Leica T 18-56 versus Olympus 12-40

[tgm]

Advertisement (gone after registration)

How on earth can one compare the MTF graphs of lenses for different sensor formats?

 

I think the plots are not MFT graphs, please look on the scale, in other words they are indeed comparable.

[01af]

Do I read this properly?

No, you don't.

 

 

Jaap is quite right—you certainly can't compare MTF figures across formats.

Indeed you can't—but ... there are no MTF figures here. Of course, neither you can compare resolution figures across formats and/or across different cameras.

 

So this whole thread is totally pointless. The Olympus M.Zuiko 12-40 mm is an excellent zoom lens, and so is the Vario-Elmar-T 18-56 mm Asph. If you want to know which of these two bests the other then you have to try them yourself, on the very same test subjects, and compare carefully. Looking at other people's test results won't help.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jonoslack]

No, you don't.

 

 

 

Indeed you can't—but ... there are no MTF figures here. Of course, neither you can compare resolution figures across formats and/or across different cameras.

 

So this whole thread is totally pointless. The Olympus M.Zuiko 12-40 mm is an excellent zoom lens, and so is the Vario-Elmar-T 18-56 mm Asph. If you want to know which of these two bests the other then you have to try them yourself, on the very same test subjects, and compare carefully. Looking at other people's test results won't help.

 

Hi O1af

I've done this - I wasn't very much the wiser - as you say, they're both good lenses.

 

. . . . and I never post results from these posts as I'm not rigorous enough to make it meaningful.

[jdlaing]

No, you don't.

 

 

 

Indeed you can't—but ... there are no MTF figures here. Of course, neither you can compare resolution figures across formats and/or across different cameras.

 

So this whole thread is totally pointless. The Olympus M.Zuiko 12-40 mm is an excellent zoom lens, and so is the Vario-Elmar-T 18-56 mm Asph. If you want to know which of these two bests the other then you have to try them yourself, on the very same test subjects, and compare carefully. Looking at other people's test results won't help.

 

No thread is ever pointless. One bit of information and it's worthwhile .....

[IkarusJohn]

No thread is ever pointless. One bit of information and it's worthwhile .....

 

 

I've benefited from both Olaf's and Jono's contributions ...

[JonPB]

Indeed you can't—but ... there are no MTF figures here. Of course, neither you can compare resolution figures across formats and/or across different cameras.

 

So this whole thread is totally pointless. The Olympus M.Zuiko 12-40 mm is an excellent zoom lens, and so is the Vario-Elmar-T 18-56 mm Asph. If you want to know which of these two bests the other then you have to try them yourself, on the very same test subjects, and compare carefully. Looking at other people's test results won't help.

 

 

01af,

 

Respectfully--

 

I find your post to be disturbingly irrational.

 

These are very likely MTF graphs, just in a different format from the kind Leica sometimes publishes. Of course, Imatest is highly configurable and these charts do not specify all of their parameters, but it is likely that these are MTF50 graphs, varied by image-height rather than absolute-width-on-sensor and by aperture rather than distance from axis. The charts themselves do not provide sufficient information about their parameters to definitively say.

 

The fact that these are image-height relative rather than absolute-width-on-sensor relative means that they are directly comparable in imaging terms. (Comparing absolute-width measurements requires scaling by a factor of the sensor size difference.)

 

This graph does not specify what the "edges" are. If they're the near edges, far edges, or average, they will favor one format or another depending on how one values the different zones of the different aspect ratios.

 

There is one point you're right about: we can't compare lenses by way of Imatest, which compares images produced by a specific sensor. However, we can compare imaging systems, in which case this is a valid comparison of an Olympus lens + camera and a Leica lens + camera. And, the further implication is that we can't compare lenses by way of careful examination of images, either.

 

This is all subject, of course, to repeatability by independent sources, which (in theory) will in the long run average out variations in device manufacture and testing practice, of which this particular example may be representative of or an extreme outlier: we have no way of knowing.

 

But, really, what I object to is the idea that "looking at other people's test results won't help." This completely negates the scientific method. Other people's tests need to be interpreted in light of conditions such as I have mentioned above, but they can tell you things. Other people's tests cannot tell you how something will meet your needs, but that isn't what you wrote.

