Zeiss ZF 25mm test

[michaelschrag]

From one of my favorite websites....

 

http://www.16-9.net/lens_tests/zf25mm/index.html

[ShunCheung]

The very first sentence of this test report immediately bothers me:

"<I>Special thanks go to Frank Fremerey who conducted this test on a Nikon D2X using a pre-production test sample number 000009 from Zeiss in Germany.</I>"

<P>

If you read Thom Hogan's Nikon tests, he makes it very clear that he always buys Nikon cameras and lenses from camera stores and then tests them. In the case of the D80, since his sample shows the amp glow problem, he accepted a replacement directly from Nikon and he makes a full disclosure of the situation. See his D80 test report: <A HREF="http://www.bythom.com/d80review.htm">

http://www.bythom.com/d80review.htm</A>.

<P>

The ZF 25mm/f2.8 may indeed be an excellent lens, but I don't like the fact that the tester has this special relationship with Zeiss.

[glynnsheridan]

Very synical Chun....

[oceanphysics]

I'm not sure who "Chun" is, or what it means to be "synical," but I agree it loses some credibility.

[oceanphysics]

However, at least the relationship, and the fact that it was a preproduction model, was disclosed.

[eric friedemann]

Zeiss has answered yet another question no one's asking- at a time when Nikon's discontinued almost all its 35mm film cameras for lack of interest in favor of producing APS-C DSLRs, who will make a full frame-compatible 24mmish f/2.8 manual focus Nikon mount lens with no CPU? The Zeiss lens is better than a 24mm f/2.0 Nikkor- who cares? I don't own one of those anymore either.

 

 

If Zeiss wanted to wow me, it could make an 18mm f/2.0 DX lens with a CPU. A lens that equals 37.5mm on a Nikon DSLR is unlikely to blow anyone's skirt up.

 

 

As an aside, what's funny is that I've heard Zeiss is hiring an in-house U.S. rep to handle the flood of business it apparently expects to generate by making overpriced, technologically-deficient Nikon mount lenses. But I guess we should give Zeiss credit for making the best buggy whip.

[blumesan]

As a long time devotee of Zeiss/Contax (I shot Contax 35mm with Zeiss glass for 50 years)I think this is a particularly foolish, disappointing and misguided attempt to re-enter the photography business after their recent break with Kyocera. However it is not surprising given that their marketing talent was always markedly inferior to their technological ability.

 

Cheers/Mike

[oceanphysics]

Idunno -- I still want the 100 f/2 macro. The MTF's look out of this world.

[triode___]

It may not be as foolish as it seems. Why?

The target market for these lenses are really the Japanese. They are still very much into film and manual focus cameras.

 

The fact that they have at least two photo magazines dedicated to the Voightlander-Cosina line of equipment (and nothing else) attests to the amount of interest there. Enough interest for Cosina to justify manufacturing and marketing the Voightlander equipment and now these "outdated" Carl Ziess lenses.

[eajames]

The MTF's what looks out of this world.

[harvey_edelstein1]

I agree that for a macro lens MF is fine. Also, its seems there are a lot of people on this forum who still have manual focus cameras and I am sure a lot of them are used for B&W and as backup to digital bodies. For these people being able to access designs that were C/Y on their favorite cameras is a treat that isn't harming anyone.

[walterh]

I think every person is welcome to speculate if the "high" price of a Zeiss lens is "worth" it (even considering that a comparison of the price of a new lens and a used Nikkor one offered on ebay may be not very fair).

 

Everyone is welcome to speculate if Zeiss makes the right "move" or if one wants to put an MF lens on an AF body or if "anybody" would "ever" "need" an MF lens. All these speculations are personal choice.

 

There is one thing I can tell you from over 30 years of experience with Zeiss people and products( including electron microscopy- light microscopy, technical optics and optics for cameras) : Zeiss is DEFINITIVELY not the company that would present a selected (on excellence) specimen for tests and produce the shit out of the lenses that go to the stores. I have had many quarrels with Zeiss people in the past but one thing is for certain - I have always been given an honest and educated answer and opinion about technical specs and details. Any implication of your statement Shun I have to reject. As has been stated above the connection was also mentioned in the text where appropriate.

