Some tests rank-ordering a variety of classic and modern LF lenses

[Joseph Kashi]

With some trepidation, given the flame-wars that seem to occur too often when discussing lenses and other photographic topics, I am posting the rank-ordered results of tests that I have made of the LF lenses that I've accumulated over the past 45 years. I am not slighting lenses not listed here - I just don't have those lenses available for a variety of reasons, such as unafforability ( Schneider XL, etc.).

I wanted to provide such information as I could, in the hope that it might be useful to others but with the caveat that I am not a professional testing lab and that I recognize that some others on this forum have a great deal of knowledge. I'm sorry for this lengthy introduction, but I hope that doing so will reduce misunderstanding and consequent posts.

Although I have numeric data for each lens, I instead rank-ordered these tested lenses because the final results on the 5x7 Delta 100 negative were often very closely bunched when made at my normal shooting aperture of f/22. Rank-ordering was used as well to avoid pixel-peeping arguments to the extent possible and in the recognition of potential sample variation.

I did review every test, retested every suspect initial result, and then reviewed the rank-ordering three times before this OP to reduce inadvertent cognitive bias. Rank-ordering results were consistent from review to verification review, suggesting that they were stable results. When re-tests were made of a particular lens, I always used the best negative as that was more likely to reflect inherent optical quality rather than poor technique on my part.

These tests and any re-testing for verification, were done over the past several months in a consistent manner biased toward end-use rather than pixel-peeping. I basically wanted to sort out and cherry-pick the best lenses for a variety of backpacking and studio 5x7 camera outfits, with a stray 4x5 or two thrown in.


Equipment used: Toyo G 5x7 and Rittreck View 5x7, verified film holders, ISO 12233 target, Delta 100 film developed in XTOL 1:2 with intermittent agitation for best acutance. Each final test negative was reviewed on a light table on several times with a well-corrected Wista 10X loupe. Rank-ordering were made directly from the best negative, not from scanning or post-processing, to avoid false results. All camera movements centered.


I'll likely not reply to argumentative posts nor feel the need to get the "last word" nor the need to feel superior. The best-performing lenses are listed on top. With exceptions as noted, these are all very usable on 5x7, although the image circle on some is rather tight, not allowing much movement even though manufacturer-rated for 5x7. 4x5 lenses like the Angulon were shot on 5x7 Delta 100 for consistency but only the central 4x5 section was evaluated for these lenses.)


All of the lenses in the top two groups are very close to each other in resolution at working apertures, with only mild to moderate differences between those in each group. The top lenses in the second group are very nearly as good as the bottom lenses in the first group, but I thought that several groupings would be a bit clearer.

Lenses in the third group are still usable but would not be my first or second choice. The top several lenses in the first group are discernibly sharper than the bottom lenses of the second group, but all of the lenses in groups one and two are entirely usable and differences would not be very noticeable, if at all, in normal use. Really, how many angels can dance on a pin-head? At f/22 on a 5x7 Delta 100 negative, virtually all of these lenses are more than good-enough and LF pixel-peeping seems rather beside the point of actually using them to make good images.


A few observations: There are several instances where I had and tested two copies of the same lens model, including the Kodak 203/7.7, Fujinon 125/5.6, and Fujinon 150/5.6. In all three instances, performance was virtually identical with each copy, suggesting not a lot of manufacturing variation in more modern optics. Although the rank-ordering suggests variation, the performance of Group 1 lenses are in fact are all very close to each other at f/22. (Dagors were stopped down to f/28.)

There was more variation between similar older lens designs, such as Protar VIIa, with some in the top ranking and some well below that.

The outstanding ultra-wide-angle lens in this group was clearly the Nikkor 90mm/f8, which was the sharpest UWA lens tested and covers 236mm despite its f/8 aperture.

