35mm lens resolution measurements

[MattKing]

Unfortunately, virtually all published lens tests are done on sample size one (or, at very best, n<10) and are therefore completely meaningless.

Well, they might permit you to show off to your friends .
Assuming you have nerdy friends.
Or for that matter, assuming that lens testers have friends .
More practically, they can tell you if the lens you bought on eBay is performing reasonably.

 

[Sal Santamaura]

Once you test your lens, then what?

You test the next one, etc., etc. Then you sell off those that aren't "good."

 

[Paul Howell]

10 maybe years ago, maybe longer I tested all my 50mms, Konica, Yashica, Mamyia, Pentax, Miranda, Minolta MD, Sigma SA, Minolta A mount 50, Petir, Kowa, Canon QL 1.7, Konica S3 with my last few rolls of microfiche film, used Air Force test chart and my son's vintage microscope. I have the test results stashed way, in a storage locker, need to find them. I also used a color test chart, and contrast test target. Here are a couple of my charts, as you see they have seen better days, I need to remount them as the foam core has been badly treated. As you can see they are not properly set up. My tests were at best a fair estimate and were not as technical as the tests performed by Modern and Pop. I don't have a test bench used my patio, overcast day, the test charts were on a easel, 90 degrees, camera on tripod, at recommended distance.

I don't have the exact numbers from the Air Force test target as my notebook is packed away, but all the lens I tested could resolve Tmax 100. The Sigma 1.4 not the Art version the older version was the clear winner, of my legacy lens, the Konica lens, 50 17 in AR mount and 38mm on the S3 were the best. The Kowa and Petri did much better than I would have thought, of course Minolta and Pentax were very close to the Konica. For color I used Ektar 100, I thought the newer lens had the best color and contrast, the Sigma 1.4 was outstanding as was the Minolta A mount. The Petri and Kowa were softer, but pleasing.

I did not have a Nikon, Canon SLR or Leica to test, and wish I had tested at least one of my 35mm point and shoots.
What I learned is that modern lens are sharp, contrast and color rendering are the most variable and subject to personal preferences.

 

[Alan Johnson]

http://www.photodo.com/category_2.html
Some modern magazines present resolution as LW/PH or line widths per picture height.
Assuming a picture height of 24mm, to convert LW/PH to lppm (line pairs per millimeter), divide LW/PH by 48

 

[faberryman]

You test the next one, etc., etc. Then you sell off those that aren't "good."

I thought you kept the ones that aren’t “good” because they have character.

 

[AgX]

Unfortunately, virtually all published lens tests are done on sample size one (or, at very best, n<10) and are therefore completely meaningless.

But this is the issue with all consumer tests.
To be indepent of the manufacturer the samples must be acquired in the wild, which means bought. And with a 2-digit sample size such buy will often be expensive, as will be all these tests.
If you know of consumer tests which such a of number of samples per model, I would like to know.

 

[Les Sarile]

I test all my used gear so that I would know what I can expect from them for my own use.

I was curious just how much detail some super cheap local listed used lenses I purchased can achieve? I tested one of them - the Pentax 50mm f4 macro using Kodak Techpan @ ISO25 processed in Kodak Technidol. I was surprised to learn that this lens was not going to be the weak link when it comes to capturing detail onto the film. Most likely if I didn't use a tripod, have optimal conditions, a subject with sufficient detail, good focus and good film that I may not get the most out of it. And of course extracting the detail to make good use of it would be the challenge . . .

Resolution testing my SMC Pentax-M 50mm F4 macro lens by Les DMess, on Flickr

Left bottom is the target and 100% cropss above it using DSLRs and Coolscan 4000dpi. The big 100% crop to the right is about a 4.5X optical magnification of the center area that clearly shows the detail I couldn't extract with the methods I used.

I tested my other bought cheap manual focus lenses and those are not going to be the weak link when it comes to achieving detail . . .

 

[BradS]

But this is the issue with all consumer tests. ...

Yes, obviously.

I consider it a devastating failure of modern consumerism that the vast majority of people...

• fail to understand basic statistical concepts and,
  
• fail to understand that design, engineering and mass production is all about economics and,
  
• fail to understand the very most basic concepts of economics.

...but, these failure are exactly what enables the mindless consumption of goods. Indeed, it is the primary motive force of irrational over consumption.

 

[faberryman]

I don’t see how those factors meaningfully affect the decision of your neighbor to buy a bigger lawn mower that you have. I’d wager the answers lie in psychology rather than economics and statistics.

 

[BradS]

I don’t see how those factors meaningfully affect the decision of your neighbor to buy a bigger lawn mower that you have. I’d wager the answers lie in psychology rather than economics and statistics.

Yes, obviously. They FAIL to understand basic concepts of economics and statistics....

 

[wiltw]

http://www.photodo.com/category_2.html
Some modern magazines present resolution as LW/PH or line widths per picture height.
Assuming a picture height of 24mm, to convert LW/PH to lppm (line pairs per millimeter), divide LW/PH by 48

The term measured is explained as

• Line Widths or Pairs per Picture Height (LW/PH or LP/PH)

 

[Sal Santamaura]

Once you test your lens, then what?

You test the next one, etc., etc. Then you sell off those that aren't "good."

I thought you kept the ones that aren’t “good” because they have character.

Note that I put "good" in quotes. Some people consider "character" to be "good." Other people's aesthetic preferences are exactly opposite.

 

[flavio81]

In my opinion, best tests are pictorial tests that show real-life images and let your eyes be the judge.