 

Shun I think if you have to resort to the implications of such a comment I personally take this as a very high recommendation for Zeiss lenses for anybody who has a sense for your emotional attachment to Nikon products in the past here at PN :-P

[ShunCheung]

Walter, as you know, I have two Zeiss lenses myself and like them a lot. My comments are about this particular test/review; IMO it is flawed in many ways. I am not at all commenting on whether the 25mm ZF is good or not, and I made that very clear in my first post in this thread. When Zeiss first announced the ZF lens plan without the details, I was very excited, until I found out later on that they have no AF and CPU. Your suggestion that I am emotionally connected to Nikon lenses is flat out wrong. My main interest is to create great images. Cameras and lenses are merely tools. If tomorrow I decide Canon or something else are better tools, I wouldn't at all hesitate to switch.

 

Ideally, a lens test should involve multiple random samples purchased from regular channels. Since buying so many lenses may indeed be too costly, at least it should be 1 random sample purchased from regular channels. That is what Thom Hogan has been doing and for the most part, Bjorn Rorslett has been doing.

 

When a lens manufacturer supplies a pre-production sample, it is not random. That is problem #1. You wonder whether Zeiss had tested a bunch and picked the best as the test sample. Moreover, the tester has a special relationship with the manufacturer to get a pre-production sample. You wonder whether the tester is totally independent or not. If you want a Nikon equivalent, Moose Peterson has had a long-term relationship with Nikon and if you read Moose's tests, every Nikon lens is described as excellent and wonderful. I simply take those reviews with a grain of salt. To say the least, there is a strong appearance of conflict of interest in this case.

 

Finally, I think that test shots are not that well done. We have gone thru this a few times. If you test a wide angle, I wouldn't do the "newspaper test." In those situations you have to shoot from a foot or two away (well within 1 meter). If you test a macro lens that way, it is perfect. Most wide angles are not designed for that type of macro shots and will show a lot of distortion. So why test a lens that in a condition not representative for its typical use?

 

So, I am merely rejecting this particular test, not the 25mm ZF, although regardless of how great they may be optically, I doubt that I'll ever buy a ZF lens without any CPU. The two Contax Zeiss lenses I have are both AF with in-len AF motor and CPU.

[walterh]

Shun I think as a first impression the "test" is not so bad because not only newspapers were imaged - but I also share your opinion that several specimens should be tested more often. I have seen specimen variation too often , especially in wide angle lenses. If I may speculate the variation among Zeiss lenses will be smaller than what one will find from some other brands.

[ilkka_nissila]

Well I thought the whole point of the high price of the Zeiss lenses was so that the manufacturer can afford to dump those lenses that don't perform well.

 

It looks good but it would be better if it had been an f/2 IMO.

 

Zeiss (like Nikon) doesn't seem to have an interest in DX/APS sensor primes. I guess they think that the Zeiss lenses are supposed to last a lifetime while DX cameras may be considered consumer toys in 10 years.

[mark_welsh1]

As the owner of 16:9, I can shed some light on the test: I came across the test at Flickr and contacted the contributor directly with a view to reproducing some of the images on my site. AFAIK, these are the first independently shot ZF25 images to be shown side-by side with their Nikon equivalents.<p><p>

 

I asked Frank the same question: how did you get hold of one?! I quote his reply verbatim below:<p>

<blockquote>"How I got the ZF? I met some influential Zeiss-guy at the

Photokina and somehow managed to get hold of a test sample.

I hope to lay my hands on the 2.0/100mm Macro as soon as

possible. At Photokina there were only dummies of this

lens, so I gues the first prototypes are in the internal

testing phase now. My Testsample of the 25mm s/n 00000009 I sent back already - I had it for one month."</blockquote>

 

The full, original test in German is published here: http://www.nikon-fotografie.de/vbulletin/showthread.php?t=22105<p>

I should point out that there are some focusing discrepancies in the original review and so I have only published a limited number of images from the original test that I feel make an accurate and meaningful distinction. The test seemed so timely, however, that I thought it deserved wider exposure and discussion.<P><P>

 

Whatever the provenance, these results, plus the MTF charts, strongly indicate that Zeiss/Cosina finally has a new ZF lens worth shouting about: ie, one which genuinely outperforms its Nikon equivalent. If only the same could be said of the ZF50 and ZF85!