Later Fujinon EBC Plasmat designs generally did quite well and were very consistently sharp and contrasty. Fujinon's manufacturing variation seems low, suggesting good quality control. There was not a great deal of difference between Fujinon's earliest 250/6.7 single-coated Plasmat and its last-generation cousin, the 250/6.3 CM-W EBC-multicoated Plasmat. Some of the 1950s era US-made lenses, such as the Goerz 7" Dagor and 14" Red Dot Artar, and the 203/7.7 Kodak Ektars lived up to their reputations and performed on par with later-model Fujinon W and Rodenstock Sironar-N Plasmats.

Sorry for the length below but I could not set up columns for readability - extra spacing was the next best option.


Group 1 - First-Choice: Best-performing lenses in descending rank-order, all very close in performance at f/22.


210/5.6 Fujinon NWS (Plasmat style, EBC, markings around outside of barrel)

250/6.3 Fujinon CM W (Plasmat style, EBC, markings around outside of barrel)

150/5.6 Fujinon NWS (copy 1) (Plasmat style, EBC, markings around outside of barrel)

150/5.6 Fujinon NWS (copy 2) Plasmat style, EBC, markings around outside of barrel)

180/5.6 Sironar N (Plasmat style, EBC, markings around outside of barrel)

90 /8 Nikkor SW (105 degree UWA coverage)

125 /5.6 Fujinon NWS (Plasmat style, EBC, markings around outside of barrel)

183 /6.8 Dagor (Goerz American Optical Co, coated, Ilex 3

203 /7.7 Kodak Ektar (copy 1) (Copy 1 in Prontor, single-coated, European made?)

250 /6.7 Fujinon W (Original Plasmat formula, single coated, marked on front retaining ring)

360/9 Kern Apo (Dialyte Style process lens, remounted in Copal 3 by Grimes)

90 /5.6 Fujinon SWD (Later model Fujinon 105 degree UWA lens, EBC, marked on outside of lens barrel)

105/8 Fujinon NSWS (Later model Fujinon 100 degree UWA lens, EBC, marked on outside of lens barrel)

125 /5.6 Fujinon NWS (copy 2) (Plasmat style, EBC, markings around outside of barrel)

203 /7.7 Kodak Ektar (US-made Dialyte-style lens, Supermatic shutter, 1947, single-coated)

305 /9 Schneider G-Claron (Plasmat-style process lens, factory mounted in shutter, single-coated)

305/9 Schneider Repro-Claron, mounted in Compur #1 (Another sharp Dialyte-pattern lens also fits Copal 1)

355/9 Goerz Red Dot Artar (US-made Dialyte-style lens, factory-mounted in Ilex shutter, single-coated)

240 /9 Schneider G-Claron (Early Dagor-style process lens, later fitted into Copal shutter, single-coated)

165/290/290 Zeiss Protar VIIa (Triple convertible lens, uncoated, pre-WWII Zeiss in small Compound shutter)



Group 2 - Second-choice: Nearly as Good, ranked in descending rank-order, again very close in performance



75 /8 Fujinon SW (4x5-only, early model Fujinon 100 degree UWA lens, single coated)

210 /9 Schneider G-Claron (Plasmat-style process lens, factory mounted in shutter, single-coated)

90 /8 WA Dagor (Rated for 4x5 but OK on 5x7 small apertures, Goerz Am., coated, Rapax shutter)

115 /5.5 Voightlander Ultragon (1950s model, single-coated, Compur #2 shutter, nice 5x7 wide angle)

120 /8 Fujinon SW (early model Fujinon 100 degree UWA lens, single coated)

152 /6.8 Dagor (Goerz American Optical Company, single-coated, factory mounted in Rapax shutter)

183 /350/290 B&L Protar VIIa (US-made triple convertible lens, uncoated, factory-mounted in Acme #3 shutter)



Group 3 - Third-Choice: Decent and Usable, but discernibly not as good as above, again in descending rank-order


90 /6.8 Schneider Angulon (Decent 4x5-only lens, but not as sharp as W.A. Dagor and Fujinon 75mm/f8 SW)