The site http://www.artaphot.ch has very good tests of various lenses under identical conditions, although only for landscape (infinity) use, and he often looks only at the corners. Long story short -- often the Minolta and Canon FD lenses are the best in terms of sharpness and contrast, many times equalled (or closely followed) by the Nikkors and Pentax lenses, behind comes the other good brands like Konica, Olympus, Yashica, etc. And far, far behind, the third-brand/discount lens makers.

The camera repairer "Richard Haw" (google his website) has many lens tests of Nikkors and IMO shows how lens tests should be done -- in actual real life usage, looking also at distortion, rendering, flare, ghosts, vignetting, CAs, etc.

 

[flavio81]

PS:

http://www.artaphot.ch/systemuebergreifend/objektive

For example:

http://www.artaphot.ch/systemuebergreifend/objektive/612-135mm-1-2-and-1-2-5-lenses

 

[wiltw]

Yes, obviously. They FAIL to understand basic concepts of economics and statistics....

Gosh I would be happy if most folks understood the uses of

• they're vs. there vs. their

so many cannot get a simple concept, 'contraction leaves out a letter or two...substitute an apostrophe'
So more complex concepts are hopeless.

 

[faberryman]

Yes, obviously. They FAIL to understand basic concepts of economics and statistics....

I still don’t understand why you think a failure to understand concepts of economics and statistic explains why your neighbor wants to have a bigger lawn mower than you have.

 

[BradS]

I still don’t understand why you think a failure to understand concepts of economics and statistic explains why your neighbor wants to have a bigger lawn mower than you have.

I don't own a lawn mower.

 

[Alan Johnson]

The term measured is explained as

• Line Widths or Pairs per Picture Height (LW/PH or LP/PH)

See the section on Resolution here:
https://www.wavechaser.xyz/optics-imaging/2019/12/ultron50
Line widths is 2x Line Pairs.

 

[Dali]

See the section on Resolution here:
https://www.wavechaser.xyz/optics-imaging/2019/12/ultron50
Line widths is 2x Line Pairs.

I like the last sentence: "This test does NOT represent all Ultron units, and DO be careful buying should you choose to do so."

 

[RMD]

Simple non technical soul that I am I liked the tests done weekly by Amateur Photographer magazine in the U.K.
They used to shoot a series of shots of a permanently moored ship on the river Thames from their office window on the
opposite bank.

 

[btaylor]

Long story short -- often the Minolta and Canon FD lenses are the best in terms of sharpness and contrast, many times equalled (or closely followed) by the Nikkors and Pentax lenses, behind comes the other good brands like Konica, Olympus, Yashica, etc.

If you can’t get a great picture with any of the lenses made by the manufacturers listed above, the problem certainly isn’t your gear.
My 35mm f3.5 Zuiko vignettes significantly, I should get a better one I guess- but I can make it work.

 

[wiltw]

See the section on Resolution here:
https://www.wavechaser.xyz/optics-imaging/2019/12/ultron50
Line widths is 2x Line Pairs.

I merely quoted what was stated on the photozone.de website.

 

[beemermark]

Once you test your lens, then what?

You realize it's terrible and sell it here, cheap.

 

[whojammyflip]

You can do this yourself nowadays with MTF Mapper. Sharpness is quantified by the MTF curves, which is the relationship between contrast and spatial frequency of an image. You need to be able to mount the lens on a digital camera body, and have a print shop print an A1 target and then mount this on a flat board. Its easy to do this wrong and have to start again.

https://sourceforge.net/projects/mtfmapper/files/windows/

A simple alternative is to print some text on a sheet of paper, with decreasing font size, and then see at what point the text is no longer readable with your lens. You can calculate the effective resolution limit from that. I guess the two methods will be correlated, but MTF Mapper does a whole lot more.

 

[alanrockwood]

If one can determine the modulation transfer function (MTF) of a lens it is possible to calculate the resolution. In fact, the modulation transfer function is contains more information than a simple resolution specification, so in a sense the MTF is even better that a simple resolution spec.

There is a method of determining the MTF to a good approximation. It is called slant edge MTF measurement. It is widely used to determine resolution of optical systems, and it's actually not too hard to do because there are canned programs out there that can do the calculation. All one has to do is to acquire an appropriate slant edge image.

Jackson K.M. Roland has shown that the slant edge MTF method is not influenced much by noise or the contrast ratio of the image. Google his write up titled "A Study of Slanted-Edge MTF Stability and Repeatability". I think that one implication of this work is that it should be possible to print slant image targets on an ordinary inkjet printer and photograph those with a digital camera to determine the resolution of the lens on the camera.

This obviously implies that one must have a decent lens-to-camera adapter if the lens mounts don't match between the lens and the camera. I think a mirrorless 4/3 body camera would be ideal for this application because there are a lot of adapters out there to mount a full-frame lens on a mirrorless 4/3 body.

ImageJ is a free program image processing program that has an MTF method available. I have used it, but I don't remember if it was a plug in or a built-in application.

One interesting feature of the slant edge method is that it allows one to measure the resolution of a system, even if the sensor pitch does not satisfy the Nyquist bandwidth limitation. This means that you don't need to have a really high megapixel camera in order to do the measurement.

I'm not going to try to discuss the magic that allows one to beat the Nyquist limit, but basically I believe it has something to do with the fact that imaging of the slant edge basically makes it possible to (conceptually speaking) interpolate between the sampling points without introducing aliasing, at least up to a point. The phrase "phase offsets" often pops up in discussions of this.

One does have to make sure that all digital sharpening schemes are turned of everywhere in the system. I think some digital cameras automatically apply some sharpening, which could be a problem.