210mm Lens Test - draft doc posted

[Christopher Perez]

Esteemed Ladies and Gents,

I just put the finishing touches on the draft version of my 210mm test:

http://www.hevanet.com/cperez/test/BigMash210.html

It may not be in it's final state. But it's getting closer. There may be a few points to hone or a few topics to yet include. But you'll get the general idea.

I just hope this does NOT drive up the value of old Symmar Convertibles. Afterall, don't many "boomers" who want to relive their past want/need a convertible? :-)

[paulr]

Hi Christopher,
thanks for posting all this. it's always interesting to see these kinds of results. but i'm confused about some of your conclusions and the methodology that led to them. you draw conclusions about the performance of these lenses vs. modern apo wonder lenses, but you didn't test any of the latter. and more generally, you did not do a comparison of in-focus areas off axis, which is where the differences between lenses are most significant and obvious. If you look at Schneider's MTF curves for the xenar, you'll see that on axis it's comparable to the apo symmar. but its performance falls off faster off axis, and its image circle is much smaller. so you would expect softer corners, at least if you're using camera movements.

i'm intrigued by your idea that aperture blades and not aberrations are responsible for nice looking out of focus areas. this would be worth testing directly at some point.

[Ed K.]

Hi Christopher -
Very interesting, nicely done essay. I wonder about the f11 super closeup with the Nikkor - it has what look like sharpening artifacts ( specks? ) that the others don't, although I'll take your word on the general results.

Also, it would be nice to have a few more bokeh examples. If one looks close, the Docter optic bokeh does look interesting - more rounded; on the other hand, the 7 bladed Xenar bokeh at times looks smooth, no? I wonder if your test will really create a Docter optic frenzy - I'm still kicking myself for passing up a Docter optic 1000mm on z' blah.

Cool contribution, indeed!

[tim atherton]

Ed:

"I noticed that I under-exposed the Nikkor 200 M f/11 image. I don't know by how much I underexposed, but it feels like nearly two stops. I must have set the shutter speed the wrong direction when I changed the aperture. "

[Ralph Barker]

Do Symmar convertibles have fins? ;-)

[medform-norm]

After performing this test, I believe I begin to understand why old photos can have those qualities. Small blade number apertures can produce rapidly changing dark to light transitions that can be perceived as "harshness". On the other hand, old lenses tend to have many many aperture blades that define a nearly round aperture shape. This is the single most important influence on out of focus area rendition. Further, I believe it is possible to duplicate this effect with modern optics by simply mounting them in multi-blade apertured shutters.

For me this is a very significant finding.

Congratulations Mr. Perez! - you have found out the very reason why we (and perhaps some others?) are looking for older lenses and older shutters with more than 8 blades. You have taken an awful lot of effort to achieve this result. Unfortunately, your test may drive up prices on older prontors :-(

(as a side effect, we may now expect to see a new phenomenon developing on eBay with sellers stating the number of iris blades of their lenses, together with a link to your article - I can already see dagor77's version of such an auction: 'not 5, not 8, not 11, but - holy smokes - 19 iris blades!)

To our experience, the best way of testing a lens for iris shape influencing the out of focus areas is to take shots of small branches of trees positioned in the off axis area of your lens. (Something simple like a portrait of someone sitting on a park bench with shrubbery in the background will do nicely).

We think lens manufacturers have switched to fewer blades for economical reasons and because of the development of the (auto-aperture) automatic shutters: an iris with 5 or 8 blades has a lot less resistance in operating and is faster in opening and closing than a multi-blade version.

Claims that bad bokeh is the result of residual spherical abberrations should be tested with different lenses mounted in multi-bladed 'round' irisses.

There, we've said it. Modern shutter prizes will soon hit an all-time low.... let's all go back to the good old waterhouse stops. We've just made a nice set of 'imitation imagon' stops for an aerial lens with promising signs of softness.

[Bill_1856]

I don't understand. Why test wide open and f:11, if most LF pictures are actually made at least at f:22 or f:32?

[Andre Noble]

My guess is that differences between lenses are less obvious at f22 & f32.

The one thin I keep seeing with CP's resolution tests is a slight tendency for older runs of a particular lens model testing better than morre recent production runs. What's going on? Is quality control down, now that we're moving onto smaller formats and the digital age?

[Michael S. Briggs]

First, let me thank you for the work that you have done and made available to us. I have some ideas that you might want to consider in your writeup.

If you look at your ranking the lenses by sharpness wide open, the three that are nearly tied for being sharpest are the three with the slowest maximum apertures. The f8 Nikkor-M and the f9 Germinar W have an easier job of being sharp wide open since wide open isn't that fast for these two lenses. This way of comparing the lenses is a bit unfair to the faster lenses. A different test would have been to stop all of the lenses to the aperture of the slowest lens of the group. Of course, the Xenar is impressive, beating more complex designs, and lenses with slower maximum apertures.

None of your the lenses in your sample are that old -- all are at least single coated. It has been discussed before that LF lenses don't have a lot of air/glass surfaces and so single versus multicoating isn't that important.

I think the bigger difference between the lenses in your sample is the different designs: one triplet, three tessars of different speeds, one normal speed (f5.6) convertiable plasmat, and one slow (f9) symmetrical plasmat.

One thing that you haven't tried to measure is the different coverages of the different designs, and how the coverage changes as you stop the lens down. I have done this in a qualitative manner, by visual inspection of negatives, with several lenses that I own. I find a significant improvement in off-axis sharpness as I stop down, with an implied increase in coverage. This is probably why many people say that better results are obtained by stopping down a few stops. Perhaps this is showing up in the last column in your resolution tables for the 200 Nikkor-M in the section "USAF Resolution Test Chart Comparison". The center resolution doesn't change much, while the edge is increasing at smaller apertures (higher f-numbers). (You might want to make it clearer how far these measurments are off-axis). My visual examinations gives me the impression of a more dramatic effect. One description (of the little brother of the 200 mm Nikkor-M) is here: http://www.largeformatphotography.info/lfforum/topic/494534.html#373881

In your photos, how much can you tell how the sharpness varies with distant from the optical axis? It must, but perhaps the 3D subject and the depth of field are making this difficult to discern.

As paulr says, in your conclusions you make a statements about the Symmar and the modern Apo "wonder", but you didn't test the modern equivalent of the Symmar -- a currently manufacturered f5.6 plasmat from a major manufacturer. And if you wanted to do this test, the place to look for a difference would be off-axis, either in better sharpness or increased coverage.

[Per Madsen]

Interesting !!!

Regarding multicoatning versus single-layer coating I have only seen visible benefits

in high contrast situations and in "against the light situations".

I.e. my 65 mm multicoated Nikkor can handle situation with some of the sun in the

picture, but my 90 mm single-layer coated Super Angulon flares out with reflections

in the whole picture.

I think that the results very much is the same as the test Zeiss Historica once did of

a prewar uncoated Sonnar, a postwar coated Sonnar and a modern multicoated Planar

(all either 1.4 or 1.5):

The modern lenses have better edge performance and the differance between a uncoated

lens and a coated lens is much larger than between a single-layer coated lens and a multicoated

lens, except for high contrast situations. The differance in center performance was much

smaller than on the edges.

[medform-norm]

I don't understand. Why test wide open and f:11, if most LF pictures are actually made at least at f:22 or f:32?

Contrary to your belief, LF pictures are indeed made with these large apertures - and not just with f:22 and larger. There are lots of people who like and use the shallow dof. Has this escaped your notice?

[Richard Schlesinger]

I wonder about converting/using Waterhouse stops, rather than an adjustable iris. It would seem to offer a perfectly round (if that is desirable) shape.