 

I'm a fan of superb photographic optics generally and Leica optics specifically. In this case, this test--with all its limitations, most notably that it only measures one facet (MTF50) of one particular aspect of image quality (resolution) and that it uses a statistically insignificant sample size--shows that the Olympus imaging system outperforms the Leica imaging system. To say anything else is to denigrate the work of Leica engineering by erecting and defending a straw man next to the true genius of their labor.

 

I have not carefully examined the two lenses. But I know that, while I value sufficient resolution at MTF50, I also value other attributes not address here, such as resolution at MTF80 or MTF20, mention distortion, astigmatism, lateral and longitudinal chromatic aberration, out of focus rendering, flare and haze, focal plane curvature, and other factors and their mutual interplay.

 

So, I say that these charts show one sample point that the Olympus outperforms the Leica. Whether that tells you anything that is meaningful to you is for you, and only you, to decide; the same applies to me, and to every other individual. Personally, I believe that this small point is meaningless in terms of my evaluation of the two optical systems, but that does not mean it is an invalid datum.

 

As I said, I'm a fan of Leica. But I'm even more a fan of rational thought.

 

Regards,

Jon

[01af]

Advertisement (gone after registration)

I find your post to be disturbingly irrational.

Irrational!? Huh?

 

You could have said that you don't agree ... but "irrational"? Now that's funny.

 

 

These are very likely MTF graphs, just in a different format from the kind Leica sometimes publishes.

They most definitely are not MTF graphs in another format but resolution graphs. They may be derived from MTF data in some way or another ... or not. We can't say. Either way, what we see are resolution graphs, not MTF graphs.

 

 

The charts themselves do not provide sufficient information about their parameters to definitively say.

Exactly.

 

 

The fact that these are image-height relative rather than absolute-width-on-sensor relative means that they are directly comparable in imaging terms.

No, it doesn't. The data presented are incomplete (as you just said), hence meaningless, hence incomparable to any other other kind of lens performance data.

 

 

There is one point you're right about: we can't compare lenses by way of Imatest, which compares images produced by a specific sensor. However, we can compare imaging systems, in which case this is a valid comparison of an Olympus lens + camera and a Leica lens + camera. And, the further implication is that we can't compare lenses by way of careful examination of images, either. This is all subject, of course, to repeatability by independent sources, which (in theory) will in the long run average out variations in device manufacture and testing practice, of which this particular example may be representative of or an extreme outlier: we have no way of knowing.

So if you agree with me about this vital point then why are you calling my post "disturbingly irrational"?

 

 

But, really, what I object to is the idea that "looking at other people's test results won't help." This completely negates the scientific method.

No, it doesn't. Obviously you believe that every lens tester always was applying scientific methods, and properly so. I negate neither science nor scientific methods. Instead, I negate most lens testers' ability to apply scientific methods properly—even when they publish neat-looking charts.

 

And even if two lens testers both are working carefully and accurately then their results still won't be comparable. Lens testing is just too complex. If you want the test results for two different lenses (or lens-and-camera combos) to be comparable then the tests have to be carried out after the same method by the same person on the same day ... which basically means that you have to do it yourself.

 

I am very serious: Do not try to assess the relative performance levels of two or more lenses after other persons' test results. The lens that looks worse according to test data may still be the better lens in your hand, for your work, on your camera.

 

 

In this case, this test—with all its limitations, most notably that it only measures one facet (MTF50) of one particular aspect of image quality (resolution) ...

Err ... the facet measured being MTF50 is just an assumption of yours, right?

 

 

... and that it uses a statistically insignificant sample size—shows that the Olympus imaging system outperforms the Leica imaging system.

No, it doesn't. System performance is far more than just one single number describing one single facet being greater than some other single number describing a similar facet.

 

 

To say anything else is to denigrate the work of Leica engineering by erecting and defending a straw man next to the true genius of their labor.

Huh!? Didn't you say you were a fan of rational thought? So why this irrational loss of rationale here? When I say a test result was meaningless then I don't mean the contrary result was correct. In fact, I don't even mean the test result was wrong. So if you are a fan of rational thinking then please don't fly in the face of plain logic.

 

To be clear: I don't think it's impossible for an Olympus lens to outperform a Leica lens. In fact, I own the M.Zuiko 12-40 mm lens, and I do think it's the greatest zoom lens I ever used. I'm just saying the test charts referred to above are meaningless, that's all.