M lens performance on the SL

[k-hawinkler]

Advertisement (gone after registration)

Hmm - well I agree with Mladen - wide open several lenses will show slightly unsharp corners on the M as well (mostly due to curvature of field).

Also worth mentioning that if you want a torture test you're better to do a distant subject (where the rear element of the lens is closer to the sensor). 

I've got some proper tests I'm hoping to put together into an article (but I'm not sure which is more boring, taking the pictures or writing about them!

 

"(but I'm not sure which is more boring, taking the pictures or writing about them!"

 

Jono, I can assure you, both! 

 

Hiep and I did infinity shots for about 30 Leica lenses on 2 cameras for all apertures.

Some surprising results though, presented here.  http://www.fredmiranda.com/forum/topic/1392833/2#13289195

 

[thighslapper]

At great personal cost to my sanity I have tested corner performance on all the sub 50mm lenses I possess.

The whole exercise is fraught with technical difficulties and it is hard to completely exclude subjective opinion.

As I don't possess a 50m tower so that I can ensure the whole frame is at infinity ..... and no handy enormous cliff or building to ensure all four corners have something in that you can compare, I've had to resort to the local church.

Close tests were done indoors .... and even with a tripod it is difficult to be sure everything is perpendicular in all planes.

I've taken the performance in the best corner as the comparator ...... some lenses show variation and I am unsure if this is due to the lens .... or me. 

I'd nearly lost the will to live by the time I had done these and stared endlessly at 100% crops ...... I may do the 50's and above at some stage. 

Basically only the 28/2 and 35/2 are a disappointment ...... all the others seem fine in comparison to the M

Close corner performance is poor on almost all lenses ...... probably more due to field curvature than sensor issues ...... but who uses a 21mm lens routinely at 1m...... and wide open at that .... ????

In general use for landscape the top corners will be lost in sky and bottom probably OOF as the will be near foreground ....... and if you are using a 35/1.4 wide open it's not because you want the corners sharp ..... usually the converse. 

Hope this helps some of you with doubts/concerns ...... or confirms your fears ..... 

 

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!

 

grade 1:

 

 

grade 2:

 

 

grade 3:

 

 

grade 4:

 

 

grade 5:

 

Edited October 8, 2016 by thighslapper

[Advertisement]

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

[thighslapper]

These were the original subjects ........

 

Done at iso 200 on both cameras ..... distance altered so the same image was in each frame. 

As I am not 3m tall I cannot avoid some discrepancy in tilt in the focal plane (nearer below, further above), which is why I compared all 4 corners and took the best for comparison. 

Doing this absolutely accurately requires an anally retentive personality and more gear, time and patience than I possess. 

No adjustments in LR except to equalise exposures. No sharpening ..... basic OOC DNG's

Voigt coded as 28/2.8 (gives best results) and Zeiss 35/1.4 as 35/1.4 asph.

No observable colour casts in the periphery of any SL (or M) images 

I've not bothered with centre image definition ..... there is little to suggest it would be worse than on the M ..... and in many cases probably better. 

 

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!

        

 

I'd be interested to see what other users find ..... I may no great claims to absolute accuracy .... this is mainly a comparison and how it impacts on normal lens/camera use. 

 

The example grading photos above are from the bottom left of the church below the small window ....

Edited June 28, 2017 by thighslapper

[edwardkaraa]

Fantastic work. Thank you!

[ramarren]

Thanks for the ratings. Lens testing is a serious pain in the butt, I agree. 

[k-hawinkler]

Hi thighslapper, many thanks indeed.
Could you please give more details about your infinity comparisons. TIA.

 

 

Edited November 16, 2015 by k-hawinkler

[k-hawinkler]

Advertisement (gone after registration)

 

Edited November 16, 2015 by k-hawinkler

[hiepphotog]

A lot of what you say is very true and I totally agree. My only point is that the sensor and lens need to be optimised.

 

In theory if the cover glass is the only issue and refraction is reduced, it should make the kolari mod Sony A7 equal to the M240 atleast in the corners. But its not. The kolari mod A7 is not even equal to the M240 in the center (this is with lux 50). Thats why I believe there are other factors at play rather than a simple cover glass issue.

