M lens performance on the SL

[IkarusJohn]

Advertisement (gone after registration)

We can (apparently) expect the R lenses to perform just fine with the SL.

 

The truism that M lenses perform best on the M cameras (by reason of thinner cover glass on the sensor if nothing else), followed closely by the SL and a long way by the Sony A7 series cameras (we probably don't need to be concerned over them, but as many here use them as alternatives to the M it does no harm to put this on the record) has been floating around for a while with limited real testing.  From what I've seen, you need to look for the difference between the M and SL cameras using M lenses, but it is there.

 

So, how many are good, and how many not so good?

 

Looking purely at corner smearing (which is fatal) we know, or can assume:

• lenses 50mm and longer should be no different whether used on the M or SL

• older 35mm designs like the 35 Summicron ASPH have issues in the corners, shot wide open

• the 28 Summilux actually seems to have better resolution in the centre and in the corners on the Sl than on the M(240) - it's marginal, but it is there

• the older 28 Summicron ASPH does not perform quite as well in the corners on the SL as it does on the M(240) - again, you need to look for it, but it's there

Apparently, we can expect the WATE to perform well, the 21 Summilux as well, but perhaps not the 21 SEM (based on its performance on the A7 cameras when tested on the release of the A7r).

 

Please add here user experiences or other conclusions from looking at files online.

[thighslapper]

I don't think you can look 'purely at corner smearing' .......

A significant part of the problem in the final image is correction of vignetting with shadow recovery and smoothing ..... so it is hard to unpick true optical/sensor issues and those created by firmware correction.

..... and are we taking things to the usual somewhat unreasonable 'pixel-peeping' standards .... or what is practically an issue in real world photography. 

With the odd exception I am sure anything under 50mm will be worse to some extent than on the M ...... the question is whether it is enough to be an issue or not ...... and that depends how high you set the bar .....

[Advertisement]

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

[IkarusJohn]

Nah, it's not that linear.  Some are good, some are not so good.

 

The problem with smearing is it really can't be rectified (other than by cropping), and yes most of the time it is little more than pixel peeping - people spending more time reading reports and doing test charts than just getting out and taking pictures.

 

My pick?  I'm going to be very happy with the SL and I will stuff up far more in processing than I will ever notice in off axis problems.  We're really dancing on the head of a pin.

[wstam]

Is it possible to re-calibrate the M lenses so that there are more optimised to SL?

 

I don't have enough technical knowledge on optic design to know if this question make sense. Few years back I bought an Elmarit-M 90/2.8 from someone but found the focus was off on my M cameras. I sent it to Leica for calibration. It is spot on after that. I remember the staff was telling me something like Leica re-calibrated it for the digital sensors instead of film. Plus now I read somewhere that the new Summilux-M 28mm performs even better on SL than M camera (not sure how true). So I am thinking that may be M lenses can be optimised for SL by some calibration. Whether one wants to do that is another question.

 

Thanks.

[mmradman]

Is it possible to re-calibrate the M lenses so that there are more optimised to SL?

 

I don't have enough technical knowledge on optic design to know if this question make sense. Few years back I bought an Elmarit-M 90/2.8 from someone but found the focus was off on my M cameras. I sent it to Leica for calibration. It is spot on after that. I remember the staff was telling me something like Leica re-calibrated it for the digital sensors instead of film. Plus now I read somewhere that the new Summilux-M 28mm performs even better on SL than M camera (not sure how true). So I am thinking that may be M lenses can be optimised for SL by some calibration. Whether one wants to do that is another question.

 

Thanks.

Short answer - no.

 

In case of your Elmarit 90mm it was focus adjusted for tighter tolerances required by digital sensor (tighter than film) but hardware optimization for non native sensor would most probably require lens re-design to some degree.  Having software optimization is pretty good modern solution, of course you can disable lens recognition and have image with lots of "character".