[mark_welsh1]

In response to the general 'Whither ZF?' vibe: ironically, I detect a lot more love coming from Canonites than Nikonians toward the whole ZF enterprise. They're prepared to go adapted to yield any benefits visible on those demanding FF sensors.<P><P>

From the Nikon perspective, though, other than as a status symbol, it's hard to see the point of ZF, unless they are genuinely, significantly better lenses, and even then the majority of Nikon users would rather not pay the penalty of manual operation. The consensus seems to be that, optically, the ZF50 and ZF85 don't bring much more to the table than was already there.<p><p>

I bought a ZF85, fully expecting it to outperform by Canon 85L on the 1Ds II and 5D, but it was a waste of time (http://www.16-9.net/lens_tests).<p><p>

Once again proving that the pitch works better to Canon shooters than Nikon, 1 Ds II owners are justifiably excited by the prospect of a state of the art 24mm, but a top class 37.5mm just isn't quite as sexy is it?

[mark_welsh1]

In response to your points, Shun: in my experience, any flat field test is less revealing than 'real world' 3D scenes, particularly USAF charts - unless macro or 'copy' performance is the specific application under review.<P><P>For me, one of the most significant aspects of lens behaviour is its ability to draw a vivid and realistic 3D image. The best Zeiss glass excels uniquely in this area, creating a distinctively tangible sense of plasticity and depth, coupled with biting microcontrast. On the other hand, the best Leica glass has a tonality and resolution that matches reality more closely than most Nikon or Canon optics can. I agree entirely with your: 'horses for courses' argument, and I don't run tests this way myself.<P><P>

Sample variation is an insoluble problem. On rare occasions we have borrowed lenses from Canon, fresh from their service department, and found them to range from excellent to decentred. In the case of the Sigma 12-24mm, we tested six samples before finding one that worked well enough to review - and that was purchased directly from Sigma. Other tests are conducted on a one-hit basis: the cost of buying six 21mm Distagons is prohibitive. So each test is usually declared as best-available-sample or otherwise. In the case of the 17-40L test, the sample was clearly defective (though it took Canon three months to admit it), so the results appear somewhat unfairly skewed against it, but then again, there are many duff samples out there: unless you're prepared to test a batch at a cooperative dealer, you take pot luck every time you buy.

[ShunCheung]

Mark, as I have poined out earlier, any valid lens test should use a random production sample purchased from a camera store. If you use a sample directly from the manufacturer, especially when you get special treatment to have a pre-production model, people will always feel that there is an obvious conflict of interest.

 

And as I have already pointed out, any newspaper test is inapproprite for a wide angle. For more distant subjects, I would use a wall or some buildings; subjects that don't move (with respect to the camera). Or you can shoot at night and use a flash as the only source of light to freeze any movement. When your subjects are trees and grass, just a bit of bleeze can introduce subject movement. That can affect your results under high magnification.

 

As some of you know, I have tested a ZF 50mm/f1.4. On my D2X, I had a very difficult time to focus that lens manually. I ended up with bunch of slightly out of focus images and I never posted any one of those, as it would have been very unfiar to Zeiss. I did post one that shows near zero distortion on my gargae door.

 

I think another interesting aspect for the ZF 25mm/f2.8 to test is how the lack of a CPU and D information affect flash photography. I can see myself using a 24mm/f2.8 for indoor event flash photography. In fact, I frequently use my 17-55mm/f2.8 DX for that purpose. Most people feel that D does not make that big a difference. It'd be interesting to do an A/B comparison between the 25mm ZF and 24mm/f2.8 AF-D under those conditions.

[ShunCheung]

BTW, B&H already is already "accepting orders" for the 25mm/f2.8 ZF:

http://www.bhphotovideo.com/bnh/controller/home?O=Search&A=details&Q=&sku=472571&is=REG&addedTroughType=search

The price tag is US$874, yikes.

