.
That’s a “sampler”. Bottom line is that when a lens adapter is inserted into the equation it is one more variable: one that can worsen an asymmetry or work to mitigate it! And by “left/right” I really mean any asymmetry in which the relevant geometric planes are not plane parallel, and that can mean through two axes.
As for Nikon versus Canon mount, a lens in Nikon mount is definitely more flexible in the range of cameras to which it can be adapted. But given the size and weight of the Otus 55/1.4, it is probably best to consider exactly which lens adapter is to be used, e.g. what the total system will look like (camera + adapter + lens), since the choices could vary in quality or availability. I’d want a very robust adapter for a heavy lens. An adapter for a Nikon lens will be 2mm thicker, and that could in theory work in its favor, depending.
In the end one just has to try, and also accept that things might not be perfect. Even my brand-new Leica 50mm f/2 APO ASPH on a brand-new Leica M240 showed lens skew! It’s one of those facts of life. If Leica can’t guarantee perfection, then perhaps no one can except for those $25K cine lenses.
All that said, remember that the lens is projecting an image that is sharp (assuming it is performing properly). The targets that show a problem are those in which the subject matter is planar: a wall or mosaic, a distant landscape, a group of people lined up, and so on. Otherwise, one is not so likely to notice.