Measuring lens resolution...

[Misha_M]

Good evening!

Lens resolution in lp\mm is displayed at technical passport information for many older lenses.
For example, Helios 44M7 lens has resolution of 50/30 lp\mm centre\side.

How does one measure the lens resolution?
Can a test for resolution be performed at home, or is special precision equipment needed?

[hasan]

http://www.imatest.com/

and here infos on how-to:
http://www.imatest.com/docs/lens_testing/#download

[danfromm]

Testing resolution used to be easier. Just buy a USAF-1951 chart, set up camera, lens and chart as directed and shoot, process, and look at the negative with, ideally, a microscope and look up the smallest group and element that can be read.

For an alternative that's less expensive than Imatest and that gets to lp/mm, not MTF at some spatial frequencies, see http://www.normankoren.com/Tutorials/MTF5.html

Resolution figures given for Soviet lenses were measured with the lens wide open. I believe that the reason for this is to indicate the worst resolution a user can expect to get. I've never been clear on which target was used or on how the film was developed. All of these will affect measured resolution, as will exposure. There was a standard ...

[Misha_M]

Well, it is kind of logical, to measure wide open and get the worse, because it can only get better from that, so if your worst is better that the average, or even best, of your competitor, you will be much superior!

Anyway, can the testing be done with digital camera and not film?
Meaning, can I test on camera and print on A4 page?

[iangreenhalgh1]

Yes, you can do it with a digital camera, you don't need to print anything, you shoot a test chart then examine the RAW file.

I don't see any point at all in measuring resolution if you're going to be using the lens for normal photography, if it's sharp enough, it's sharp enough.

[scsambrook]

[twinquartz]

[Misha_M]

[iangreenhalgh1]

It really is a waste of time. What does it matter if a lens has 45 or 50 lp/mm central resolution?

[ForenSeil]

Because it's a waste of time which makes fun imho
I've also experiemented with that.

I've used a DINA4 test chart from Ebay which I currently can't find anymore for that. It directly showed lp/mm for a special fixed magnification (1:40 in that case). Stopped down with good 50mm lenses my NEX-5N sensor was outresolved that much that I had to use Efke 25KB film which worked well. But then lenses outresolved the test chart and I had to repeat the test at lower 1:80 magnification and recalculate the test results

Digital cameras generally have much lower resolution than low iso B/W film, wich makes it generally very hard to count lines in the tests charts.
Ultra high resolution lenses like Biogon 25/2.8 for example are impossible to get outresolved with any recent digital sensors (it would need about let's say 100 megapixels).
For Helios lenses normal digital sensors like from your camera might have just enough resolution, but only wide open only if at all, forget testing their max. center resolution at F5.6.
Another big problem with both digital and film is shutter vibration. Can be only avoided by very fast shutter speeds (1/500s) or special tricks like HSS flash setups etc..
Of course self-timer, tripod, m-mode, external ambience lightmeter/luxmeter or flashmeter for flash-setups, water-level, folding rule etc., fixed artificial light in 45° angle (to get a reproducable contrast),... have to be used to get reproducable "scientific" results.

But I found that a brickwall- and/or white wall & newspaper test gives more meaningful results - clarity, corner performance, distortion, fine contrast and sharpness are more important than pure resolution, especially on digital. Also some sources of multispectral light (candle light, halogen, tungsten, sun reflected by metal) in the setup give some very good feedbacks about CAs and flaring. Such a test setup is much easier, faster and gives more feedback at once.

Btw. I found that Helios 44-M7 was slightly softer than my 44-2, while it was supposed to be sharper in the center. The M7 was slightly better in the corners though. I think quality spreading might be one reason for that.

[Ray Parkhurst]

I like to use a 2x Teleconverter to aid in checking lens resolution. I've never cared to check quantitatively, only qualitatively. The teleconverter gives you double the spacial resolution, so the camera is not as limiting. It makes it easier to compare lenses, or to find optimum apertures when the lens is better than the camera...Ray

[scsambrook]

[twinquartz]

[kds315*]

Well, "testing lenses" once and foremost tests the ability of the tester to conduct a test correctly.
If you don't have experience with it - leave it to the professionals as the results will be at best
misleading, at 2nd best, embarrassing - not for the lenses but the tester.

Just talking about my own experience....

[iangreenhalgh1]

I agree Klaus. The sharpness you achieve in the end result has many factors, lens resolution being one of them, but I tend to think that it is less important than other factors such as technique.