Panasonic S 1.4/50mm or Apo-Summicron-SL 50mm

[Ivar B]

Advertisement (gone after registration)

In these quiet times I have read through some magazines which I did not have the time to read in detail previously. One such is the German magazine Color Foto, which as far as I know, is a well respected publication. One issue which has interested me a bit are tests of the Panasonic S-Pro 1.4/50mm and the Leica Apo-Summicron-SL 2.0/50mm ASPH; as I was in the process of buying a 50mm lens myself. It seems that Color Foto compute a weighted average of certain performance indicators, such as resolution, contrast, chromatic aberration, vignetting etc. and this is reported as point score. Both these lenses were tested on a Panasonic S1R. The Apo-Summicron-SL scored 100 points (a very good score) while the Panasonic S Pro scored 112. Apparently the Apo-Summicron loses Points as it needed to be stopped down to 5.6 before sharpness was at its optimum. Now, this difference may be due to the Leica lens being tested on a Panasonic body (?), the test methodology may be open to question; I don`t know. Perhaps it is fair to argue that these lenses are for all practical purposes identical when it comes to performance. I then wonder, how long can  Leica keep on charging 2 - 3 times the price for their lenses compared to the competitiors? An SL2 with one native lens is going to set you back many thousand dollars or euros.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[Advertisement]

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

[BernardC]

26 minutes ago, Ivar B said:

Apparently the Apo-Summicron loses Points as it needed to be stopped down to 5.6 before sharpness was at its optimum. Now, this difference may be due to the Leica lens being tested on a Panasonic body (?), the test methodology may be open to question; I don`t know.

It may be a question of firmware too. Panasonic may have provided a later update that improves performance with the Apo-Summicron-SLs.

Just about anybody can do automated "lens testing" using commercial software. The problem with this is that they are testing the camera/lens combination, not just the lens, and they are only testing one copy, usually at a single distance, pointed at a flat chart, with flat lighting. They also ignore some of the most important deciding factors (at least for me), like flare, handling, build quality.

In other words, these tests may prove to be very informative if you plan to use the same camera/lens to copy flat artwork in a studio, at a distance of 2m (for instance). They lose a lot of value when you change any or all of those variables, including firmware versions. They also don't take into consideration that "more accurate" is not always "better." We are not all doing scientific documentation, after all.

 

I recommend that you combine these types of lens tests with other sources, such as actual pictures taken in conditions that you may encounter (Sun in the frame, strong side light, etc), and actual handling (in-person preferably, but some video reviews can be helpful).

[Jared]

You need to be cautious with lens tests in general, not just this one.  First, you mentioned the “weighted average” score.  Does their weighting even vaguely match yours?  For example, let’s assume they give a relatively high importance to corner sharpness at moderate aperture.  In a landscape lens, I would agree that’s quite important.  In a portrait lens?  Wholly irrelevant.  What about flatness of field (if they even test it—most don’t)?  Again, in a landscape lens this might be important.  For a sports lens?  Or for a photo journalist?  Or street?  Probably not. 

What do you value in a lens? Resolution and sharpness?  In the corners? In the center? Wherever you choose to put your focus point? Do you always focus in the center then recompose, or do you compose then move your focus point? If the former, flatness of field would be quite important. If the latter, almost irrelevant.  Do you care about flare resistance? Most don’t check for that since there is no standard, repeatable test. But it probably ruins more photographs than lack of resolution.

You just need to be cautious about reading too much into lens tests. Like MTF charts themselves they can be extremely useful but do not tell you the whole story of a lens.  Want to know whether a lens will be good for astrophotography? An MTF chart will give you a heck of a start since it will easily show a curved field and give you a good idea about astigmatism when shot wide open; both of these are critical for showing point light sources at infinity, and shooting wide open is important in terms of signal to noise.  But a landscape photographer very rarely shoots wide open and may be more concerned with flare control than “sharpness” since sharpness will usually be limited by diffraction more than any other factor.

Even my statements are generalizations about what factors will be important to a given type of photographer but may not apply to any given photographer and certainly won’t apply to any given photograph.  Read the lens tests to understand the parts that are important to you. Ignore the overall score whether good or bad unless the testing protocol has the exact weighting you would want. And if they don’t publish their weighting formula? How can you possibly know whether their point system is a reasonable match for your type of photography?