146 /290/220 Zeiss Protar VIIa (Triple convertible lens, uncoated, pre-WWII Zeiss in small Compound shutter)

300 /9 Nikkor Q (Tessar-style Nikkor, predecessor to 300M, single-coated in Copal 3)

450/9 Nikkor Q (Tessar-style Nikkor, predecessor to 450M, single-coated in Copal 3)

400/8 Fujinon T (classic telephoto design, covers 5x7 barely, yet better than I expected)

305 /6.8 Dagor (German-made bulky early uncoated Dagor in large Acme shutter)

[BrianShaw]

What was your DV or DVs?

[Joseph Kashi]

What was your DV or DVs?

Not sure that I understand the abbreviation? Are you referring to variance?

[BrianShaw]

Not sure that I understand the abbreviation? Are you referring to variance?

Oh sorry... DV = Dependent Variable... that which was measured. What did you measure to establish this ranking? I’m not picking on your ranking... just interested in your method.

Interestingly, I don’t think you included a single lens that I use. :)

[Jody_S]

Interesting that all of your Tessars are in the bottom group, I would have expected modern coated Tessars to rank equally with coated Dagors, at f22-28.

[cowanw]

Oh sorry... DV = Dependent Variable... that which was measured. What did you measure to establish this ranking? I’m not picking on your ranking... just interested in your method.

Interestingly, I don’t think you included a single lens that I use. :)

Well I wondered the same thing. You mentioned in passing Resolution, Sharpness and Contrast and rank them by Performance (and goodness); so I also wondered if and how you measured your grades of performance or if the ranking was descriptive in nature. I don't have any of these lenses either and so no opinion as to the ranking although as my interests are in portraiture I suspect I might rank them in reverse.:)

[Joseph Kashi]

Interesting that all of your Tessars are in the bottom group, I would have expected modern coated Tessars to rank equally with coated Dagors, at f22-28.

There were only two Tessar types, the Nikkor Qs, which are older single-coated pre-M variants in a factory Copal 3 mount. They are certainly usable optics but that lower ranking surprised me as well. However, the sample size is too low to draw any conclusions about Tessars generally. I surely would have liked a modern Nikkor M or a Commercial Ektar, but did not have any.

There were enough Dialyte variants to form a broad, generally favorable conclusion about them in terms of consistently good sharpness in longer focal lengths. (Two 203/7.7 Ektars, a 14" Red Dot Artar, and a 360mm Kern Apo, with a 305mm Schneider Repro-Claron enroute.)

One modern Dagor, the 7", was exceptionally good, near the top in fact, and one Dagor, the 6" was pretty good, but not quite as sharp as the 7". The third Goerz Dagor, an old German 12" in an Acme #4 shutter, would be good enough for 8x10 and 11x14, but both too bulky and too soft to be optimal on smaller formats. The 90mm Goerz American Wide Angle Dagor was better than expected but still not quite up to the more modern UWA Nikkor and Fujinon lenees. Of the two Dagor variants made by Schneider, the early 240mm G-Claron was very sharp. The 90mm Angulon performed as expected for a good copy, sharp enough for field work, but the newer wide-angle lenses were very noticeably better.

The three tested G-Clarons (including the 240mm Dagor variant) and the three Protar VIIa lenses showed similar variation among focal lengths. Although this may simply be copy to copy variation or small sample size, there was more variation than shown by the more modern Fujinon W Plasmats. I also have a Protar VIIa 12"-19"-23" set in a CLA'd Volute and a Protar V that I have't tested yet because they are currently unmounted. These and the 12" Dagor are part of a currently unused 11x14 outfit. I will add the other Protars and the 305mm Repro-Claron when I have a chance to test them as well

[Joseph Kashi]

Oh sorry... DV = Dependent Variable... that which was measured. What did you measure to establish this ranking? I’m not picking on your ranking... just interested in your method.

