“Image quality” - SL2+M lens vs S1R+SL lens?

[Jon Warwick]

Advertisement (gone after registration)

I’d really appreciate your thoughts.

I’m in a position to spend c £ 5.5-6k on a digital set up.

I already own an M 75mm 2.4 Summarit + M-L adapter.

My choice that I’ve come up with is 

(1) buy a new SL2 body + use my existing M 75mm lens

(2) buy a used S1R body + buy a new 75mm SL Summicron lens

Clearly the biggest different is manual vs autofocus. But let’s leave that aside.

** What I’m interested in, is, which combo do you would give me the best bang for my buck, PURELY in terms of image quality? I print big to c 50” wide, and would use the pixel shift technology on either camera for landscapes.

The MTF on the M 75mm Summarit is pretty impressive (at f5.6) and the microlenses and thin glass stack on the SL2 should mean the M lens works well .... but do the SL Summicrons simply provide a better level of clarity and tonal depth that one sees the difference, even if it means I have to go “non-Leica” on the body?
 

[Advertisement]

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

[Photoworks]

Pixel Shift is not a good idea for landscapes, since the subject moves. SL2 will have this function in the next firmware update.

I have used the SL2 + 90 apo M all last week and was surprised about the slight miss focus even at 8 feet, but little movement of me and the person sitting across gave me a hit rate of 75%. I was shooting 2.8-4f

I think the best performing lens will be the one designed for the camera, so SL2 and the 75mm SL

I am not big fun of the look of the Panasonic , but I would invest in lenses.

[Rob L]

The used S1R  with an APO Summicron SL 75 will outperform the SL2 with Summarit M 75. I can't say by how much at f/5.6 though. The cameras are basically the same. SL Summicrons are leaps ahead of M lenses.

[caissa]

The best bang for the buck is very personal. Most people would say S1R plus a new L-mount lens (maybe a Sigma instead of a Summicron. Maybe Sigma 1.4/28 and 1.4/85 ).

But for me personally the camera is important and somehow I like the SL cameras. So for me I would rather buy the SL2, temporarily use the M or R lenses that are already around and add later more modern Leica SL lenses (I would probably start with SL 16-35 and SL 75mm).

But there is no “objectively” best solution. No possibility to “optimize” it. But compare to the money you spend on transportation (car), if you optimize there, then you don’t need to optimize your hobby. (Sorry, if you are a professional user. Then I would have assumed you know better than we what is best for your business.)

Edited February 28, 2020 by caissa

[Jon Warwick]

1 hour ago, Photoworks said:

I have used the SL2 + 90 apo M all last week and was surprised about the slight miss focus even at 8 feet, but little movement of me and the person sitting across gave me a hit rate of 75%. I was shooting 2.8-4f

Good to know. I’d often attempt to use either M or SL 75mm wide open, so sounds like autofocus is a real benefit for that situation 

[Jon Warwick]

1 hour ago, caissa said:

But there is no “objectively” best solution. No possibility to “optimize” it. But compare to the money you spend on transportation (car), if you optimize there, then you don’t need to optimize your hobby. (Sorry, if you are a professional user. Then I would have assumed you know better than we what is best for your business.)

Yes, it’s as a hobby!. I owned a M240 in the past, but the SL2 / S1R + SL prime seem a large upgrade in terms of image quality (to my eyes at least), so I’m coming back into colour digital after a several year absence - currently using E6 in my M7 film camera and 5x4. 

To get to an SL2 + SL prime combo, one option would be for me to sell my M7 film camera ....especially as I’m not sure how much I’d continue using 35mm film, given (1) the relative image quality of the SL2+SL primes, and (2) if taking the time & effort (and drum scanning costs) with film, I’d prefer to use the 5x4.

Edited February 28, 2020 by Jon Warwick

[douglas ball]

Advertisement (gone after registration)

If you have the patience to learn about the added complexity of the S1-R, then that should be your choice, along with the SL75.

I am an industrial designer, so I do prefer the looks of the SL to the Panasonic, but I prefer what the S1-R offers, especially the tilting screen.  I still have an SL but always go with the Lumix.

In the high resolution mode on the S!-R there is a 'Mode 2' which takes into account  a little movement in the scene. 'Little' meaning a flag blowing in the wind for example. It then uses just one frame of the 7 for the flag, when it combines them all into one image. If the wind is blowing enough that grasses and tree branches are moving, then forget the high resolution mode.

 

[Macberg]

@Jon Warwick You might also want to have a look at this - the same lens (in the first link, it's the Leica 35mm Summicron SL, in the second link, it's the 28mm Summicron M Asph.) performs visibly better on the SL2 (than on the S1R).

