You can go to cameraeccentric.com and read the wollensak catalogs . They refer to their soft focus effect being obtained with lens abberations. I don't recall if they mentioned anything about aperture shape. But it is worth a look just to see the older company catalogs.
"You need rise that gets your 210mm to the limits of what a plasmat gives. I don't"
That's not really the point. I'm talking about what happens when you get anywhere near the image circle of the Xenar.
If you look its MTF chart at f11 and infinity, you'll see that by the time you 80% of the way to the edge of the circle, which is not much displacement of a 4x5 neg, performance is already abysmal. MTF-20 is already at zero; MTF10 tangential lines are at zero, and MTF10 radial lines are at 25% ... barely there. At that same distance from the axis on the apo symmar, you're only 60% of the way to the image circle, and you're still in the flat-response area. Meaning, performance is practically as good as it is on-axis.
This is just one example. If you compare closely you'll find many more. Obviously, if you don't use movements, only photograph at infinity, and at f22, etc. etc. ... none of this will matter. The modern design is for people who want more flexibility than that.
Your conclusions about bokeh in LF lenses still seems to be based mostly on conjecture. It contradict a lot of annecdotal experience as well as the more serious investigations i've seen. You could be right, but I'd want to see some evidence beyond a casual comparison of two lenses to be convinced.
By the way, what are these tradeoffs you see in the small format lenses? LF lenses make huge design tradeoffs to get their big image circles. This is why the new crop of digital lenses is so good. In general, the best small and medium format lenses are better corrected than LF lenses, for this reason.
I just did some math: at f22 at infinity, the top corners of your neg will hit 80% of the image circle ... and the resulting terrible performance ... with less than one inch of front standard rise. at f11 things are much worse.
With the apo symmar, you don't hit the zone of that kind of poor performance until you've raised the lens over three inches.
Not everyone needs all three inches of rise ... but a lot of people would be happy to have more than one inch. especially in any kind of urban situation, where there's going to be more than sky in those top corners.
That's not really the point. I'm talking about what happens when you get anywhere near the image circle of the Xenar.
As I have already explained, the tessar lenses perform very similarly toward the edges of their fields of coverage as plasmats do.
If you look its MTF chart at f11 and infinity, you'll see that by the time you 80% of the way to the edge of the circle, which is not much displacement of a 4x5 neg, performance is already abysmal.
A 210mm tessar at wide apertures covers between 52 and 57 degrees (depending on who made it and what marketing literature you find). This is at least 230mm's of coverage. This is more than enough room for movement in 4x5 work (which, BTW, only requires 155mm of coverage straight on). Drop off at the edges of the field in some tessars is pretty fierce. But it's not this way in all tessars. Some drop off more gradually. Same holds true for plasmats. It all depends on how the lens was designed and implemented.
The modern design is for people who want more flexibility than [just straight on shooting].
Both Rodenstock and Schneider have recently (over the past two decades) introduced certain implementations of plasmats that increase their coverage. This much is true. The old Symmar Convertibles cover 70 degrees, while the new APO-S or Symmar-L cover 75 degrees. I see this as a potentially useful, but not huge, improvement over earlier implementations. For some reason you don't seem to want to recognize that old lenses can provide more than enough movement that achieves a similar level of flexibility as modern lenses. Perhaps I'm mis-reading you?
Your conclusions about bokeh in LF lenses still seems to be based mostly on conjecture. It contradict a lot of annecdotal experience as well as the more serious investigations i've seen. You could be right, but I'd want to see some evidence beyond a casual comparison of two lenses to be convinced.
For my own education, research, and understanding I am in the process of swapping modern lenses into round apertured shutters. Thus far my results have been absolutely consistant with my statements. This surprises me. I would have thought that out of focus area rendition would be more influenced by optical design or some other factor (such as coatings). But thus far in my investigations it is not.
The lenses I have swapped between shutters include an ever increasing list of fun things. Thus far I have swapped a 150mm Symmar Convertible, 150mm Germinar W f/9, a 150mm Fuji W/EBC, a 300mm APO Germinar f/9, a 300mm Nikkor M f/9, a 210mm Symmar Convertible f/5.6, a 240mm Germinar W f/9, and a 200mm Nikkor M f/8.
