Has anyone take the plunge on the Zeis 50?

[Mark_Swindler]

Sorry, here are the photos-Zeiss, then Leica

[diglloyd]

Originally Posted by Mark_Swindler

These next two shots are cropped from the middle of both images.
First is the Leica, then the Zeiss:
I don't know if you can tell, but the Leica completely blows away the Zeiss! I expected it to be closer, really, but even at f8 the Leica glass is amazing, and really doesn't get much softer at f1.4.
What really surprised me was that the cheap Nikkor 1.8 beat the Zeiss at f8 and wide open! This leads me to believe that I may have a lemon-what do you think?

FYI, I've tested two copies of the 50mm f/1.8 Nikkor (and six (6) manual focus variations of the 50mm Nikkors), and all were notably inferior to my 50mm/f1.4D at all apertures, the critical difference being a big difference in achieved depth of field in real shots when stopped down, but with the 50mm f/1.4D obviously beating them all at f1.8 as well. The advantage of doing "3D" shots is that you can detect focus error much more easily.

Sample variation/assembly error? I don't know, but the difference was obvious, and repeated tests to confirm the results showed the same thing.

I would recommend repeating your test to check focus error. For flat subjects, extremely precise focus is mandatory, and curvature of field can also come into play. Just try it first with one lens, and see for yourself what an apparently trivial adjustment can do. Don't count on autofocus to be anything near accurate enough to deliver optimal optical performance (sometimes it does, often it doesn't). That's why I sometimes check focus accuracy too; it can obviate all the optical advantages of the very best lens.

Did you bracket focus (and I mean by *tiny* amounts, the barest minimum you can manage to achieve by manual focus)? If you didn't do so, the results are suspect; I have repeatedly confirmed this with many different lenses (Canon, Nikon, Leica). A 1/3mm (on the focusing ring) error can easily change the conclusions at f8 when comparing two lenses on a high-res camera like the D2x.

[Mark_Swindler]

Lloyd;
Thanks for your reply, I tried the focus "bracketing" you suggested, and I could see some difference. After retesting for front or back focusing, and finding that to be accurate, I retested the lenses @f8, and f2.8, and included the 60mm micro as well.
This latest test showed the Zeiss to be sharper at f8 than the 50mm1.8 and a tiny bit sharper than the 60mm micro. This surprised me, as the 60mm is very sharp! I didn't have the Leica Summilux to include in the re-test, though.
At f2.8, the results were pretty much the same as the first test, i.e. the Nikkor 501.8 was surprisingly sharp, the 60 micro pretty much fell apart wide open(not surprising), and the Zeiss is just not very sharp. I even tried the 1.8 Nikkor at 1.8 compared to the Zeiss at f2, and the Nikkor still won out.!
I'm thinking a call to NPS to loan me a 50mm f1.4 should give a really fair comparison between Nikon's sharpest 50 and the Zeiss, but I suspect that the Nikkor will win out.
I'm thinking eventually to try the same test on the D70 to rule out a misalignment in the D2x body.
I'll continue to test and we'll see.
Mark S. (crosseyed!)

[diglloyd]

Originally Posted by Mark_Swindler

Lloyd;
Thanks for your reply, I tried the focus "bracketing" you suggested, and I could see some difference. After retesting for front or back focusing, and finding that to be accurate, I retested the lenses @f8, and f2.8, and included the 60mm micro as well.
This latest test showed the Zeiss to be sharper at f8 than the 50mm1.8 and a tiny bit sharper than the 60mm micro. This surprised me, as the 60mm is very sharp! I didn't have the Leica Summilux to include in the re-test, though.
At f2.8, the results were pretty much the same as the first test, i.e. the Nikkor 501.8 was surprisingly sharp, the 60 micro pretty much fell apart wide open(not surprising), and the Zeiss is just not very sharp. I even tried the 1.8 Nikkor at 1.8 compared to the Zeiss at f2, and the Nikkor still won out.!
I'm thinking a call to NPS to loan me a 50mm f1.4 should give a really fair comparison between Nikon's sharpest 50 and the Zeiss, but I suspect that the Nikkor will win out.
I'm thinking eventually to try the same test on the D70 to rule out a misalignment in the D2x body.
I'll continue to test and we'll see.
Mark S. (crosseyed!)

Well, testing lenses is VERY HARD. Far, far harder to do it well, but after quite some time it becomes easier! At least that's the way it was for me.

My 60mm/f2.8D is an outstanding performer, even wide open at close distances (1:10 or so). It does have a very, very slight curvature of field though.

That's one of the confounding factors; a lens that appears to have soft corners might in fact have exquisitely sharp corners except that curvature of field means they're out of focus! This can be detected by making minute focus adjustments; if you see a sharp center in one frame, but sharp corners and a soft center in another (nuts anyone?), then there is curvature of field at play.

[Mark_Swindler]

Lloyd;
You are absolutely right about lens testing being HARD! Much harder than I thought. My 60mm doesn't seem to be terribly sharp @2.8, but maybe it was either curvature or camera shake. I've found that it's much harder to maintain sharpness with digital than with film, especially compared to my Leica M's.

Mark

[diglloyd]

Originally Posted by Mark_Swindler

Lloyd;
You are absolutely right about lens testing being HARD! Much harder than I thought. My 60mm doesn't seem to be terribly sharp @2.8, but maybe it was either curvature or camera shake. I've found that it's much harder to maintain sharpness with digital than with film, especially compared to my Leica M's.

Mark

Camera shake should not be an issue up to 600mm or so, provided you use a solid tripod, and mirror lockup. Any results obtained without mirror lockup (using at least a 3 second pause from mirror-up to release, more for longer lenses) should be considered suspect. Obviously, a remote release must also be employed.

For that matter, several models of Gitzo tripods have unacceptably strong resonance which is clearly shown by careful testing. That is why mirror lockup with a pause is critical as the focal length increases. [My paid (not free) research article, The Sharpest Image delves into these issues with unequivocal results, one reason the only Gitzo tripod I'll use without concern is the G1325].

Even mirror lockup can be insufficient beyond 600mm. I was using a 500mm + 1.7X teleconverter combo in Norway recently, and it was not sufficient to prevent camera shake; the shutter itself became a problem (Canon 5D). That was at shutter speeds of around 1/4 - 1 second.

Many optically top-quality lenses (even the $5000 ones) have a quite poorly-designed mounting foot, as well as having resonances that are kicked off by the mirror. That's right, the lens itself resonates. Some are really bad that way. It's easy enough to see with the naked eye. The 200-400mm/f4 AF-S is one culprit.