[jonoslack]

Hmm well

 

I don't think you can look 'purely at corner smearing' .......

Hmm well. I've not seen anything I would call 'smearing ' in any of the lenses I've tested. 

"Smearing' seems to be rhe sole preserve of the A7 family of sensors (where it can be very obvious)

 

What I see in the SL (when compared to the M240 ) is a slight reduction of both contrast and detail which I would think was commensurate with the need for a little more vignettting correction. The result is marginally softer corners than you'd get with an M.

 

Being an odd kind of guy I quite often like to plonk my motif right in a corner, so this stuff matters to me in more than pixel peeping terms.

 

When I was using an A7 I finally decided that even 50mm were guilty occasionally (I have a dreadfully smeary corner taken with the 50 APO at f8 for instance.) I still suspect there is more to this on the Sony than just the cover glass issue which I would expect to produce 'soft' rather than 'smeary'. The smeariness is unpredictable and really does spoil images. So I gave up!

 

on the SL I've now checked at infinity:

WATE

28, lux, elmarit and summicron

35 lux and summicron

50 lux and APO

 

I'd consider all these lenses to be perfectly good on the SL (with the caveat that it's worth stopping down a bit if you're focusing near infinity). 

[k-hawinkler]

Advertisement (gone after registration)

Any evidence to share? TIA.

[svenjosh]

"Smearing" seems to be rhe sole preserve of the A7 family of sensors (where it can be very obvious)

 

When I was using an A7 I finally decided that even 50mm were guilty occasionally (I have a dreadfully smeary corner taken with the 50 APO at f8 for instance.) I still suspect there is more to this on the Sony than just the cover glass issue which I would expect to produce 'soft' rather than 'smeary'. The smeariness is unpredictable and really does spoil images. So I gave up!

 

 

 

Thanks Jono, I have the A7s and A7rII and totally agree. Smearing is unpredictable in the A7 series and the loss resolution afftects the center too. The 50 lux is a good example for this. There is more to this than just "thick cover glass". Even the Kolari mod does not get rid of these issues. In fact Kolari mod affects color cast more than smearing, which proves Jono's point.

Edited November 10, 2015 by svenjosh

[powermacs]

We can (apparently) expect the R lenses to perform just fine with the SL.

 

The truism that M lenses perform best on the M cameras (by reason of thinner cover glass on the sensor if nothing else), followed closely by the SL and a long way by the Sony A7 series cameras (we probably don't need to be concerned over them, but as many here use them as alternatives to the M it does no harm to put this on the record) has been floating around for a while with limited real testing.  From what I've seen, you need to look for the difference between the M and SL cameras using M lenses, but it is there.

 

So, how many are good, and how many not so good?

 

Looking purely at corner smearing (which is fatal) we know, or can assume:

• lenses 50mm and longer should be no different whether used on the M or SL

• older 35mm designs like the 35 Summicron ASPH have issues in the corners, shot wide open

• the 28 Summilux actually seems to have better resolution in the centre and in the corners on the Sl than on the M(240) - it's marginal, but it is there

• the older 28 Summicron ASPH does not perform quite as well in the corners on the SL as it does on the M(240) - again, you need to look for it, but it's there

Apparently, we can expect the WATE to perform well, the 21 Summilux as well, but perhaps not the 21 SEM (based on its performance on the A7 cameras when tested on the release of the A7r).

 

Please add here user experiences or other conclusions from looking at files online.

 

It looks like you have summarized the findings of Sean Reid. If so, why don't you quote him? 

[ramarren]

Once my SL arrives, and because I have the M Adapter T and a bunch of LTM and M-Mount lenses, I'll be testing how the SL compares when using it vs the M-P typ 240. 