As for the use of M-Lenses on the SL - Leica claims that their micro-lenses on the sensor were especially designed for the use of M-lenses on the SL2 (see the technical chart when you scroll down about one third in this review): https://neunzehn72.de/leica-sl2-review-mein-praxistest/

 

P.S.: I just saw, that you started the first thread...well, this might help others, too...😉

 

Edited February 28, 2020 by Macberg

[Jon Warwick]

I guess what’d be interesting - has anyone done a direct comparison of 

- SL2 body + M lens (in the 50-75mm range) vs.

- SL2 + SL Summicron prime

....I’ve tried out the latter, in isolation, and thought the image quality was incredibly good compared to my M240 set-up.

It’s just not clear to me how much of that relative image quality (over my M240) comes from the SL2 body, and how much from the SL prime lens,

ie, am I compromising much image quality (resolution, micro contrast etc) if I started with an SL2 + M lens (rather than an SL prime) even at f5.6? From Rob L’s comment, it sounds like I might be?

Edited February 28, 2020 by Jon Warwick

[caissa]

Maybe you lose some resolution with wide aperture. But for these topics resolution (in the corners) is usually not crucial anyway. But if  you use small apertures then the difference will be minor. (probably even non-existent).

The way I photograph is usually far from perfect (regarding technical prowess). I'm doing it kind of "laid back". So I mix and match lenses, e.g. I still use an old M 28mm from the 90s, although the modern ones (and the SL16-35 or Sigma 24-35 are closer to perfection). But if it is your style to push to the limits in each second, then you will be happier with a pure SL prime setup. (only primes, although the 90-280 and 16-35 are better than any other lens I ever had before in that range) 

Edited February 28, 2020 by caissa

[bags27]

There is a bit of thread drift here, and I apologize to the OP for participating in it since I'm not addressing 75mm focal length.

I won't volunteer his identity, but another member of the LUF has convinced me, through careful and detailed testing and comparison with the L35 f/2, that there is a fall-off of acuity around the edges of even the sharpest M lenses at least up to and perhaps including 50mm lenses. I'm not sure in the end this will stop me from eventually buying an SL2 to pair with my M and R glass, but, despite the assurances of several very trustworthy reviewers, a difference in edge sharpness between M and L lenses is there.

Edited February 28, 2020 by bags27

[caissa]

Yes, there is a difference, but it depends mainly on the topic and the aperture if it is important (visible) or not.

Edited February 28, 2020 by caissa

[bags27]

1 minute ago, caissa said:

Yes, there is a difference, but it depends mainly on the topic and the aperture if it is important or not.

To an extent I certainly agree. But I was surprised to see the difference even when the lenses were not at their widest apertures.

Edited February 28, 2020 by bags27

[caissa]

Was it a "good" photo, or just a boring unpleasant "test" shot ? Who cares if the test shots are inferior ....     

Many of my "favorite" photos that I look at again after years are not perfectly sharp (and it is not really clear if it is the fault of the lens ). At least not everywhere and certainly not at the corners/edges. Often the contrast  is much more important for the result. (vs edge sharpness)

I don't know if you know Ernst Haas (member of Magnum) and his photo style. When I was young, he was the big master photographer. And I still like this style. He used Leica, but in his pics sharpness is not the dominant element. So from his photos I know that an artist can cope very well with M lenses that are not perfectly sharp (I mean not to the outer rim).   La Suerte de Capa (Spain 1956)  is such an amazing photo I still like after all these years. (mainly shape and color) So to say I like it against my will.

Edited February 28, 2020 by caissa

[bags27]

8 minutes ago, caissa said:

Was it a "good" photo, or just a boring unpleasant "test" shot ? Who cares if the test shots are inferior ....     

These were intentional photos. I don't know what the line is between the two. As I wrote, it hasn't prevented me from thinking about the kit myself. But I think anyone who is going to pay 6k for a camera body and expect their very expensive M glass to perform as well there as on an M might want to consider this. Everyone (at least Jono and Sean Reid) says that M glass still works "best" on M bodies. How much is "best" and "good enough"? Falsely attributed to Plato and then to Voltaire, who does actually quote it: the perfect is the enemy of the good. 😀

Edited February 28, 2020 by bags27

[thighslapper]

I'd go for the S1R + 75/2 apo. 

The results are impeccable.

The only differences between the S1R and SL2 are the interface and ergonomics ...... and that's just a matter of taste and preference. 

[Jon Warwick]

1 hour ago, bags27 said:

To an extent I certainly agree. But I was surprised to see the difference even when the lenses were not at their widest apertures.

That’s very interesting, thanks. Certainly I’d expect the wider M lenses (I assume you’re mainly talking about 28/35/even 50 focal lengths) to be compromised due to the thicker sensor glass of even an SL2 compared to the M.