Control lenses (ie: those already mounted in round aperture shutters) have thus far included 15cm Kodak Anastigmat f/4.5, 150mm Symmar Convertible, 21cm Voightlander Heliar, 210mm Symmar Convertible.
As I said, thus far my results are absolutely consistent. Frankly, I remain shocked. It seems too easy. It seems like a too simple explanation for something we as humans might want to make more "artistic" or "complex" just because we believe a certain way.
If anyone is interested in this, all they have to do is take their new WonderLens and mount it up in a round apertured shutter. This is quick, easy, and completely verifiable as you can readily inspect the results from the comfort of your own home or business.
By the way, what are these tradeoffs you see in the small format lenses?
Any lens that needs to wrap itself around a flapping mirror runs the risk of being nothing more than a well thought out set of compromises. Many 35mm and 120 format lenses fit this catagory. Though the "best of class" Mamiya 7 lenses have been designed along the lines of the more symmetrical LF designs (plasmat, biogon).
LF lenses make huge design tradeoffs to get their big image circles.
There is nothing inherent in lens design or physics that support this claim.
Regarding the coverage of tessar designs, bear in mind that some tessarscover better than others. The Nikkor 450-M tessar, for example, has rather generous actual coverage, even by plasmat standards.
Christopher, i feel like you're muddying the waters by bringing up all these other tessar designs, etc... i'm just talking about your own example of the 210 xenar and apo symmar. My point is a simple and verifyable one ... you're limited to under an inch of front rise at infinity and f22 with that xenar, before you start seeing dramatic image degradation.
"I see this as a potentially useful, but not huge, improvement over earlier implementations. For some reason you don't seem to want to recognize that old lenses can provide more than enough movement that achieves a similar level of flexibility as modern lenses. Perhaps I'm mis-reading you?"
The old lenses do not offer comparable performance off axis to the new ones. I'm not looking at the size of image circle; I'm looking at the size of the circle that represents excelent performance. Fifty year old tessars and plasmats do not come very close to what modern designs can do.
As far as bokeh, I'm not convinced by your tests is largely because I don't know how you're evaluating your results. There's a lot of subjectivity involved in evaluating bokeh; there are also a lot of objective parameters that influence it, including subject contrast, magnification range, distance in front of/behind plane of focus, and aperture. I don't know what your standards are, or what conditions where used to make the testt images. The sites I've seen that demonstrate spherical aberration analyze the conditions, and they show the results ... in pictures.
"Any lens that needs to wrap itself around a flapping mirror runs the risk of being nothing more than a well thought out set of compromises"
true for wide angle lenses, but not most normal lenses and longer.
""LF lenses make huge design tradeoffs to get their big image circles."
There is nothing inherent in lens design or physics that support this claim.""
the researchers at Schneider would be astonished to hear this. According to them, image circle is the single biggest design compromise in any LF lens design. It's specifically why all the small camera lenses have dramatically higher MTF performance, why the Schneider and Rodenstock digital lenses have dramatically higher MTF performance, and why it's impressive that modern plasmats have been able to get wider without getting softer.
I'm not an optical engineer, but I'd be happy to put you in touch with one of the engineers who educated me on the topic.
... i'm just talking about your own example of the 210 xenar and apo symmar. My point is a simple and verifyable one ... you're limited to under an inch of front rise at infinity and f22 with that xenar, before you start seeing dramatic image degradation.
Wow! Really? Not in my experience. Not with the 210 Xenar I own. It's not a super lens or a Linhof select or a unique design or anything special. It's just an off the shelf Schneider Xenar f/6.1. That's it.
The old lenses do not offer comparable performance off axis to the new ones. I'm not looking at the size of image circle; I'm looking at the size of the circle that represents excelent performance. Fifty year old tessars and plasmats do not come very close to what modern designs can do.
I have evidence to the contrary. You're free to review my VOE any time you like.
As far as bokeh, I'm not convinced by your tests is largely because I don't know how you're evaluating your results.
I know nothing from "bokeh". I assume, until someone shows me otherwise, that many people equate "bokeh" with what others of us phrase as the rendition of the out of focus areas.