 

The lenses I'll be testing are explicitly the following: 

 

Elmar-M 24mm f/3.8 ASPH

Color Skopar 28mm f/3.5

Color Skopar 35mm f/2.5 (LTM)

Summilux 35mm f/1.4 v2

Nokton 40mm f/1.4 MC

M-Rokkor 40mm f/2 mark II

Color Skopar 50mm f/2.5 (LTM)

Nokton 50mm f/1.5 ASPH (LTM)

Summarit-M 75mm f/2.4

M-Rokkor 90mm f/4

Hektor-M 135mm f/4.5

 

That's a diverse bunch of lenses, ranging in age from current to about 1959 production, with a diverse range of optical designs. It will prove interesting to see how they perform in comparison on these two cameras. I'll enable the same lens code corrections that I use with the M-P when I render them. 

 

In the end, it is always all about the specific lens to sensor match that determines its best performance. No amount of lens correction can get around a fundamentally bad match between sensor and lens. 

 

My experience with these same lenses fitted via adapter on the Sony A7 model was poor. I ultimately found a lens code that allowed the Color Skopar 28 to do reasonably well for B&W, not so well for color. The Nokton 40 and M-Rokkor 40 worked satisfactorily. The Nokton 50 had some problems but was mostly quite good. The M-Rokkor 90 and Hektor 135 worked well. I bought the Elmar 24 and Summarit 75 after I'd sold the A7; never tried the 'Lux 35, CS 35, or CS 50 on it. Neither of the Ultron 28/2 or CS 21/4 worked on the A7 particularly well either. 

 

My mainstays of these today on the M-P are the Elmar 24, 'Lux 35, Nokton 50, and Summarit 75; they all work beautifully. The M-Rokkor 90 and Hektor 135 work well too, I just don't use them very much. I use the others primarily on M4-2 and CL film bodies. 

Edited November 10, 2015 by ramarren

[jonoslack]

It looks like you have summarized the findings of Sean Reid. If so, why don't you quote him? 

Sean has certainly done lots of work on this, comparing M240 with SL. I'm sure the OP didn't mean not to credit him, but people with pay sites don't usually want to be quoted directly. 

 

Still, if anybody hasn't read them yet I thoroughly recommend that you do so (at reidreviews.com)

Edited November 10, 2015 by jonoslack

[jonoslack]

Any evidence to share? TIA.

Hi there

yes indeed. Much too much evidence!  It's going to take some time to present it in a useful way.

im going to leave the M to SL comparisons to Sean Reid (who has done a fantastic job) Ie done a three way comparison at infinity. 

 

Its not something I enjoy doing all that much!

[Peter H]

It looks like you have summarized the findings of Sean Reid. If so, why don't you quote him? 

 

Hello. Welcome to the forum.

 

You'll find it's a friendly place. Hope you enjoy the conversation.

[edwardkaraa]

Sean has certainly done lots of work on this, comparing M240 with SL. I'm sure didn't mean not to credit him, but people with pat sites don't usually want to be quoted directly.

 

Still, if anybody hasn't read them yet I thoroughly recommend that you do so (at reidreviews.com)

Totally agreed. I normally don't go by others' descriptions. I like to see the results with my own eyes and judge for myself how acceptable are the results by my own standards. I think Reid's comparisons between the M and SL are very revealing.

[IkarusJohn]

It looks like you have summarized the findings of Sean Reid. If so, why don't you quote him?

I'm not, actually. I have referenced him elsewhere and certainly his findings are consistent on a couple of issues. As Jono says, his review is worth reading, but this thread is intentionally not about Sean. The above summary includes the findings of others.

[Peter H]

I'm not, actually..............................

 

In which case, and ironically enough, Powemacs now has.

[jonoslack]

Totally agreed. I normally don't go by others' descriptions. I like to see the results with my own eyes and judge for myself how acceptable are the results by my own standards. I think Reid's comparisons between the M and SL are very revealing.

HI Edward

I think it's a pity not to show the results from the Sony A7 in the same breath though. Sean is exhibiting issues which are rarely going to be a problem in the real world (at least, not to me). From my observations, if you arbitrarily put the M at 100% and the SL at 75% I think I'd put the A7ii at 25% on the same scale.