Part of the reason for my question about how the longer M lenses like my M 75mm (Summarit) would perform on the SL2 compared to the 50-75mm SL Summicron primes is the following ....whilst it’s not quite as perfect as the SL primes, the MTF for the M Summarit (specifically at f5.6 .... see the link) is surprisingly and remarkably good at almost 80% at 40 lines/mm, and if it doesn’t pose problems with the SL2 (perhaps due to the 75mm focal length being more telecentric compared to the wider M lenses) maybe the difference with the SL lenses isn’t so far off in this specific 75mm instance? .....

https://uk.leica-camera.com/content/download/131042/1644356/version/1/file/Datenblatt_Summarit-M+2.4_75mm_e.pdf

 

Edited February 28, 2020 by Jon Warwick

[Donzo98]

S1R + SL glass.... no question in my mind!!

Stellar, stellar, stellar.... 

Only reason to buy an SL2 is for wider M glass. 
 

[Advertisement]

Hallo,

Look here: Leica L-Mount

[bags27]

What my friend who did the extensive testing suggests is that most of the shorter focal length M lenses (we didn't check them all on the wiki, but it is a reasonable generalization) have a concave final element. This is to be distinguished from L glass, which has a flat final element. There's only so much that software and microlenses (what are those, really?) can do to change the direction of light that is designed to go, in a very, very small space, in a different angle.

So, it may not just be the thickness of the glass covering the sensor, but the design of the M glass itself, that is a compromising variable.

[Tailwagger]

I sadly feel the need to step out of the shadows.

AFAIK, the issue has nothing to do with aperture. Changing aperture might sharpen the result because the lens itself performs better, but otherwise it is not in the equation.  The core problem is a simple one. M lenses were never designed to deal with a piece of glass between them and their point of convergence. Digital change all that. Light striking glass at an angle mean refraction. Thats physics. Any lens, M or otherwise, that projects light onto the sensor at an angle other than perpendicular presents a problem for a digital sensor. The more extreme the angle, the greater the problem. The thicker the cover glass the greater the amount of refraction. Now that I've come to get a laymen's understanding of the issue, I think it's frankly a marvel, all applause to Leica, that the wide M lens work as well as they do on the M.

I've shot two M lenses side by side with their native equivalents on an SL2. There is no comparison. A 1000 pixels in from the corners, the Pano zoom at 35mm crushes my Zeiss Distagon-M 35mm, the sharpest M lens I own.  And by crush I mean destroys, no comparison, throw the Zeiss in the garbage can. The same is true when comparing the nearly equally sharp SEM 21 to the Pano output at 21mm. Is the same true for all other M lenses? No. My 75mm 'lux, for example seems to be fine and can carry its head high, at reduced apertures, when compared to the incomparable SL-75. Given the 75mm 'lux's rear optic is tucked well up in the barrel, presumably light exiting it moves perpendicular or at least far closer to it, than the others mentioned.  Hence no ill effects.  The 50mm Lux in my opinion is compromised though others report it's fine. The ZD and SEM have concave rear elements that protrude into the camera body. Interestingly, the final element of the 'lux is concave as well, though not as extreme.  Draw your own conclusion on the angle of incidence each of these lenses produce.  All of these optics are fine center frame, the physics tells us why that is so and why they suck in the corners. Leica has done the best they can, but they can not repeal the laws of physics. Why they could not employ the same technology on the SL that they did on the M is completely beyond my understanding, but presumably it introduces some other form of compromise for the native L glass they refused to entertain. Hence the SL2 seems to be a compromise, whereas the M is a specific dedicated, best possible solution. Again, draw your own conclusions.

I'm not an optical engineer, just a logician and a real world user of these lenses. What I can state from my own experience, and Ken can confirm, is that if you shoot with an SL2 using either of the two M lenses I've mentioned ($5600 worth), you risk giving up a significant, visible, clear for anyone to see, amount of performance as compared to a $1500 zoom, let alone the SL35 which I also own, no matter the aperture. If you are unconcerned with the corners, no problem. If you shoot on a CL, likely not a problem. If you shoot landscape... well. My 'experiments' were carried out in actual shooting scenarios, real world shots doing landscape and concerts. I found the results of those two lenses to be completely unacceptable and certainly nothing I would consider to be remotely worth an investment of $9K for camera and lens. I have never seen any such ill effects when using either of those lenses on the M10 and have used both those lenses in the same contexts for many years and many thousands of frames.  

Make of this what you will.  I frankly have no desire to get into any arguments. I discussed this topic with Ken in background at length over several weeks. I carried additional gear, shot a number of additional frames on his behalf to test things as I was well positioned to help him decide if an SL2 + M lenses rather than an M was a sound decision. I'm personally fed up with this topic, which is why Ken did his best to keep me out of it. But in the interest of community, I thought it a responsibility to simply state what I've seen. Bottom line, like Reid and Jono, I shoot M on M, L on SL. Would I shoot my 75 lux on the SL2? If I didn't own a pair of Ms and an SL-75, sure. Focal lengths below that? No. I didn't pay all this money to entertain any additional compromise. YMMV.

 

Edited February 28, 2020 by Tailwagger