Interesting Leica lens tests

greenspun.com : LUSENET : Leica Photography : One Thread

Paul Chefurka did some interesting Leica lens tests that I thought you all might want to have a look at. (Forgive me, Paul!)

http://members.home.net/chefurka/Photo/LensTests/LensTests.html

Comments?

-- Steve Hoffman (shoffman2@socal.rr.com), July 30, 2001

Answers

P

aul's results look about what I'd have expected. T h e p o o r performance of the Nikkor 50/1.8 at f/4 is surpr is in g .. . I su sp ec t a focus shift or something like that m igh t b e r esp ons ibl e. And it 's go od to see that the Nik kor 85/2 AI- S, o ne o f my fav orit e Nik kor le n ses, prov e s to work very well . It' s a s w e e t l e n s.

Howev er, the re is a reason I've s old off my N ik on g ear and use on ly Leica M kit f or my sy stems ca mera .. . And t

his t es t does demonst ra te why th at woul d be sati sf ying, a t leas t in the si mple case.

Godfrey

-- Godfrey (ramarren@bayarea.net), July 30, 2001.


Paul,

Thanks for the test information. I have trouble though, seeing where the Nikkor 85/2 shows an edge over the 90 Apo that you mention in your conclusion. I thought the pictures indicated a slight improvement for the 90 Apo over the Nikkor 85. Very close though.

-- Sam Smith (Ruy_Lopez@hotmail.com), July 30, 2001.


Interesting test, but I agree with the above comment, using the sample you gave us, both the 90mm Summicron and the Apo-Lanthar seem to be sharper than the Nikkor to me.

-- Robin Smith (smith_robin@hotmail.com), July 30, 2001.

I can't really tell anything from looking at those scanned images at screen resolution except for a couple of obvious things such as that in the 90 shots there was a difference in exposure by about 1 stop, and in the 50 shots the Nikkor was mis-focused. If I were doing side- by-side comparisons between a Leica M and Nikon F3 I would have done 3 things additionally: first, I would not have used a 100-speed film because the Leica shutter at 1/1000 varies at least 1/3 stop from shot to shot while the F3 is dead-bang-on each time. Also, I would have locked the mirror up on the F3 because even at 1/1000 the Leica has a distinct edge in the vibration department. Lastly, I would not have chosen small treetop branches and leaves as my test target because even on a still day there's some subject movement.

-- Jay (infinitydt@aol.com), July 30, 2001.

to go through all that effort and use a subject that moves, is highly questionable methodology. seems an odd choice to reference against, though the Leica lenses certainly rise to the top.

-- daniel taylor (lightsmythe@agalis.net), July 30, 2001.


There's no pleasing everyone, I guess. At least I used a tripod :-) The criticisms of using a moving subject are valid. And yes, I may have misfocussed that Nikkor 50/1.8.

In order to satisfy myself and to calm the carps, I'll be doing another test series in a week or so. The lenses will include the 35 Summilux ASPH, a 35/2.0 Nikkor, the 50 Summicron and Nikkor, and the 90 AA and the 85/2.0 Nikkor. Test film will be Velvia. I will use a stationary subject (probably a street scene) and I'll lock up the mirror on the F3. I'll try to focus the lenses before I take the pictures :-)

My main motivation for doing these tests was all the bleat on the net about how you just can't tell a Nikkor from a Leica lens at "normal" apertures. At least the 50mm test showed this (the 55 Micro example at least, even if the shots with the 50 were suspect), though the 90mm test was a bit of a shock. If anything, doing this has made me a lot more skeptical about anybody's lens performance claims.

The other thing that came through loud and clear to me is how little resolution degradation is required to eliminate the superiority of really good glass. It's one thing to hear it, but it's another thing to see how close the 3E is to Summicron/Summilux performance.

-- Paul Chefurka (paul_chefurka@pmc-sierra.com), July 30, 2001.


BTW, in the true spirit of the scientific method, anyone who disagrees with my methodology or conclusions should feel free to rerun the experiment and post their results. The more data points we get, the more reliable the curve fit becomes.

Any takers?

-- Paul Chefurka (paul_chefurka@pmc-sierra.com), July 30, 2001.


I found the test very interesting. The 90mm Apo-lanthar proved the match of the other lenses, even at f4 which was almost wide open compared with the other lenses being stopped down 2 stops, which makes it excellent value for us CL/CLE users. One thing that may be misleading is the 55mm f2.8 Micro-Nikkors perfomance, as at infinity its at its very worst perfomance. With its close range correction system at a meter and less I would think it will be king. Mine is one of the sharpest lenses I have when used as an actual macro lens rather than a regular standard lens, where then my summicron is the lens of choice.

-- Joel Matherson (joel_2000@hotmail.com), July 30, 2001.