Interestingly, I don’t think you included a single lens that I use. :)

I measured visually perceived acutance and resolution directly on Delta 100 5x7 negatives on a light table with a loupe. I used sufficient ISO 12233 targets (modified to include some additional continuous tone and other imagery), with the targets laid side to side to cover the horizontal field of view. Each negative was scored on an arbitrary numeric scale, compared with similarly scoring negatives, and then sorted in rank order.

Although this sounds too subjective in some ways to be entirely satisfactory, you could think of it as consistent with expected real-world usage and as a crude sort of Bayesian approach. Mostly, I wanted to cherry pick the best lenses of the lot for a studio 5x7 kit and for light and medium weight 5x7 and 4x5 field kits. Putting these results out was very much an after-thought to the practical sorting out of lenses.

Multiple negatives were made of nearly all lenses and the sharpest/highest acutance negative for each lens was selected for ranking to reduce the effect of any poor technique on my part.

What most surprised me was that I re-evaluated these many final negatives three times over several weeks and the scoring and ranking were pretty consistent each time, which gave me a bit of confidence that this was not so subjective as to be useless.

[Bob Salomon]

I measured visually perceived acutance and resolution directly on Delta 100 5x7 negatives on a light table with a loupe. I used sufficient ISO 12233 targets (modified to include some additional continuous tone and other imagery), with the targets laid side to side to cover the horizontal field of view. Each negative was scored on an arbitrary numeric scale, compared with similarly scoring negatives, and then sorted in rank order.

Although this sounds too subjective in some ways to be entirely satisfactory, you could think of it as consistent with expected real-world usage and as a crude sort of Bayesian approach. Mostly, I wanted to cherry pick the best lenses of the lot for a studio 5x7 kit and for light and medium weight 5x7 and 4x5 field kits. Putting these results out was very much an after-thought to the practical sorting out of lenses.

Multiple negatives were made of nearly all lenses and the sharpest/highest acutance negative for each lens was selected for ranking to reduce the effect of any poor technique on my part.

What most surprised me was that I re-evaluated these many final negatives three times over several weeks and the scoring and ranking were pretty consistent each time, which gave me a bit of confidence that this was not so subjective as to be useless.

Tested within the optimization range of the lenses? Optimal magnification ratio? If not for copy work at different field curvatures?

[Joseph Kashi]

Tested at 1:20, approximating my normal working distance and with flat field target perpendicular to lens axis. I was basically sorting out which lenses would work best for my own use and working style.

[Bob Salomon]

Tested at 1:20, approximating my normal working distance and with flat field target perpendicular to lens axis. I was basically sorting out which lenses would work best for my own use and working style.

Then why flat field?

[Joseph Kashi]

Then why flat field?

Sorry, perhaps I mis-phrased, apparently, in responding to what I thought was a question about testing for "field curvature". I used standard flat test targets and did not separately test for field curvature beyond what could be observed at the edges and corners.

[Drew Wiley]

Thanks for your effort, but unless you had a true optical bench system and an ability to keep film truly flat (which no ordinary holder provides), you've got some misleading variables inherently involved. There are quite a few other variables to contend with too; so alleged objectivity depends on how one defines the rules. You'll no doubt hear other such comments. But you obviously have plenty of lenses to choose from for you own preferred usage, and that's what counts. In large format work, the particular "rendering" or "look" a lens provides is often more important than nitpicking sharpness.

[cowanw]

You have obviously put a great deal of effort into this and I applaud that. For portraiture I have done something similar with my lenses, this for babies, this for 20 somethings, this for 40 somethings and the Verito for 70 year olds and so on. Broadly speaking, descriptive personal preference and experience can be useful and certainly has validity. Well done.

[Mark Sawyer]

Waiting for the resolution tests of Verito vs. Plasticca vs. Pinkham & Smith...

[cowanw]

I rank them by age and sometimes sex of the sitter.
Verito at the far end, Kodak commercial for newborns , Heliars for twenty year olds, Kodak portrait or Imagon for thirty year olds, and the forty to fifty age group has to be individualized. Add 10 years for smokers. Pinkham are best overall.
Your mileage will vary.

