Leica SL2 and SL 35/50/90mm - Chasseur d`Images

[Ivar B]

Advertisement (gone after registration)

I noted in another forum thread that the French magazine Chasseur d`images had tested the Leica SL2 with Apo-SL 35/50/90. My French is certainly not fluent, but I decided to buy the magazine in any case and I believe I was able to get the message somehow. As I read the review, the SL2 is certainly praised. The lenses also get very favourable reviews, with the 90mm scoring the highest, and the 50mm at the other end. 

What surprises me a bit is the comments that the SL lenses do not perform as well on the Panasonic S1R as they do on the SL2, despite sensors with the same resolution. It is guessed that the design of the sensor of the SL2 with M-optimization also gives advantages with other lenses, and that the thicker glass in front of the Panasonic sensor also reduces sharpness. I use the S1R and the Leica CL and I would love to use the Leica lenses on the Panasonic, but it is a pity that they perhaps then cannot be used to their full advantage. 

I don`t know if the differences are purely academic and observable in practice or not, but are there any advantages to the Pansonic design? Why a thicker glass if it impairs quality?

 

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!

[nicci78]

Thicker stack filter = better video

Thinner stack filter = better photo

So pick your poison 

[Jeff S]

I would expect Leica to be better able to incorporate software and/or other enhancements to optimize lens and system performance in subtle ways that Panasonic cannot.  I stick with native lenses on all my systems; in fact I choose those systems in part because of native lens options. Sharing a mount is only part of design/engineering considerations.

Jeff

Edited October 18, 2020 by Jeff S

[Jon Warwick]

I ended up with an SL2 instead of the S1R, and use it with a SL 50 APO.

The S1R produced VERY sharp RAW files to my eyes, but i personally found it challenging to tame the files to not look “too sharp”. In that sense I am am referring to edge sharpness / acuity, not resolution per se.

The combo of the SL2 and SL APO lens produces a very detailed image, but with greater smoothness and less aggressive acuity / edge sharpness to my eyes, which i personally prefer because i found it easier to get a more natural looking output.

As an aside, I thought I’d use M lenses on the SL2, but quickly moved to the native SL prime ....better image quality and balance with the SL2 body were my reasons.

I kept the M lens to work with my M analogue body - again, a superb balance and work flow.

From all the above, I’ve certainly learnt to keep to native body & native lenses, and not try and mix and match across the Leica or other systems.

 

[scott kirkpatrick]

I was using the S1R with the 35 APO before the SL2 came out, and have some very impressive images that resulted.  So my vote is for the "too close to make a fuss over" position.

Examples here (with 35 and 75 APOs).  https://flic.kr/s/aHsmGdkgCQ 

[Stuart Richardson]

I have not noticed any meaningful difference in sharpness on my S1 and SL2, other than those inherent in the sensors. That does not mean it is not there, however. I certainly have noticed it for M lenses.

Am I reading it correctly that they only gave the 50mm APO Summicron SL 3 out of 5? I find that pretty crazy...it is probably the best lens I have ever used.

[didier]

Advertisement (gone after registration)

On 10/19/2020 at 6:11 PM, Stuart Richardson said:

I have not noticed any meaningful difference in sharpness on my S1 and SL2, other than those inherent in the sensors. That does not mean it is not there, however. I certainly have noticed it for M lenses.

Am I reading it correctly that they only gave the 50mm APO Summicron SL 3 out of 5? I find that pretty crazy...it is probably the best lens I have ever used.

Crazy ... I used to buy Chasseur d’Images regularly for years.

I stopped as I found they got a real bias against Leica : with every new Product from Leica, I found they criticised too much, losing their objectivity. 

I just stopped buying the magazine for that very reason. Some other French magazines are a better read in my View...

[BernardC]

On 10/19/2020 at 12:11 PM, Stuart Richardson said:

Am I reading it correctly that they only gave the 50mm APO Summicron SL 3 out of 5?

That's their "coup de coeur" rating, which takes price into account. The technical rating is 4 of 5 stars. Presumably they would give-out the fifth 5 star if the corners were as sharp as the center.

These DxO-type lens ratings are a bit obsolete. It's telling us that the lens is sharp enough for pixel-perfect A3+ or A2 prints, what Americans know as 16x20. I don't know what their standard is, but I suspect that this indicates the sensor's full resolution. In other words, the lens "out-resolves" the sensor at every aperture until F:8.

Back in the film days Cd'I tested with higher resolution stock, and lenses were not as good, so having a "sharpness" number made some sense, even if it could only be achieved with Tech Pan and a good tripod.

