lens testing

[RobC]

Leica show 5, 10, 20 and 40 cycles/mm

Zeiss show 10, 20 and 40 cycles/mm

Zeiss also use lenses taken from the production line to measure MTF which is shown in the lens data sheets. Apparently Leica don't so their charts are ???

Also Leica by using 5 cycles/mm in their charts always have a close line to 100% which to anyone who doesn't quite get MTF charts (which includes me) can be highly misleading.

For me its the 40 cycles/mm line that dictates mostly how good a lens will be for overall resolution. Am I wrong?

 

[Nodda Duma]

That's good stuff RobC. Schneider also reports as-built values by the way. I'd prefer to read that data. It'd be interesting to measure a Leica lens for which nominal (ideal) MTF data is available to see how real performance compares.

No you're not wrong at all. Most important: Equate those values to what you've seen in your experience to benefit your own evaluation. The reported values depend entirely on your application. I suppose 40 lp/mm is a fair value for evaluation of photographic performance of a 35mm camera, though there's obviously no standard. That value equates to imaging each line of the pair onto a 12.5 micrometer pixel pitch or spot (1/40 divided by two..remember it's line pairs). So you can get a feel for the contrast at the equivalent imager resolution. I used 40 lp/mm on a recent design effort. On another I tracked it out to 220 lp/mm. On still another I only tracked to 15 lp/mm. All dependent on the imager resolution the lens would be used for.

The key is that you can suitably interpret the data and how it translates to how you would use the lens in the real world

 

[Dan Fromm]

Ok fair enough. It seemed that he was saying you don't need QC at all if the parts are in all tolerance,

That was Ralphie, not me.

 

[Nodda Duma]

Great posts, Nodda Duma!!

Every time i dissasemble a lens for cleaning, i wonder if i am not decentering the lens when i assemble it back... But most of those lenses seem to have a very tight fit for the elements inside their locations!

Here's something to consider: there's a 99% chance the lens you take apart was designed to be assembled on a production floor with minimal amount of labor. So taking them apart and putting them back together will maintain the decenter within the original acceptable tolerance. So disassemble and reassemble with confidence that the lens will still perform as intended.

Additionally, in my experience mounting decenter isn't always the driver of performance. Wedge and airgap are also big drivers, so changing the decenter within the acceptable tolerance will only perturb the MTF a minor amount. That's what I've seen in testing and analysis.

 

[Jim Jones]

For DIY types, a precision test chart can be projected by a lens onto a large screen to instantly detect obvious flaws. I believe this was one of the tests used long ago by Leitz. In the early 1970s I used a 2" square microfilm that included the complete Bible as a test chart for a quick check of 35mm camera lenses. The results were obvious, although not quantitative: numerical values have little direct application in practical photography. Of 30 or 40 lenses tested, four were distinctly superior: El-Nikkor 50mm f/2.8, Micro-Nikkor 55mm f/3.5, 45mm GN-Nikkor, and Elmar 50mm f/2.8. Most of the standard fast 50mm lenses fell slightly behind that first group of four. I don't recall how most wide angle and telephoto lenses fared. An early Nikkor 8 element f/4 21mm was very sharp in the center, but dismally unsharp at the edges. Of course there were several critical qualities that such a test doesn't evaluate, such as back focus or rangefinder coupling.

 

[Nodda Duma]

Hi Jim,

That's a great qualitative test that I've used in the past for quick-check of performance at short object distances. Just one caveat that it tells you what the performance is at the distance you are projecting... Typically short-range. The performance can be slightly different at long-distance.