50 Comparison: APO-Summicron trounces Zeiss Otus

[Paul J]

Advertisement (gone after registration)

A test and article from Roger Cicala at Lens Rentals has appeared. Testing with the Imagemaster Optical Bench reveals the Leica APO-Summicron trounces the Zeiss Otus. The Summilux also beats it, fairing exceptionally well.

 

The other wonder of the bunch is the Sigma ART, especially so at the price. Beating the Otus and Summilux at f2.

 

Find Roger's very well presented article here - LensRentals.com - Comparing Rangefinder and SLR 50mm Lenses. (Version 0.7)

[Jeff S]

http://www.l-camera-forum.com/leica-forum/leica-m-lenses/332355-50mm-f2-apo-asph.html#post2707958

[Advertisement]

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

[4X5B&W]

Interesting results, while I love my Summilux ASPH, I must say the test show what continues to amazes me……that my 1979 Leitz Canada Summicron is still a stunning lens.

 

While these new SOTA designs are impressive, the better the sensor, the better older designs like the Ver. 4 Mandler Summicron look

[asiafish]

Interesting results, while I love my Summilux ASPH, I must say the test show what continues to amazes me……that my 1979 Leitz Canada Summicron is still a stunning lens.

 

While these new SOTA designs are impressive, the better the sensor, the better older designs like the Ver. 4 Mandler Summicron look

 

Agreed. My 50 Summicron (v5) performs brilliantly on the M Monochrom and gives up precious little to these modern expensive wonders.

 

Of course Roger mentions that the "lowly" Summicron at $2400 isn't so lowly when it comes to purchase price. Still, with the old Mandler Cron you still get what you pay for.

[tgm]

A test and article from Roger Cicala at Lens Rentals has appeared. Testing with the Imagemaster Optical Bench reveals the Leica APO-Summicron trounces the Zeiss Otus. The Summilux also beats it, fairing exceptionally well.

 

The other wonder of the bunch is the Sigma ART, especially so at the price. Beating the Otus and Summilux at f2.

 

Find Roger's very well presented article here - LensRentals.com - Comparing Rangefinder and SLR 50mm Lenses. (Version 0.7)

 

 

this review comes to a different conclusion:

 

Google-Ergebnis für http://farm8.staticflickr.com/7373/11687601664_62d4936880_o.jpg

 

 

any idea?

[menos I M6]

Careful with the test - Roger explicitly mentions the restrictions given with this test (lenses tested at infinity only) and his little experience with the new testing method, his procedures being still a development in progress.

 

What indeed surprised me was how so very good the old 50/2 Summicron design from the late 70's showed and how so close the Summilux ASPH is to the 50/2APO.

[tgm]

Advertisement (gone after registration)

Careful with the test - Roger explicitly mentions the restrictions given with this test (lenses tested at infinity only) and his little experience with the new testing method, his procedures being still a development in progress.

 

 

My suspection, Roger has determined the resolution were it is best, so field curvature is not taken into account.

[jaapv]

this review comes to a different conclusion:

 

Google-Ergebnis für http://farm8.staticflickr.com/7373/11687601664_62d4936880_o.jpg

 

 

any idea?

You'll have to provide some other link, the Google Images interference messes up this website and freezes it.

[pgk]

What indeed surprised me was how so very good the old 50/2 Summicron design from the late 70's showed and how so close the Summilux ASPH is to the 50/2APO.

At the high quality level of lenses you mention I would actually be surprised if the differences were substantial. To marginally increase performance beyond that of the 'older' lenses and current state-of-the-art lenses such as the Summilux-ASPH requires not only improved designs, but also less tolerance variation in component manufacture and very high standards of assembly and quality control too. I have a friend who designs lenses and with whom I have discussed increasing lens 'quality'. He has said for a long time that the precision to which lenses need to be built to improve their performance will result in escalating costs due to the complexity of ensuring sufficient precision in manufacture, assembly and QC. The Apo-Summicron would appear to indicate that he is correct.

 

Its also worth remembering that 'good' and very acceptable 50mm lenses can be bought for a fraction of the price of even an ordinary 50mm Summicron, so we are already comparing the very 'best' here.

[tgm]

At the high quality level of lenses you mention I would actually be surprised if the differences were substantial..