 

Have done many careful tests myself at infinity and seeing results from Zeiss MTF, Jim Kasson and Lloyd Chambers, I am not sure I would believe that the A7 can't match the M240 in the center with the Lux 50. Even before the mod, they should be very close; the difference is in the AA filter.

 

Brian Caldwell (brilliant optician himself) and Roger Cicala already showed the strong effect of stack thickness in the past. Going past that, microlens array's difference can only account for the micrometer scale difference, which might be picked up on an optical bench but not likely on simple visual check. Lastly, after the mod, the total stack thickness on the Kolari mod is still slightly thicker than most M digital (0.8mm + 0.5mm = 1.3mm). From hearsay, the M240 has a 1mm stack. I imagine this SL has aroundn 1.2-1.3mm to have better IR rejection, more durable, and somewhat more compatible with the old T lenses.

 

Edit: FWIW, both the A7S/II and A7RII are better than any Leica M digital at handling vignetting and color cast. Of course, on the other hand, smearing (can't be fixed in post) can be a problem on Sony body, pre or post mod.

Edited November 16, 2015 by hiepphotog

[scott kirkpatrick]

@thighslapper -- Is it fair to assume that because the distant shots (church wall) were taken at a slight slant and you chose the best corner, that this tended to overcome some of the effects of field curvature.  Was it generally a bottom corner that came out best in the church wall shots?  That could be why your short distance corner tests were generally worse than the distant example, although we expected the opposite, if the angle of the light rays at the edges of the image sensor were the biggest problem.  It may be that this is no longer such a problem.

 

I hesitate to suggest that you focus on the corners, although I might try that next month. (In the SL, not in the M.) I find boring lab work a bit soothing sometimes.

 

scott 

[thighslapper]

Hi thighslapper, many thanks indeed.

Could you please give more details about your infinity comparisons. TIA.

 

I have nowhere suitable to accurately test true infinity ...... at the distances I did these most of the w/a lenses were close to the infinity stop anyway ....... the WATE 16 at 2m is in focus from there to infinity.

Not sure there would be much difference really.

In your neck of the woods you seem to be blessed with endless horizons/hills/canyons and plenty of suitable test spots ...... where I live is much more problematic ....

[k-hawinkler]

Scott, it makes a lot of sense to do two shots, one focused in the center, the other in a corner.

We - Hiep and I - did both.

Edited November 16, 2015 by k-hawinkler

[jonoslack]

 

Edit: FWIW, both the A7S/II and A7RII are better than any Leica M digital at handling vignetting and color cast. 

I'm interested in this statement - I've not tested either of those cameras, but I have tested the A7ii, and at least at vignetting it's much much worse than either the M240 or the SL (I wasn't looking at colour casts). Leica impose vignetting correction for each lens individually for different apertures, which, of course, Sony do not. It's hard to see how Sony have managed to make their cameras do better for vignetting with M lenses without any firmware corrections.

[thighslapper]

@thighslapper -- Is it fair to assume that because the distant shots (church wall) were taken at a slight slant and you chose the best corner, that this tended to overcome some of the effects of field curvature.  Was it generally a bottom corner that came out best in the church wall shots?  That could be why your short distance corner tests were generally worse than the distant example, although we expected the opposite, if the angle of the light rays at the edges of the image sensor were the biggest problem.  It may be that this is no longer such a problem.

 

I hesitate to suggest that you focus on the corners, although I might try that next month. (In the SL, not in the M.) I find boring lab work a bit soothing sometimes.

 

scott 

 

The 'grading' shots are just for examples of the degree of crappiness. I compared the best corners of each ..... and anyway this is a comparison .... with the same factors in both sets of shots, not an absolute test of each lens.

If you read through my commentary, I mention all of this ...... and an acknowledgement that this is not perfect .... and if anyone wishes to waste a few days of their life repeating all this more accurately then they are welcome to. 

The take-home message is that some lenses WO are a bit worse on the SL (and the usual suspects), but mostly not by much, and by f5.6 they are all just as good, if not better on the SL.

I included the near tests only because I did these first because the weather was awful ..... and whilst they exaggerate the differences seen at mid distance they hardly represent the normal distance these lenses are used at ....

[k-hawinkler]

I have nowhere suitable to accurately test true infinity ...... at the distances I did these most of the w/a lenses were close to the infinity stop anyway ....... the WATE 16 at 2m is in focus from there to infinity.