That tells me very little about how I will get-on with a lens. It's even worse than comparing car by top speed: it's saying that "the top speed is faster than the speed limit." What about every other aspect of a lens' performance? Colour, saturation, contrast, flare, handling, etc?

[Ivar B]

On 10/19/2020 at 6:11 PM, Stuart Richardson said:

I have not noticed any meaningful difference in sharpness on my S1 and SL2, other than those inherent in the sensors. That does not mean it is not there, however. I certainly have noticed it for M lenses.

Am I reading it correctly that they only gave the 50mm APO Summicron SL 3 out of 5? I find that pretty crazy...it is probably the best lens I have ever used.

I think the 3 out of 5 is their purely subjective sentiment - considering prize and cost. The "scientific" test is 4 out of 5.

The German Magazine also tested this lens (then on a Panasonic S1R) and their verdict was about the same as this one.

A very good lens, but could not reach the score of the Panasonic Pro 1.4/50mm. I expect it may be partly due to using a Panasonic body then.

Edited October 21, 2020 by Ivar B

[Ivar B]

On 10/19/2020 at 11:09 AM, scott kirkpatrick said:

I was using the S1R with the 35 APO before the SL2 came out, and have some very impressive images that resulted.  So my vote is for the "too close to make a fuss over" position.

Examples here (with 35 and 75 APOs).  https://flic.kr/s/aHsmGdkgCQ 

Beautiful photos! Makes me want to visit!

[Stuart Richardson]

Thank you for helping. To be honest, it kind of reinforces the idea to me that lens tests are generally useless to me. I can learn more by MTF and looking at pictures than by any ratings. I am using the 50mm APO on the SL2, and using a sturdy tripod, it is sharp at f4 over the whole frame at 200 megapixels, with no measurable distortion, no color aberrations and it can hold that performance at any focusing distance in a lens that uses a 67mm filter ring, focus silently and quickly and balances perfectly on the cameras it is intended for. I do not understand how it can come out at 3 out of 5 unless every other lens is a 1.

[Stuart Richardson]

I shot these yesterday. They are taken on the SL2 with the multishot mode with the 50mm at f4 at almost 200mp. Crops from the center and upper left. It is not as sharp at the extreme corner as it is on center, but it is 200mp and f4. At this level, I think that minor difference is more than acceptable. Stopping down to f8 reduces the center contrast a bit, but evens out the sharpness. It is a trade off. 

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!

Edited October 21, 2020 by Stuart Richardson

[FlashGordonPhotography]

The S1R has one extra layer of glass over the sensor. In my personal tests I have seen that my SL2 can produce images with slightly more accutance some of the time with SL lenses. But not always and the difference is subtle. In some tests I could see it. In others I couldn't. I started to think I might be imagining it but it's there. At first I thought it was probably due to slightly different focusing but more structured tests do show a very very very slight advantage to the Leica.

So slight it would not cause me to choose one camera over the other.

Gordon

[jrp]

12 hours ago, Ivar B said:

A very good lens, but could not reach the score of the Panasonic Pro 1.4/50mm. I expect it may be partly due to using a Panasonic body then.

Like others, I struggle to relate what I see in real life to the test results.  (We had a similar disconnect with DXO's test of the 35mm Summicron SL.)

To add to the confusion, here is a "rather disappointing"  test of the supposedly better Panasonic 1 .4/50mm 

https://www.lesnumeriques.com/focales-fixes/panasonic-lumix-s-pro-50mm-f1-4-p52031/test.html

 

 

[BernardC]

5 hours ago, jrp said:

Like others, I struggle to relate what I see in real life to the test results.  (We had a similar disconnect with DXO's test of the 35mm Summicron SL.)

To add to the confusion, here is a "rather disappointing"  test of the supposedly better Panasonic 1 .4/50mm 

https://www.lesnumeriques.com/focales-fixes/panasonic-lumix-s-pro-50mm-f1-4-p52031/test.html

I don't find it strange that they give 4 of 5 stars as a technical score. As they explain, the corners are slightly softer than the center at some apertures, so it's shy of a perfect score. Do they give 5 stars to other lenses? I imagine that this score is reserved for some APO telephotos that are sharp edge-to-edge wise-open, and maybe some slower macro lenses. That's how it used to be when I last read Cd'I.

The "value" score is 3 of 5, but that's hardly surprising.