Edited April 9, 2020 by Jared

[Ivar B]

May be their exact formula for computing the score is published some place, I don`t know. Of course the individual preferences could be quite different. There can be little doubt in my mind that other producers now match the performance of Leica lenses. I believe I have read that users see no difference between Summilux-SL 1.4/50 and Panasonic S 1.4/50, and often the Zeiss Otus 1.4/55 is thrown into the same group. I had perhaps expected that the Apo-Summicron-SL would distance anything else, but this appears not to be the case. Let us see how this change when Color Foto test on a Leica SL 2 body. Where the magazin tests the same lens on different bodies, results are often quite different. 

[Jared]

2 hours ago, Ivar B said:

May be their exact formula for computing the score is published some place, I don`t know. Of course the individual preferences could be quite different. There can be little doubt in my mind that other producers now match the performance of Leica lenses. I believe I have read that users see no difference between Summilux-SL 1.4/50 and Panasonic S 1.4/50, and often the Zeiss Otus 1.4/55 is thrown into the same group. I had perhaps expected that the Apo-Summicron-SL would distance anything else, but this appears not to be the case. Let us see how this change when Color Foto test on a Leica SL 2 body. Where the magazin tests the same lens on different bodies, results are often quite different. 

It has been a very long time since any given manufacturer was able to make lenses that others could not match.  Basically, ever since software could do the ray tracing for you.  They all have access to the same glass from Schott, Ohara, Hoya, and various other manufacturers. They all have the ability to manufacture accurate aspheres. They all have the ability to design a lens in a computer and evaluate very precisely how it will perform.  What varies is only the priorities.  Cost? Weight? Aperture? Size? Tolerances? Materials? These are the different levers a designer can pull.  Sigma and Panasonic are, and have been for quite a while, every bit as capable of designing and producing lenses that are just as good as those from Zeiss and Leica.  Same for Canon, Tamron, Nikon, Tokina, Sony, Cosina, Fuji, Nittoh (Hasselblad subcontractor), Schneider, Rokinon, and probably several other Chinese manufacturers.  They just have different priorities as they make products for different segments of the market.

Sigma ART lenses have been every bit as good optically as anything Zeiss or Canon or Nikon or Leica can make ever since their inception. But they can’t yet charge the price premium of the more famous brands just because of their history.  Given enough time that will change. At least Sigma hopes it will.

Leica has taken a calculated risk in forming the L mount alliance.  So far, I expect it is paying off for them.  It may well be the difference between their long term survival or not.  They never did have a monopoly on optical quality.  The thing that was different about them historically was not that they could make great lenses, but that they never made bad ones.  Now nobody makes bad lenses.  So they differentiate on other factors.
 

[Ivar B]

43 minutes ago, Jared said:

It has been a very long time since any given manufacturer was able to make lenses that others could not match.  Basically, ever since software could do the ray tracing for you.  They all have access to the same glass from Schott, Ohara, Hoya, and various other manufacturers. They all have the ability to manufacture accurate aspheres. They all have the ability to design a lens in a computer and evaluate very precisely how it will perform.  What varies is only the priorities.  Cost? Weight? Aperture? Size? Tolerances? Materials? These are the different levers a designer can pull.  Sigma and Panasonic are, and have been for quite a while, every bit as capable of designing and producing lenses that are just as good as those from Zeiss and Leica.  Same for Canon, Tamron, Nikon, Tokina, Sony, Cosina, Fuji, Nittoh (Hasselblad subcontractor), Schneider, Rokinon, and probably several other Chinese manufacturers.  They just have different priorities as they make products for different segments of the market.

Sigma ART lenses have been every bit as good optically as anything Zeiss or Canon or Nikon or Leica can make ever since their inception. But they can’t yet charge the price premium of the more famous brands just because of their history.  Given enough time that will change. At least Sigma hopes it will.

Leica has taken a calculated risk in forming the L mount alliance.  So far, I expect it is paying off for them.  It may well be the difference between their long term survival or not.  They never did have a monopoly on optical quality.  The thing that was different about them historically was not that they could make great lenses, but that they never made bad ones.  Now nobody makes bad lenses.  So they differentiate on other factors.
 

Very well said!