Not sure there would be much difference really.

In your neck of the woods you seem to be blessed with endless horizons/hills/canyons and plenty of suitable test spots ...... where I live is much more problematic ....

Thanks again. If I were in your situation I would ask a person who could give me access to a top window in a tall church tower. Then you can easily do infinity shots. :-)

Edited November 16, 2015 by k-hawinkler

[hiepphotog]

I'm interested in this statement - I've not tested either of those cameras, but I have tested the A7ii, and at least at vignetting it's much much worse than either the M240 or the SL (I wasn't looking at colour casts). Leica impose vignetting correction for each lens individually for different apertures, which, of course, Sony do not. It's hard to see how Sony have managed to make their cameras do better for vignetting with M lenses without any firmware corrections.

 

Jono, quite simple: A7S-big fat pixel well, less cross-talk (corner color cast), more light for oblique ray region (corner vignetting)

 

A7RII - BSI, shallow well: less cross-talk, less light loss (no need to go through the underlined circuitry like in the old front side illuminated)

 

Both of those cameras can handle my ZM 15 with ease without the need for color cast correction. Vignetting can be handled using the normal LR's vignetting slider. Same can't be said about the M9 or M240. I have seen the same thing on Biogon G 21, ZM 25, C-Biogon 21/4.5.

[AtelierM]

When using a M lens on the SL will it function in Aperture priority?

[jonoslack]

I have nowhere suitable to accurately test true infinity ...... at the distances I did these most of the w/a lenses were close to the infinity stop anyway ....... the WATE 16 at 2m is in focus from there to infinity.

Not sure there would be much difference really.

In your neck of the woods you seem to be blessed with endless horizons/hills/canyons and plenty of suitable test spots ...... where I live is much more problematic ....

It's a real pain doing that stuff, and you seem to have done a great job (and I like the rating system)

But there's a difference between 1km and infinity! (as I know to my cost). Let alone a church wall and infiity.

I did a whole range of tests in Venice - taking pictures from the S-Elena vaporetto stop to San Giorgio Maggione . . . and it wasn't really far enough. 

Edited November 16, 2015 by jonoslack

[jonoslack]

Jono, quite simple: A7S-big fat pixel well, less cross-talk (corner color cast), more light for oblique ray region (corner vignetting)

 

A7RII - BSI, shallow well: less cross-talk, less light loss (no need to go through the underlined circuitry like in the old front side illuminated)

 

Both of those cameras can handle my ZM 15 with ease without the need for color cast correction. Vignetting can be handled using the normal LR's vignetting slider. Same can't be said about the M9 or M240. I have seen the same thing on Biogon G 21, ZM 25, C-Biogon 21/4.5.

Hang on - Leica don't provide vignetting correction for Zeiss lenses! Of course you can code them as something or other, but that's only going to be an approximation. I don't doubt that Zeiss lenses on Leica is interesting, but it's not what this topic is about (is it?)

[hiepphotog]

Hang on - Leica don't provide vignetting correction for Zeiss lenses! Of course you can code them as something or other, but that's only going to be an approximation. I don't doubt that Zeiss lenses on Leica is interesting, but it's not what this topic is about (is it?)

 

. I don't want this to be another Leica vs. others talk. Even with native Leica lenses, Leica M body needs the 6-bit code to do "in-camera correction" for color cast and vignetting. Those two Sony bodies don't need that because they can physically cope with these errors.

 

And yes, Zeiss do make M-mount lenses as well. Again, let me stress that both vignetting and color cast can be fixed in PP even if you decide to turn off 6-bit coding. With smearing, you can't. So M cam is still the best platform if you want to extract the best details from M optics. It's just that the digital sensor using on those M cam is not the best at handling vignetting and color cast.

[Advertisement]

Hallo,

Look here: Leica L-Mount

[k-hawinkler]

It's a real pain doing that stuff, and you seem to have done a great job (and I like the rating system)

But there's a difference between 1km and infinity! (as I know to my cost). Let alone a church wall and infiity.

I did a whole range of tests in Venice - taking pictures from the S-Elena vaporetto stop to San Giorgio Maggione . . . and it wasn't really far enough.

Hi Jono, you should visit here sometime.

On a clear day you can see for one hundred km. :-)

That should be enough distance for infinity shots.