 

In the image center, I would agree, but wide open at the edges this seemd not to be the case, the differences are substantial:

 

Most Adorable 50s - Zeiss Otus & Sonnar, Leica Noctilux & Summilux, SLR Magic HyperPrime - Chart Test

 

At least for the summilux the results make a big difference if you focus at the center or at the edge. I think the main achivement of the most recent lenses such as Apo Summicron or Zeiss Otus etc. is the strong performance all over the field

[pgk]

In the image center, I would agree, but wide open at the edges this seemd not to be the case, the differences are substantial:

 

Most Adorable 50s - Zeiss Otus & Sonnar, Leica Noctilux & Summilux, SLR Magic HyperPrime - Chart Test

 

At least for the summilux the results make a big difference if you focus at the center or at the edge. I think the main achivement of the most recent lenses such as Apo Summicron or Zeiss Otus etc. is the strong performance all over the field

The problem with tests like the one you link to is that whilst it illustrates that there are differences in the flatness of field at a specific distance and aperture, it doesn't actually indicate 'real world performance. Its perfectly valid, but a 'snapshot' of a specific performance indicator.

 

When I changed from 50/1.4 pre-aspheric to aspheric, it was the edge performance which was distinctly better wide open which was most noticeable - in real world photos that is - not that I often use it wide open where it matters but on occasion I think that it does.

 

Apo flat field lenses are nothing new but in the past they have mostly been designed and used for very distance specific copying requirements which has probably helped significantly in their design. To design a general purpose lens and ensure flat field performance over all apertures and distances is far more problematic - as of course is testing to see whether is has such a characteristic.

[NightSun]

In the image center, I would agree, but wide open at the edges this seemd not to be the case, the differences are substantial:

 

Most Adorable 50s - Zeiss Otus & Sonnar, Leica Noctilux & Summilux, SLR Magic HyperPrime - Chart Test

 

At least for the summilux the results make a big difference if you focus at the center or at the edge. I think the main achivement of the most recent lenses such as Apo Summicron or Zeiss Otus etc. is the strong performance all over the field

 

Also one has to be cautious with such a test, as the guy made the comparison on a Sony A7R, which does not really work optimally with many M lenses when it comes to edge performance, even with some 50mm lenses. Actually, if I remember well, some months ago I read reports that Summilux 50 ASPH does not behave close to what it is capable of on a Leica M body. In reality, what such a test measures is the coupled performance of the optics comprising lens and filter/glass stack in front of the sensor. In addition image quality can be affected by a subpar lens adapter, as well. That means, if you were to do these tests on a Leica M, for example, chances are that you will get a different performance for some lenses compared to what you see here.

 

E.g. check this test done by Ron Scheffler, and look at Summilux 50 ASPH, there are also links to download images:

 

Sony a7R vs. Leica M9 infinity test with 20+ Leica, Zeiss, Voigtlander rangefinder lenses

Edited June 9, 2014 by NightSun

[tgm]

Also one has to be cautious with such a test, as the guy made the comparison on a Sony A7R, which does not really work optimally with many M lenses when it comes to edge performance, even with some 50mm lenses. Actually, if I remember well, some months ago I read reports that Summilux 50 ASPH does not behave close to what it is capable of on a Leica M body. In reality, what such a test measures is the coupled performance of the optics comprising lens and filter/glass stack in front of the sensor. In addition image quality can be affected by a subpar lens adapter, as well. That means, if you were to do these tests on a Leica M, for example, chances are that you will get a different performance for some lenses compared to what you see here.

 

E.g. check this test done by Ron Scheffler, and look at Summilux 50 ASPH, there are also links to download images:

 

Sony a7R vs. Leica M9 infinity test with 20+ Leica, Zeiss, Voigtlander rangefinder lenses

 

 

Yes, I agree, with such tests you always measure the performance of the combo lens plus sensor, but that's what counts at the end. What realy puzzles me, that the egde performance of the summilux in combination with the sony a7r substantially improves when focused to the edge. This is a clear signature of field curvature, and if you look on the MFT curves published by Leica, the edge performance of the summilux wide open is not so great. According to the MFT curves, the resolution of the apo summicron across the whole field is much better. This is in contrast to the test on the optical bench which was the starting point of this thread and what puzzles me.