[Sal Santamaura]

...unless you had a true optical bench system and an ability to keep film truly flat (which no ordinary holder provides), you've got some misleading variables inherently involved. There are quite a few other variables to contend with too; so alleged objectivity depends on how one defines the rules...

A reprise of Drew (who knows everything about everything) missing the point and not understanding that how LF lenses are used in real life is what's important. We've been through this -- painfully -- before:


https://www.largeformatphotography.info/forum/showthread.php?71322-f64&p=1321449&viewfull=1#post1321449

At the end, I noted that everything Drew posts must be evaluated critically and, when appropriate, ignored.


...You'll no doubt hear other such comments...This is post #17. So far, yours is the only one.


...In large format work, the particular "rendering" or "look" a lens provides is often more important than nitpicking sharpness.In large format work, sharpness is often the driving factor motivating use of larger film. Wasting that potential on unsharp lenses is self defeating.

[Joseph Kashi]

The reason that I took my personal results and started this thread was because I have noticed over the years that there's not much in the way of even basic comparative LF lens data on the web since Perez and Thalmann twenty years ago, although there's much strongly expressed opinion. I look forward to others providing more comprehensive and more precise data to the LF community in the future.

My broad take-away is that most modern-era lenses are more than good enough for practical use. Secondarily, reiterated comparative tests help us find subtle discrepancies such as small mis-alignments and improve the precision of our daily technique.

I used only rank-ordering because I made these tests as own-system practical tests similar to those Zone System tests that we made at MIT with Minor White many years ago. Yes, I have taken courses on experimental design when I was an undergraduate and then graduate student at MIT, I do understand how to design and conduct proper experiments, and can handle apparatus without breaking it. Living in a semi-rural part of Alaska 150 miles out of Anchorage, the nearest major city, I also understand the practicalities of use outside the lab.

Because these are practical usage tests, it seemed sensible to make them with the same equipment that would be used in the real world and with a similar (and verified) technique comparable to that I would use in the field. Were I publishing data in a peer-reviewed journal, the tests would have been done rather differently, of course, and I would have given numerical data as precise as I could make it. But, I'm not Roger Cicala and LensRentals, and, Roger's not testing LF lenses nor publishing LF lens MTF and other data. Also, to minimize copy to copy variation, Roger tries to test ten new copies of each lens, something not feasible for anyone in the LF community these days.

In the end, the reason that most of us even care about the relative quality of a lens is to make real-world photos. That's easiest to do by starting out with the sharpest, most contrasty lens possible - such lenses provide maximum information on the negative for later use/manipulation. If the lens is not a good one, then the information is never there and later analog/digital processing can't compensate for lost data.

[BrianShaw]

... :)

[Drew Wiley]

There's a lot more to it than that. Yes, extreme sharpness might be the priority if I want to use the same lens in my 8x10 kit for roll film holder shots destined for a much higher degree of print magnification. But high contrast can sometimes be a detriment, especially to color transparency work. I'm sure the usual mosquito will bzzzz in to contradict anything I say, but whatever ..... "Quality" involves a lot of subjectives, which are very difficult to "rank" unless everybody thinks and composes photos in exactly the same manner; and thank goodness, we don't. Don't get me wrong; I once specialized in extremely sharp large Cibachrome prints. But that's just one side of the coin.

[Joseph Kashi]

There's a lot more to it than that. Yes, extreme sharpness might be the priority if I want to use the same lens in my 8x10 kit for roll film holder shots destined for a much higher degree of print magnification. But high contrast can sometimes be a detriment, especially to color transparency work. I'm sure the usual mosquito will bzzzz in to contradict anything I say, but whatever ..... "Quality" involves a lot of subjectives, which are very difficult to "rank" unless everybody thinks and composes photos in exactly the same manner; and thank goodness, we don't. Don't get me wrong; I once specialized in extremely sharp large Cibachrome prints. But that's just one side of the coin.