[Jon Warwick]

Advertisement (gone after registration)

3 hours ago, Ivar B said:

I had perhaps expected that the Apo-Summicron-SL would distance anything else, but this appears not to be the case. Let us see how this change when Color Foto test on a Leica SL 2 body. Where the magazin tests the same lens on different bodies, results are often quite different. 

I do wonder that too. 

I think the whole chain of camera-lens is key. Put it this way, i was SURE that I would get the S1R, and not bother with the SL2, given my initial assumption of same image quality but just in different clothing?!

I tried both cameras with the SL 90mm Summicron. IMHO, paired with the S1R, DNGs that were "too sharp" for my personal taste (admittedly I do come from an analogue background, and whilst i like as much resolution as possible, too much acuity stands out abruptly to my taste); with the SL2, the combo produced images that felt perfectly balanced and smooth in terms of what i was personally seeking in terms of acuity -- simply more "natural" to my eye.

I purchased an SL2, and am delighted with the image quality it produces, initially with my M 75 Summarit, and have every intention of getting an SL prime one day.

Despite my my initial thoughts that I'd definitely get an S1R to combine with an SL prime lens, i'm keeping it all "in house". 

[tom0511]

probably hari splitting. I would not think corner sharpness to be a problem of any of those lenses. Rather look at overall image impression, micro contrast, color, bokeh. And also handling. Do you want an f1.4 lens, than its Pana 50/1.4 vs Leica 50/1.4. Do you want a more compact lens with af , than its Leica Summicron. 

 

[Donzo98]

FYI... I read the reviews of the Panny 50, and tried two different copies. I also read that the sharpness and IQ in general, compared to the Leica 50 SL was similar. 
 

I directly compared the Leica and Panny at 1.4 on the S1R.  To my eye, the Leica was significantly better. I even tried a second Panny copy... but same result. 
 

Believe me... I would have been happy to have kept the Panny, especially for the price I paid, but at 1.4... Leica won the battle. 

 

Edited April 9, 2020 by Donzo98

[Fedro]

9 hours ago, Donzo98 said:

FYI... I read the reviews of the Panny 50, and tried two different copies. I also read that the sharpness and IQ in general, compared to the Leica 50 SL was similar. 
 

I directly compared the Leica and Panny at 1.4 on the S1R.  To my eye, the Leica was significantly better. I even tried a second Panny copy... but same result. 
 

Believe me... I would have been happy to have kept the Panny, especially for the price I paid, but at 1.4... Leica won the battle. 

 

what did you do in the end? Like you (if I remember correctly) I traded the Lux for the Cron; I love the cron but I am definitely missing something

[Ivar B]

11 hours ago, Donzo98 said:

FYI... I read the reviews of the Panny 50, and tried two different copies. I also read that the sharpness and IQ in general, compared to the Leica 50 SL was similar. 
 

I directly compared the Leica and Panny at 1.4 on the S1R.  To my eye, the Leica was significantly better. I even tried a second Panny copy... but same result. 
 

Believe me... I would have been happy to have kept the Panny, especially for the price I paid, but at 1.4... Leica won the battle. 

 

So your results are quite the opposite of what Color Foto reports. Then the natural question is who has completed the most stringent test procedures. You claim that there is a significant difference even, and as far as I have seen you are the only one who has come to this result. It would be very interesting if you could support this with data from your test as I find it hard to believe that this result is generalizable.

Here is a very thorough review published on YouTube, but this time the Summilux-SL 1.4/50

The tests appear to be very carefully executed and the results are very thoroughly documented. Differences are not large, but when it comes to sharpness. the winner is the Lumix. 

 

 

Edited April 10, 2020 by Ivar B

[Donzo98]

4 hours ago, Fedro said:

what did you do in the end? Like you (if I remember correctly) I traded the Lux for the Cron; I love the cron but I am definitely missing something

I sold the APO SL... and went back to the Lux

[Photoworks]

3 hours ago, Ivar B said:

So your results are quite the opposite of what Color Foto reports. Then the natural question is who has completed the most stringent test procedures. You claim that there is a significant difference even, and as far as I have seen you are the only one who has come to this result. It would be very interesting if you could support this with data from your test as I find it hard to believe that this result is generalizable.

Here is a very thorough review published on YouTube, but this time the Summilux-SL 1.4/50

The tests appear to be very carefully executed and the results are very thoroughly documented. Differences are not large, but when it comes to sharpness. the winner is the Lumix. 