[NightSun]

Yes, I agree, with such tests you always measure the performance of the combo lens plus sensor, but that's what counts at the end. What realy puzzles me, that the egde performance of the summilux in combination with the sony a7r substantially improves when focused to the edge. This is a clear signature of field curvature, and if you look on the MFT curves published by Leica, the edge performance of the summilux wide open is not so great. According to the MFT curves, the resolution of the apo summicron across the whole field is much better. This is in contrast to the test on the optical bench which was the starting point of this thread and what puzzles me.

 

As for the strange performance of Summilux 50 on the Sony A7R, I'm pretty much convinced that it is due to the thick glass cover in front of its sensor that induces astigmatism and field curvature (which is much greater than that of the lens). I downloaded the images taken by Ron Scheffler and they clearly show that Summilux on the Leica M9 is still sharp wide open at the edges, while on A7R you have the smearing effect, which is very similar to what can be seen at the linked 'Most adorable 50s' test also done on A7R, even though Mr. Scheffler did the test for infinity focus, while the other one is for close range.

 

Now, coming to the MTF curves and the optical bench test. Here confusion could come from the different way of displaying the results. Observe that Leica supplies MTF with respect to distance measured from the center in millimeters, while Roger Cicala's graphs provide angle off axis from 0 to 20 degrees along the 'x axis'. 20 degrees roughly corresponds to 18 mm from center, which means horizontal image edge in 36mm x 24mm rectangle. Now, if you check the MTF curves published by Leica for Summilux 50 ASPH, actually the big drop in the MTF curves happen between 17 mm and 21 mm, the latter corresponding to the image corners. But if you cut the curves at 18mm, then you would see that optical performance of the Summilux is still outstanding there. That's why I would dare to say that most probably things are about right in Roger Cicala's tests, if you mean image edges and not corners, the Summilux still bears the competition quite well against the APO Summicron.

[Adrian Lord]

I have to say these sorts of tests really turn me off, being almost the antithesis of photography.

[NightSun]

I have to say these sorts of tests really turn me off, being almost the antithesis of photography.

 

Then don't bother to read them, that's it.

 

In my honest opinion, though, it only becomes the antithesis of photography, if someone does not care about anything else than doing/publishing/reading such tests while buying into cameras and lenses. That is, taking a means as an end.

 

On the other hand, photography starts with optics and there should be people who care about the optics part on a very technical level. Some of them should profess photography as well, because evidently there should be an understanding that different optical qualities/performance/aberrations how do they translate into the image and what limitations they present for the image taking process. And, actually , what we see here, is that through different tests one can learn about various aspects of such optical performance and their consequences. For the photographer that could mean a more well-founded decision whether it's worth for him/her to buy a super expensive APO Summicron 50 lens, or stick to a Summilux, for example.

[pgk]

I have to say these sorts of tests really turn me off, being almost the antithesis of photography.

The problem is IMO, that if they were really comprehensive, these sorts of tests would be unbelievably tedious to wade through. IMO you 'learn' a lens through using it;). Translating technical information into its effect within images is often difficult and simple numbers or even graphs needs to be taken as indicators rather than as definitive scores:eek:.

[Adrian Lord]

Then don't bother to read them, that's it.

 

I don't!

[NightSun]

Translating technical information into its effect within images is often difficult and simple numbers or even graphs needs to be taken as indicators rather than as definitive scores:eek:.

 

For sure, you won't be able to fully describe the characteristics of a lens with sheer numbers from brick wall tests and alike when it comes to how they can be used in an artistic way. It is like giving a description about ingredients and recipe, calories, percentage of fat and carbohydrates on a food dish, which, however, won't provide you the experience that how it would taste, unless you actually eat it. BUT: if you have the experience of tasting and at the same time learn about how the technically measurable qualities of certain components influence that taste, you can become much more conscious about what is what and what you are doing. As for me, this is what these kind of optical tests are about, and I see important lessons there, notably about how much the performance of RF lenses depend on the actually used camera body (e.g. 50 Summilux on a Leica M vs Sony A7R). But, of course, I don't pretend to say that in itself such a technical knowledge is what makes somebody's photos good or not.

[Advertisement]

Hallo,

Look here: Leica M

[pgk]

I suppose my actual concern is that whilst information is in itself always useful (provided its accurate), it is all too often presented as comparative, competitive data, rather than simply for what it is - specific tests results which broaden knowledge.