Hi, Drew: I concur with everything you note in the above post. It's a matter of the right tool for the job. My starting point is high sharpness and good acutance in a relatively flare-free lens, but that's just a starting point.

I would not want my portrait taken at this stage of life with such a lens and I will likely still prefer some of those medium-resolution uncoated Protar VIIa lenses when I need to have deep shadow detail boosted out of Zone 1-2 in a black and white landscape.

[Arne Croell]

The reason that I took my personal results and started this thread was because I have noticed over the years that there's not much in the way of even basic comparative LF lens data on the web since Perez and Thalmann twenty years ago, although there's much strongly expressed opinion. I look forward to others providing more comprehensive and more precise data to the LF community in the future. ...

Tooting my own horn here, but I have done a few tests years ago, found on my web site: http://www.arnecroell.com/lenstests.pdf

It is also linked in Dan Fromm’s list of lens resources here on the LF forum lens page.

[Joseph Kashi]

Hi, Arne:

Now that you mention it, I do recall seeing this very helpful set of test results some time back. My error and my apologies - I probably should have recalled it. I don't want to think about why I forgot your tests .. .....

[pgk]

There's a lot more to it than that. Yes, extreme sharpness might be the priority if I want to use the same lens in my 8x10 kit for roll film holder shots destined for a much higher degree of print magnification. But high contrast can sometimes be a detriment, especially to color transparency work. I'm sure the usual mosquito will bzzzz in to contradict anything I say, but whatever ..... "Quality" involves a lot of subjectives, which are very difficult to "rank" unless everybody thinks and composes photos in exactly the same manner; and thank goodness, we don't. Don't get me wrong; I once specialized in extremely sharp large Cibachrome prints. But that's just one side of the coin.
Lens testing is one of the most misunderstood aspects of photography. To be really meaningful, tests should be 'targeted' in that they should be designed to provide data which is useful for the application in which they are to be used. In my experience this is rarely the case because applications are rarely defined tightly enough. So we end up trying to quantify a very complex set of characteristics in far too simplistic ways. And then as the quoted post, states there are the 'subjectives'. So any test results should be heavily caveated. That said, lists like those from the OP often seem to have particular lenses nearer the top so there are certainly trends which must say something.

FWIW I used to MTF test lenses many years ago. My experience was that the higher the trace on the graph, the happier to customer. Whether this actually relates to anything else depends on variables such as the customer's understanding of what the graph meant in practice for what they were doing.

And rather bizarrely, one photo which I sell as a greetings card was taken on an 1860s lens. I suspect that it sells because it is a pretty shot which buyers cannot easily produce for themselves.

Things are more complex that simply being the 'best', 'worst' or 'acceptable', etc., etc..

[Bernice Loui]

See LFF post:
https://www.largeformatphotography.info/forum/showthread.php?155445-Easy-Testing-Used-LF-Lenses&p=1527758#post1527758

Image results from any lens test and acceptance-ranking of lens tested is quite subjective and VERY complex. To over simplify lens rankings by resolution and contrast alone tells only a fraction of what the personality of any individual lens might be. Getting to know and understand any individual lens is to develop a relationship with that lens to know if you're both agreeable and will enjoy each others company making images.


Bernice

[Peter De Smidt]

The perfect is the enemy of the good. Great job, Joseph!

[Mark Sawyer]

Image results from any lens test and acceptance-ranking of lens tested is quite subjective and VERY complex. To over simplify lens rankings by resolution and contrast alone tells only a fraction of what the personality of any individual lens might be. Getting to know and understand any individual lens is to develop a relationship with that lens to know if you're both agreeable and will enjoy each others company making images.

Agreed. Raw resolution is only one factor, but often still an important one, and it's the one that impresses most people, perhaps because it's the one that can be expressed by a simple number. For those of us who contact print, a lens of even modest resolution can render more detail than the eye can see.