I have never found a 50mm Nikkor which was sharper at infinity than the 55/3.5 Micro or 55/2.8 Micro, though the 50/1.8AF can equal them; similarly my 60/2.8 Macro-R is as sharp if not a tiny bit sharper at infinity than my 50 Summicron-R. The 100/2.8 APO-Macro is sharper at infinity than any other lens at that general focal length of any manufacturer I've ever tried...I suspect the 90AA could equal it but I haven't tried that lens yet.

-- Jay (infinitydt@aol.com), July 30, 2001.

Paul:

Just to chime in on the moving subject thing... Unfortunately, it is obvious by the differing leaf patterns in some of the direct comparisons that there was at least a light breeze blowing... But I think the exercise is at least suggestive, if not conclusive, of how these lenses compare at the given apertures. The one other thing I perhaps overlooked in your test was what shutter speed(s) you were using.

-- Jack Flesher (jbflesher@msn.com), July 30, 2001.



The shutter speeds were 1/250 and 1/1000.

-- Paul Chefurka (paul_chefurka@pmc-sierra.com), July 30, 2001.

I have tested my Nikkors and Leica lenses on the same resolution chart, but all that really showed me was which ones reproducrd a high contrast, fairly close, flat field subject the best. Not surprisingly, my 55 f3.5 Nikon Macro did very well, and so did the 50 Summicron and 90 Elmarit. My 35f2.0 Nikkor had tremendous center sharpness but showed much less detail on the edges than even my old 2.8 Summaron. My 40 Rokkor, which is superb in the field at nearly every aperture, tested slighltly below the 50 Summicron on the chart. What I learned from all my time testing lenses is that real world results are what count the most, and is is extremely difficult to come up with any method that truly tests all the important things that make a great lens great and a mediocre lens mediocre. As a rule of thumb, it did help me to see which apertures tended to have the highest resolving power and contrast for a particular lens (again, of a flat, high contrast, fairly close subject). After using a particular lens for several hundred images, you can usually figure this out in the field anyway.

-- Andrew Schank (aschank@flash.net), July 30, 2001.

I believe that Paul’s test is very close to the actual performance of these lenses and especially between the lens quality of Leica M lenses and Nikon lenses. For about 8 years I was using Nikon lenses and for the last two years I am using a Minox 35GT and Leica M. I am printing large B&W prints on my own and in these prints I can see clearly the quality of the lenses. Also I made similar test between Leica and Nikon lenses and I had about the similar results. I take the same images with two or three lenses on the same film, I develop all the films together and then I print a part of the negative from a 20x or 25x magnification. In these small prints you can see the differences between the tested lenses. The different thing I do in test is that I shoot in different lighting conditions, I make shoots against the sun or the light, I make shoots wide open, I make shoots in different subject to camera distances, very close portraits, shoots in 2 or 3 meters and finally in infinity. I have made the test: 35/2 Asph vs Nikon AF35/2 and summicron-M 50/2 vs Nikon AF 50/1,4. In both tests Leica lenses are far better than Nikon lenses and especially in 50mm lens test. Leica lenses are also far better in infinity, wide open and in low lighting conditions. In very close and in some medium distance shoots Nikon AF35/2 is very close to Summicron 35/2 Asph, but again the difference is visible by naked eye if you compare the two trial prints. If you make only one large print with the Nikon AF35/2 and you look at it you would say that it is great, but if you compare it with a similar print with the 35/2 Asph you can see the difference between the two lenses. Leica lenses, as well as all good lenses, produce smaller grain than other lenses and I think that this is the main point as far as lens quality is concerned. I made tests only with the AF Nikons but I think that some older manual Nikkors (and especially wide angles like 20,24 and 28) are very close to Leica lenses in quality. Also Nikkor 105/2,5 and Macro AF60/2,8 which I own are great.

Paul, I am waiting for your new tests!

Street Photography by Dimitris Kioseoglou

-- Dimitris Kioseoglou (kosefoto@otenet.gr), July 30, 2001.


Paul-

When you retest the 85/90, do a test at f/2. That's where the Nikkor should die a horrible death. I used and loved the lens for 15 years - it is one of Nikon's best at f/4 and smaller apertures, but turns into a Coke™ bottle bottom wide open.

Lying down, all people are about the same height, and at f/8 all lenses are about equally sharp. 8^)

I'd also recommend some tests WITHOUT mirror lockup - in the real world what counts is system sharpness. A fairly good 3rd generation Leitz 28 Elmarit that shows almost ALL its resolution at 1/4 second is far more useful to me than a really sharp Nikkor 28 that starts showing camera shake at 1/30 or 1/15.

But do the controlled no-mirror tests, too. These were very interesting!!

-- Andy Piper (apidens@denver.infi.net), July 30, 2001.


I don't think using tree branch/leaves provide conclusive evidence of performance of lens. It is quite apparent that the leaves had moved from one camera to another. A lens thus tested to be unsharp might acutally had a breeze blowed on the branch.

It is far better to photography window frames of stationary building, a bill board at far distance, with printed words.

-- martin tai (martin.tai@capcanada.com), July 30, 2001.