The Panasonic 50/1.4 test linked above also gets 4 of 5 technical stars, although the comments are a bit more critical. It needs to be stopped-down significantly to achieve full sharpness.

...which brings me back to my original point: these lenses get 4 stars, because they aren't perfect, but they are too good to get just 3 stars. It's not a very useful scale, is it?

 

Some video tests are more useful. Resolution is barely mentioned; any modern lens should easily resolve 4K, or even 8K. Besides, cinematographers often have to work hard to reduce resolution so the images don't look to harsh (especially with beauty shots). What really matters lies elsewhere: distortion, flare, handling, blur, colour, breathing, ramping, etc. Even then, it's not a matter of finding "the best." What's important is describing what you will get. You want to establish a look for your project (or several if it suits a narrative function), and find lenses that deliver that look.

[scott kirkpatrick]

22 hours ago, Stuart Richardson said:

I shot these yesterday. They are taken on the SL2 with the multishot mode with the 50mm at f4 at almost 200mp. Crops from the center and upper left. It is not as sharp at the extreme corner as it is on center, but it is 200mp and f4. At this level, I think that minor difference is more than acceptable. Stopping down to f8 reduces the center contrast a bit, but evens out the sharpness. It is a trade off. 

I did a quick comparison of the SL2's multishot high res mode with the S1R's when the SL2 arrived, but I haven't shlepped a tripod outside for a serious landscape study.  My impression was that Leica had done a better job of controlling artifacts that you see in the Lumix multishot at 100%, but left the image slightly soft, needing some post sharpening.  What's your impression?

[Stuart Richardson]

My impression is that they both lead to sharper photos and more detail than their base images do. The SL2 is also significantly more detailed than the S3 when using the multishot mode. But yes, the sharpness impression of both is not as high as they are in their single shot modes, so it is not a true 200mp in comparison to the 47mp. My impression is that the Leica does have less of the zipper effect/weird aliasing that these files tend to have than the S1 does, but I do not have the S1R. A stable platform is critical. One of the two modes of dealing with motion is better...I think it is the mode which does not compensate for blur, but I would have to do a controlled test, as it does not seem to say anywhere in the metadata which mode was used. I have found that it is better to sharpen with a significantly higher radius and amount, but 0 detail and a pretty high masking. This does the best job of enhancing the edges while avoiding bringing out the aliasing.

[Eclectic Man]

I have an SL2 and the Apo 50mm f/2 L lens.  Testing it I am very impressed by the lack of distortion.  Photographing stars at the edges of the frame coma is very well handled, only visible on high magnifications, and absolutely no chromatic aberration evident.  Much, much better than any other 50mm lens I have tried.

[Jeff S]

7 hours ago, Eclectic Man said:

I have an SL2 and the Apo 50mm f/2 L lens.  Testing it I am very impressed by the lack of distortion.  Photographing stars at the edges of the frame coma is very well handled, only visible on high magnifications, and absolutely no chromatic aberration evident.  Much, much better than any other 50mm lens I have tried.

Unlike S lenses on an S body, which would reveal any distortions in the optical viewfinder, SL lenses on the SL/SL2 can benefit from software enhancements to eliminate distortion.

Jeff

Edited October 24, 2020 by Jeff S

[Advertisement]

Hallo,

Look here: Leica L-Mount

[scott kirkpatrick]

On 10/23/2020 at 3:33 AM, Stuart Richardson said:

My impression is that they both lead to sharper photos and more detail than their base images do. The SL2 is also significantly more detailed than the S3 when using the multishot mode. But yes, the sharpness impression of both is not as high as they are in their single shot modes, so it is not a true 200mp in comparison to the 47mp. My impression is that the Leica does have less of the zipper effect/weird aliasing that these files tend to have than the S1 does, but I do not have the S1R. A stable platform is critical. One of the two modes of dealing with motion is better...I think it is the mode which does not compensate for blur, but I would have to do a controlled test, as it does not seem to say anywhere in the metadata which mode was used. I have found that it is better to sharpen with a significantly higher radius and amount, but 0 detail and a pretty high masking. This does the best job of enhancing the edges while avoiding bringing out the aliasing.

The multishot version of "200 MPx" relies on big fat 4.3 micron pixels.  A true 200 MPx chip would have a pixel spacing of a bit over 2 microns.  Since we don't yet know how to make a pixel that small that will hold 2^14 electrons when full and give 14 stops of dynamic range, some clever image processing is needed to enhance the fat pixels.  I like your suggestion of using large radius to emphasise real edges without amplifying the subpixel artifacts.