[walterh]

Mark I can only support your observation of the Sigma 12-24mm lens. I know now 4 Canon users who bought this lens in Canon mount. All were initially very happy with this lens until they noticed some inconsistent results in their shots. It turned out upon further testing that the lenses were not well in alignment, hard to detect e.g. in landscapes. All lenses were serviced by local Sigma and came back in excellent condition. In this case we can not speak of "sample variation" because 4 out of 4 is no variation. Interestingly enough I recently tested 2 samples of this lens against the Nikkor 12-24mm zoom. My conclusion was somewhat different. The optical performance was excellent - at 12mm better than the Nikkor but the Nikkor was better at the short end. The unexpected drawback was a consistent underexposure of the Sigma of 1/3 to 1/2 fstop. Does this indicate we are back to "sample variation" ^^?

 

My personal decision at the time was to go for the Nikkor lens and after good use I see only one drawback - it is a bit slow, but so is the rest of the bunch.

[mark_welsh1]

Randomly acquiring a sample will give randomly predictable results. Any which way is wrong: if you cherry pick the best of a set of ten samples, you're being fair to the manufacturer, but unfair on the consumer. If you test a shop-bought sample, there's no guarantee it will be nominal. It's a null choice.<p><P>On an entirely unrelated point, I've never felt under pressure to bias results either way. Nearly all the tests at 16:9 were made with lenses purchased via regular channels. Generally, we only run flat field tests to assess distortion.

[mark_welsh1]

In common with the 15-30mm, the Sigma 12-24mm is well designed for consistent performance throughout its range; in fact, if anything, they are best where you would least expect them to be: at the wide end. They therefore perform comparatively well against, for instance, the widest prime lenses. There's little difference between a Sigma 12-24mm and a Nikon 14mm, for instance (apart from the fact that the Sigma is much better at controlling geometric and chromatic aberration), but the 28mm AIS is a lot better than the 15-30mm at 28mm.

[fred_mackintosh]

(Primarily in response to Mark Walsh)

 

I know this is late to chime-in, but in case anyone is still out there...

 

You indicate in your comment of Nov 27, 2006 that there were some focusing inconsistencies in the original test, but that you have selected reliable images from that test for your comparison. In fact, the images given in your test suffer from significant focusing discrepancies. Have a careful look at the originals of the very first images you give at f/4.0 on age 2 of the test. The Nikon is clearly NOT focused on the central part of the image used for comparison. This can be seen by comparing the images of the bowl to the left of the center.

 

It is unfortunate that such a flawed test, independent of any possible biases or conflicts of interest originally raised, seems to have been taken so seriously.

 

I was very excited by the prospect of the ZF lenses, and was prepared to go out and buy one. But, after following all the leads I could, short of buying the ZF just to test it myself, I can only conclude that the jury is still out. In particular, I wasted way too much time pouring over the various original images on the flickr site, only to realize that there were some apparent discrepancies. I only later noted your cryptic remark about focus discrepancies. I wish you had been more clear about this on your 16-9 site, where this test is still presented as a valid one!

 

Besides, other tests (e.g., photozone) seem to suggest that CA is a bit more of a problem with this lens. Carefully scrutinizing the results I get with my 24/2.8 AIS, I conclude for now that I am unlikely to see any noticeable improvement with the Zeiss: 4000dpi scans of my images with the 24 show detail in the f/4.0-8.0 range that is practically speaking in excess of any enlargements I am likely to try with 35mm film. (I'll use my Hasselblad SWC/M if I want A2 or greater.)

 

I also note that the 25 appears from Zeiss' own published data to be the most/only attractive option in the wide-normal range. Nikon is quite capable of excellent performance in the normal (my 55/2.8AIS never ceases to amaze me) and med-tele range. As you note, there does not seem to be a real case for the 85ZF.

 

So, I remain skeptical of the case for the ZF line. I might get interested again if they come out with a 21mm. But, for now, I can only assume the market for these relies in no small part on almost religious feelings many have for German glass... it must be better if it is German, right?

 

-fred