 

 

they have a download of the images .

I think the Leica has less DOF at 1.4 and test are closer than different. In a blind test you could not see any difference .  On the extreme sides the Lumix shows some smearing bokeh, Leica is cleaner.
At 5.6 I see a better detail in the Leica but it is like splitting hairs.

 

I wonder if the sensor of the SL2 gives different resulted 

[Ivar B]

1 hour ago, Photoworks said:

they have a download of the images .

I think the Leica has less DOF at 1.4 and test are closer than different. In a blind test you could not see any difference .  On the extreme sides the Lumix shows some smearing bokeh, Leica is cleaner.
At 5.6 I see a better detail in the Leica but it is like splitting hairs.

 

I wonder if the sensor of the SL2 gives different resulted 

Regarding your final point, I would expect so. Kind of difficult to believe otherwise. Camera firmware may be an issue. I recall Sean Reid in his test of the 1.4/50 SL was kind of critical of the substantial firmware processing which goes into an image. This is probably just as true for Panasonic as it is for Leica. But then, why should not optical designers take firmware into account when they design a lens? 

[Donzo98]

5 hours ago, Ivar B said:

So your results are quite the opposite of what Color Foto reports. Then the natural question is who has completed the most stringent test procedures. You claim that there is a significant difference even, and as far as I have seen you are the only one who has come to this result. It would be very interesting if you could support this with data from your test as I find it hard to believe that this result is generalizable.

Here is a very thorough review published on YouTube, but this time the Summilux-SL 1.4/50

The tests appear to be very carefully executed and the results are very thoroughly documented. Differences are not large, but when it comes to sharpness. the winner is the Lumix. 

 

 

My test procedures aren't stringent at all... but the results were clear to me.

I took the same exact shots on both cameras... only at 1.4 (which is why I bought the lens). I did my typical post processing to both... without pixel peeping at first. The Panny was different from how I remembered the 50 SL. There was less pop in the subjects eye. I wasn't sure if what I was seeing was real. Zooming in to 100% it was immediately obvious that the 50 SL was sharper. I read and watched al the same things you did... and had to convince myself that the Panny I had was not somehow "off". I bought another Panny and tested again.. same result. I repeated the tests using text on a box as well.

Now... I bought the Panny used for around 1500... and bought the 50 SL used for 3900 including tax. Is the 50 SL that much better... maybe not. To me... I wanted the best glass in front of my sensor. I didn't save the 100% crops... so you can believe me or not. Trust me...I bought the 50 SL after I bought  the first Panny.. so I really wanted the Panny to be as good :).

The Panny definitely has many good points... it's a little lighter and much faster AF, but  IQ at 1.4 wasn't as good to my eyes.

I voted for the Leica at a significant increased cost financially...so that should tell you how sure I was.

Edited April 10, 2020 by Donzo98

[Jared]

5 hours ago, Ivar B said:

The tests appear to be very carefully executed and the results are very thoroughly documented. Differences are not large, but when it comes to sharpness. the winner is the Lumix. 

 

 

Doing accurate lens tests like this is more challenging than you might think.  For example, in the test above it appears the subject distance is perhaps 1.5 to 2 meters.  We’ll be generous and assume 2 m.  The horizontal field of view of a 50mm lens on a 36mm sensor is just under 40 degrees.  The subject is perfectly flat—the slats of an outside fence of some sort.  At that distance. The edges of the frame would be almost 13cm farther away from the camera than the center.  Depth of field behind your subject at f/1.4 and a distance of 2m is about 7cm.  In other words, edges of the frame would be significantly defocused just by geometry.  Damn Pythagoras!

Plus, the focal lengths of the four lenses in the test varied from 40mm to 55mm (more than 25%) yet the same subject distance was used.

And what about field curvature? Did that make things better or worse for each lens?  Could be either depending on direction and complexity of the curve.

Finally, what about sample variation? Is it large or small for any given model? Is the copy used typical? Better? Worse? With a sample size of one can you draw conclusions about more than just that one lens? Roger C’s lens tests have shown sample variation, particularly with zooms but also with primes, is huge even with premium lenses.

Look, I’m not trying to say the Leica is better or the conclusions regarding the Panasonic are wrong.  But is a test like this one strong evidence in either direction? It’s one data point based on flawed methodology.  Testing lenses when you are trying to identify differences of just a few percent in MTF is really, really hard.  Be cautious about drawing conclusions.