I agree, Martin. I had assumed that a 1/1000 shutter speed would negate any motion from the light breeze that day, but in retrospect it was a confounding variable that should have been controlled. I'll be doing the next round of tests with a stationary subject.

-- Paul Chefurka (chefurka@home.com), July 30, 2001.

Paul,

I really appreciate you taking the time to evaluate these lenses. my comments about the moving subject were just observations, divorced from my gratitude. too many myths linger, and propagated by photographers who have never bothered to validate them.

thanks again.

-- daniel taylor (lightsmythe@agalis.net), July 30, 2001.


Hi, I found your test to be quite interesting. I suppose that if you take every single relevant comment into consideration for future tests, you will have achieved a higher level of scrutiny than Erwin Putts. In fact, I would be surprised if Mr. Putts ever had anyone looking over his shoulder as you do in this news group.

Anyway, in my judgment, you remain one of the guru’s of this news group and for that I am grateful.

Eddie

-- Edward Steinberg (es323@msn.com), July 30, 2001.


Between the Nikkor 85mm f/1.8 and the f/2, I recall reading in the Nikon Compendium that one of these was designed for soft focus at full aperture, to be used as a portrait lens. I *think* it was the f/2. If that's the case, I must say it sure does clean up by f/4.

-- Bob Fleischman (RFXMAIL@prodigy.net), July 30, 2001.

As others have said, many thanks for taking the time to plan and execute these tests, and to share them with us, and even receive feedback.

-- Tse-Sung Wu (tsesung@yahoo.com), July 31, 2001.

I found the test most useful and hope to see more like it in the future. I have found the Erwin Puts view, all Leica lenses are better than everyone elses a difficult one to accept. Each lens manufacturer has a lens that may excell over all others of the same focal length. To claim all Leica lenses are best is simply untrue. Nikon alone makes dozens of lenses and im sure no-one has done comparisons with them all to make that claim. I remember a time when the Nikkor 105 f2.5 had the reputation of being the sharpest lens of its focal length range. I would like to see Leica R lenses added to the comparison. I suspect that the sharpness differential may be rangefinder/SLR related due to constraints on SLR lens optical formulas which also have to illuminate the morror box. Telephoto lens designs are less affected by this constraint to which the fact that the Nikkor lens in the short telephoto range is able to compete. Many thanks to the tests author, to me it was far more believable than the way Erwin presents his views.

-- James Cooke (james.c@mis.net.au), July 31, 2001.

As a former scientist, I can appreciate the level of rigor displayed in Paul's tests. And as a matter of idle curiosity, the results were interesting and informative.

But as a Leica user who has never mounted his Ms on a tripod and who regularly shoots handheld at 1/25 with a 50 or 1/50 with 90, the tests don't mean much. If you're willing to lug around a tripod, shoot with plenty of light, and be deeply concerned with sharpness, medium or large format will do a much better job.

[Yeah, I know I'm just trying to start trouble. My geek half really enjoyed the tests, but my pragmatic half won't let me keep my mouth shut. . .]



-- Mike Dixon (mike@mikedixonphotography.com), July 31, 2001.


Hey Mike,

Don't worry, your attitude won't start any trouble with me. This was the first time in over two years that I've put a camera on a tripod. Like you, I'm a pure hand-held available-light shooter.

One of the pics that convinced me of the quality of the 90AA was a pin-sharp performance shot of a cabaret singer taken wide-open, hand- held at 1/60. That's what I buy Leica M cameras and lenses for, not this kind of stuff. If I were to put a camera on a tripod, it would be a camera that *needs* a tripod - like an RB67 or something. Leicas belong in the hand.

These tests were done essentially out of idle curiosity. The results were interesting and anomalous enough that I'm going to redo some of them this weekend with even a bit more rigour. This time I'll look at f/2.0 as well, which should give people even more to talk about.

Then I'll be done with lens testing for life. There isn't enough money in the world to make me do this full time :-)

-- Paul Chefurka (paul_chefurka@pmc-sierra.com), July 31, 2001.


Thank´s, who should we thank, well thank´s Steve for sharing, and to Paul for an bold test, I confess my self as a no test photographer, although very interested in other person´s test, and Paul´s has a lot to show for who ever want to see.

I must say I wish you had done some wide open test specialy in the 90´s, I wish I could send my 90/2 non asph to add to your tests, and my 28/2.8 (3rd version).

I agree next time you picture a static object, just beware of earthquakes.

By the way Paul, I just loved your "life" pictures, and the first "death" photograph is also beautiful, I´ll return to see your others portafolios.

Thank´s again

-- r watson (al1231234@hotmail.com), July 31, 2001.


Paul, in you next round of test result scanning, it would be very helpful to include the performance of lens at the corner or edge -- ie, sample from center or near center as you did, and one sample from edge.

Have fun !

-- martin tai (martin.tai@capcanada.com), July 31, 2001.


Moderation questions? read the FAQ