 

Put it another way - I spent 3 weeks in Crete shooting almost exclusively with the SL and M lenses, and this was before the profiles were fully optimised. . . . . . . . . I haven't got a single shot which would have been noticeably better on the M . . but I've got a lot of shots (mostly wide open in bright sunshine) . . which you absolutely couldn't have shot with the M without using ND filters (which i didn't have with me:  (remember 1/4000 with ISO 200 vs 1/8000 at ISO 50). 

[edwardkaraa]

HI Edward

I think it's a pity not to show the results from the Sony A7 in the same breath though. Sean is exhibiting issues which are rarely going to be a problem in the real world (at least, not to me). From my observations, if you arbitrarily put the M at 100% and the SL at 75% I think I'd put the A7ii at 25% on the same scale.

 

Put it another way - I spent 3 weeks in Crete shooting almost exclusively with the SL and M lenses, and this was before the profiles were fully optimised. . . . . . . . . I haven't got a single shot which would have been noticeably better on the M . . but I've got a lot of shots (mostly wide open in bright sunshine) . . which you absolutely couldn't have shot with the M without using ND filters (which i didn't have with me: (remember 1/4000 with ISO 200 vs 1/8000 at ISO 50).

Thank you Jono. That sounds really good. I agree about your assessment putting the SL at 75% and A7 at 25% from my own experience with the Sony, and from what I've seen so far from the SL. The only thing keeping me from preordering is that I'm waiting until I see your promised infinity comparisons

[k-hawinkler]

HI Edward

I think it's a pity not to show the results from the Sony A7 in the same breath though. Sean is exhibiting issues which are rarely going to be a problem in the real world (at least, not to me). From my observations, if you arbitrarily put the M at 100% and the SL at 75% I think I'd put the A7ii at 25% on the same scale.

 

Put it another way - I spent 3 weeks in Crete shooting almost exclusively with the SL and M lenses, and this was before the profiles were fully optimised. . . . . . . . . I haven't got a single shot which would have been noticeably better on the M . . but I've got a lot of shots (mostly wide open in bright sunshine) . . which you absolutely couldn't have shot with the M without using ND filters (which i didn't have with me:  (remember 1/4000 with ISO 200 vs 1/8000 at ISO 50). 

 

Well Jono, I consider this a pretty meaningless statement.

 

"From my observations, if you arbitrarily put the M at 100% and the SL at 75% I think I'd put the A7ii at 25% on the same scale."

 

Here you seem to be in an SL sales mode.  The facts please, just the facts!

 

A more interesting comparison would be with the A7r2 anyway.

I am the first to point out that there are problems with certain Leica M lenses on any of the A7/2/s/r/2 cameras to a different degree.

 

But there are also other M lenses that do just fine on a stock A7r2.

Some M lenses do not sufficiently improve with a Kolari replacement of the sensor cover glass while others do.

 

This thread is about

"M lens performance on the SL"

I respect that and only make my point, as you constantly seem to refer to Sony A7 without providing any evidence so far that I know of.

 

Edited November 10, 2015 by k-hawinkler

[Advertisement]

Hallo,

Look here: Leica L-Mount

[CheshireCat]

Smearing is unpredictable in the A7 series [...] There is more to this than just "thick cover glass". Even the Kolari mod does not get rid of these issues. In fact Kolari mod affects color cast more than smearing, which proves Jono's point.

 

As far as I know:

 

- The smearing on the A7 series is only caused by the thick cover glass. What else ?

- The color-vignetting is caused by the sensor technology. As a matter of fact, the A7R2 sensor has no color-vignetting with any of my lenses.

- The Kolari mod affects smearing and global color-cast, not color-vignetting. I guess the global color-cast is affected due to (partial) removal of the IR filter, which is to be expected.

 

Is there more to these ?