Edited April 10, 2020 by Jared

[Ivar B]

20 minutes ago, Jared said:

Doing accurate lens tests like this is more challenging than you might think.  For example, in the test above it appears the subject distance is perhaps 1.5 to 2 meters.  We’ll be generous and assume 2 m.  The horizontal field of view of a 50mm lens on a 36mm sensor is just under 40 degrees.  The subject is perfectly flat—the slats of an outside fence of some sort.  At that distance. The edges of the frame would be almost 13cm farther away from the camera than the center.  Depth of field behind your subject at f/1.4 and a distance of 2m is about 7cm.  In other words, edges of the frame would be significantly defocused just by geometry.  Damn Pythagoras!

Plus, the focal lengths of the four lenses in the test varied from 40mm to 55mm (more than 25%) yet the same subject distance was used.

And what about field curvature? Did that make things better or worse for each lens?  Could be either depending on direction and complexity of the curve.

Finally, what about sample variation? Is it large or small for any given model? Is the copy used typical? Better? Worse? With a sample size of one can you draw conclusions about more than just that one lens? Roger C’s lens tests have shown sample variation, particularly with zooms but also with primes, is huge even with premium lenses.

Look, I’m not trying to say the Leica is better or the conclusions regarding the Panasonic are wrong.  But is a test like this one strong evidence in either direction? It’s one data point based on flawed methodology.  Testing lenses when you are trying to identify differences of just a few percent in MTF is really, really hard.  Be cautious about drawing conclusions.

Yes, testing is a complex issue, I fully agree. I am not a physicist but an economics professor, so I know a little about conducting tests and experiments. Some of the arguments we see here in the thread cannot be called "testing" and probably suffer heavily from confirmation bias. Perhaps Color Foto testing are conducted under a more strict regime. They are professionals and do this every day. I agree that 40 to 55mm lenses were tested, but most of the comparisons made, were 50/50 etc. I don`t know how large sample variation is in practice. I have a friend who is an M-user over several years, and more or less every (!) Leica lens he bought in the later years needed to og back to Leica and sometimes more than once. Well, he is a specialist and critical user. I know also that sample variation is a problem for Voigtlander and it is probably so for Panasonic as well. It appears that many don`t approve of the work done by DXO, but sometimes they have tested a large number of the same lens and it has been very surprising to see how large the sample variations are.  

Edited April 10, 2020 by Ivar B

[Donzo98]

28 minutes ago, Ivar B said:

Yes, testing is a complex issue, I fully agree. I am not a physicist but an economics professor, so I know a little about conducting tests and experiments. Some of the arguments we see here in the thread cannot be called "testing" and probably suffer heavily from confirmation bias. Perhaps Color Foto testing are conducted under a more strict regime. They are professionals and do this every day. I agree that 40 to 55mm lenses were tested, but most of the comparisons made, were 50/50 etc. I don`t know how large sample variation is in practice. I have a friend who is an M-user over several years, and more or less every (!) Leica lens he bought in the later years needed to og back to Leica and sometimes more than once. Well, he is a specialist and critical user. I know also that sample variation is a problem for Voigtlander and it is probably so for Panasonic as well. It appears that many don`t approve of the work done by DXO, but sometimes they have tested a large number of the same lens and it has been very surprising to see how large the sample variations are.  

Trust me... I am a physician, not a Leica brand ambassador No bias here... at 1.4... the Leica is better. I tested two Panny 50's on the S1R. I'm convinced  

[Advertisement]

Hallo,

Look here: Leica L-Mount

[Fedro]

6 hours ago, Donzo98 said:

I sold the APO SL... and went back to the Lux

I may end up doing the same. I dont use the SL for walkabout much, and when I do, it is with the 24-90

The Lux was one of my favourite portrait lenses, much as the APO is stunning, it does not give me the same emotions

[Donzo98]

55 minutes ago, Fedro said:

I may end up doing the same. I dont use the SL for walkabout much, and when I do, it is with the 24-90

The Lux was one of my favourite portrait lenses, much as the APO is stunning, it does not give me the same emotions

Yup... no doubt. I also have the 35 SL which is even sharper...but more clinical. The 50 APO , to me is similar to the 35. The Lux is different...

Edited April 10, 2020 by Donzo98