I have read where Cosina/Voightlander thinks that multi-coatings/single coatings influence "bokeh". I have read where other people put other parameters on their definition of "bokeh". I'm sure there are as many definitions and interpretations as there are photographers with brains.
For the record, my test setup and criteria remain simple. My criteria is how lenses render out of focus areas. My references are old lenses that are claimed to be examples of the kinds of effects people describe. My test setup subject matter is around 5 feet away. My VOE are the negs I create as part of the process of comparing lenses.
When I can take a modern multi-coated optic and drop it into a different shutter and get a different result that, in fact, matches the out of focus rendering of reference lenses, well I think I've uncovered something relevent. To confirm what I think I'm seeing, I'm taking the effort entirely at my own expense to try as many different lenses as I can afford. I have nothing to loose, right? Perhaps I have something to gain; a little knowledge.
""LF lenses make huge design tradeoffs to get their big image circles."
There is nothing inherent in lens design or physics that support this claim.""
the researchers at Schneider would be astonished to hear this. According to them, image circle is the single biggest design compromise in any LF lens design.
What tradeoffs did Schneider tell you they needed to make to achieve the coverage they designed to? For the people I have spoken to, it's more along the line of ease of manufacturing traded-off against glass availability traded-off against projected ROI. Nothing really about the physics involved. That's been understood and solved a long time ago. The parameters are well known.
I'm not an optical engineer, but I'd be happy to put you in touch with one of the engineers who educated me on the topic.
Yes. Please do!
I too have been talking with lens designers. It will be very very interesting to compare notes.
"When I can take a modern multi-coated optic and drop it into a different shutter and get a different result that, in fact, matches the out of focus rendering of reference lenses, well I think I've uncovered something relevent."
i think that's a fair thing to conclude from your methods.
on the other hand, what you said earlier ... "I have empiric evidence that shows modern lenses mounted in round aperture shutters change the characterists of the out of focus areas to match earlier lenses (both single and uncoated) that typically came mounted in round aperture shutters. There is no "magic" in lens design that changes this fact."
strikes me as overreaching--a broader, more sweeping generalization than i think you can make from the information at hand.
same with this earlier statement: "For resolution, with the sun over my shoulder, I have to say there is absolutely no difference between the lenses."
that one would need a paragraph full of qualifiers to be of general use to people. because there a lot of differences between the lenses in question. you've just found specific circumstances that don't reveal the differences. someone working under other circumstances would have a very different experience.
i'm not trying to give you a hard time, Christopher. But I tend to jump when I see conclusions that fly in the face of both my own experience and established research. especially when they aren't even supported by the evidence cited. for one example, i know (from experience as well as mtf data) that a tessar derrived lens wouldn't work for me doing urban landscapes, but it would probably work fine in the plains or the desert. likewise, my current plasmat might not be enough for someone who specializes in architecture. these qualifications are important.
i'll be happy to put you in touch with one of the technicians i've spoken with. in the mean time, take a look at some typical mtf charts for 35mm lenses, and compare them to the best large format lenses. you'll see that the best of the small camera lenses have an amazing combination of higher speed, lower price, and better optical performance. It's such a glaring set of differences that some people have formed conspiracy theories around it. but the answer is a lot simpler: unconstrained by the need for large image circles, the designers can easily make much sharper lenses. the schneider digitars are a perfect example of this. take a look at the mtf charts, and make sure you notice how much higher the resolutions are that they're graphing. the schneider lenses for rollei 2-1/4 cameras provide another direct comparison.
At photo new york last fall, a technician showed me pictures from a special purpose schneider lens that could resolve over 200 lp/mm at 50% mtf--absolutely unheard of in all previous camera optics. the catch? it was designed to cover, without movements, a digital sensor the size of the fingernail on your pinky. at the other end of the spectrum. lenses that cover ULF measure much worse than ones that cover 4x5. i've heard it posited that with bigger film sizes, the designers don't try as hard ... but this is silly. their constraints are rooted in optics, and to a lesser degree, in economics.
Posting Permissions
Reply With Quote
Bookmarks