Leica Apo-Summicron-SL2.0/35mm - average performer?

[thighslapper]

Advertisement (gone after registration)

100% crop of Sammy's rat

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!

[thighslapper]

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

[thighslapper]

For some reason the crops end up showing at 2:1 ..... this is what they should look like on a screen ....

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!

 

[thighslapper]

Bearing in mind the forum compression tries it's hardest to ruin images I cannot see any difference between the S1R and Sl2 image quality or sharpness.

A startling revelation considering they use the same sensor with only minor tweaks to compensate for the vagaries of M lenses. 

Edited January 26, 2020 by thighslapper

[scott kirkpatrick]

OK, I'll see about doing two shots in which the S1R gets an extra half stop on the shutter speed, I use same ISO (probably 400), and keep the aperture at 2.0 for both.  If there is any difference in sharpness due to different sensor coverings, microlenses, etc, that would show up most wide open.

But I won't let either of our cats be photographed looking like a piece of left luggage.  Is that a black Lab (a noble creature)?

Edited January 26, 2020 by scott kirkpatrick

[Sohail]

27 minutes ago, thighslapper said:

Bearing in mind the forum compression tries it's hardest to ruin images I cannot see any difference between the S1R and Sl2 image quality or sharpness.

A startling revelation considering they use the same sensor with only minor tweaks to compensate for the vagaries of M lenses. 

Same sensor, yes, but the thickness of the cover glass is surely not a minor matter. Lloyd Chambers has looked into this fairly extensively noting that:

"Ray angle to the sensor is a factor in both color uniformity and sharpness; rays entering steeply are more refracted by the sensor cover glass, including the IR blocking layer (which tends to block more red resulting in a cyan cast). As well, higher angles impair image sharpness, a phenomenon that can be of great severity with rangefinder lenses that are adapted to mirrorless cameras. .... The effect of ray angle varies by “camera”—meaning the thickness and type of sensor cover glass. It is one reason that Leica uses a very thin sensor cover glass on its Leica M digital cameras."

Sean Reid has also commented extensively on the thickness of sensor glass for Leica M cameras -- as has Jonathan Slack. See here: https://www.slack.co.uk/lumix-s1.html

See also Kolari, a German 'company' that will modify the sensor stack to optimise optical performance. See here https://phillipreeve.net/blog/review-kolari-ultra-thin-sensor-stack-modification/

 

 

[scott kirkpatrick]

Advertisement (gone after registration)

OK.  Tai is pretty bored with all this, but two shots, both f/2.0, ISO 800.  SL2 at 1/30 and S1R at 1/20, both handheld.  I did a little highlight suppression and a tiny shadow boost on each one, copying the parameters across so the processing is identical.

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!

U1000958 by scott kirkpatrick, on Flickr

P1022795 by scott kirkpatrick, on Flickr

as before, click through to Flicker, and then click twice more to see the pixels and whiskers and plastic zipper detail.  It looks like I lifted the S1R shot a little more than was needed.  The fur in the first one is correct, the second a little light.

Edited January 26, 2020 by scott kirkpatrick

[thighslapper]

9 minutes ago, Sohail said:

Same sensor, yes, but the thickness of the cover glass is surely not a minor matter. Lloyd Chambers has looked into this fairly extensively noting that:

"Ray angle to the sensor is a factor in both color uniformity and sharpness; rays entering steeply are more refracted by the sensor cover glass, including the IR blocking layer (which tends to block more red resulting in a cyan cast). As well, higher angles impair image sharpness, a phenomenon that can be of great severity with rangefinder lenses that are adapted to mirrorless cameras. .... The effect of ray angle varies by “camera”—meaning the thickness and type of sensor cover glass. It is one reason that Leica uses a very thin sensor cover glass on its Leica M digital cameras."

Sean Reid has also commented extensively on the thickness of sensor glass for Leica M cameras -- as has Jonathan Slack. See here: https://www.slack.co.uk/lumix-s1.html

See also Kolari, a German 'company' that will modify the sensor stack to optimise optical performance. See here https://phillipreeve.net/blog/review-kolari-ultra-thin-sensor-stack-modification/

 

 

Most of this was provoked by those complaining that DXO used the S1R to test the 35/2 rather than on the SL2.

Basically we are trying to show that it makes no difference at all ..... as for L mount lenses the cameras perform essentially identically. Chaemono will no doubt chip in with assertions that shadow recovery and noise etc. etc. do differ slightly when processed, but that doesn't make much odds when it comes to testing lenses and basic sensor resolution. I use both cameras concurrently and have noticed zero difference in normal usage. Image output from both is superb. 

Edited January 26, 2020 by thighslapper

[Sohail]

14 minutes ago, thighslapper said:

Most of this was provoked by those complaining that DXO used the S1R to test the 35/2 rather than on the SL2.

Basically we are trying to show that it makes no difference at all ..... as for L mount lenses the cameras perform essentially identically. Chaemono will no doubt chip in with assertions that shadow recovery and noise etc. etc. do differ slightly, but that doesn't make much odds when it comes to testing lenses and basic sensor resolution. I use both cameras concurrently and have noticed zero difference in normal usage. Image output from both is superb. 

Except that using language like "essentially identically" or "minor tweaks" doesn't really help much if you want to resolve this question scientifically as DXO has sought to do albeit with dubious methodology.  Neither does it help to claim that you have noticed "zero differences" unless we're to take your eyes as a a scrupulously reliable measure. You may well be right (DXO too) but let's see your findings and methods first and then we can all decide and resolve this in an evidence-based way.

Edited January 26, 2020 by Sohail

[IkarusJohn]

11 minutes ago, Sohail said:

Except that using language like "essentially identically" or "minor tweaks" doesn't really help much if you want to resolve this question scientifically as DXO has sought to do albeit with dubious methodology.  Neither does it help to claim that you have noticed "zero differences" unless we're to take your eyes as a a scrupulously reliable measure. You may well be right (DXO too) but let's see your findings and methods first and then we can all decide and resolve this in an evidence-based way.

What exactly are you looking for?

What Scott & Thighslapper have shown pretty clearly is that, in isolation, it would be impossible to look at either image and say with anything more than a wild guess which image was taken with which camera.  They have also shown, side by side, the output is almost indistinguishable.

As Thighslapper comments above, these are excellent cameras, and any differences are minimal at best, and the output from 35 Summicron-SL looks fabulous.

The problem with the DXO conclusion about the APO Summicron-SL 35 is that it suggests that the lens falls well short of the designer’s claims.  DXO’s conclusions don’t stand up.  It has been a useful discussion.  I’m not in the market for this lens, as I’m not a huge fan of this focal length, and I have the pre-asph 35 Summilux-M and the XCD 45 P on order.

I am, however, holding out for the 28 Summicron-SL.  I hope it doesn’t come too soon ...

[thighslapper]

22 minutes ago, Sohail said:

Except that using language like "essentially identically" or "minor tweaks" doesn't really help much if you want to resolve this question scientifically as DXO has sought to do albeit with dubious methodology.  Neither does it help to claim that you have noticed "zero differences" unless we're to take your eyes as a a scrupulously reliable measure. You may well be right (DXO too) but let's see your findings and methods first and then we can all decide and resolve this in an evidence-based way.

I'm sorry, but I am not blind, and viewed at 1:1 on a 5k 27" iMac screen there is no difference at all. 

You can prove scientifically that there is a difference but if you cannot actually see it on a screen or in a print then it is irrelevant. 

 

[Sohail]

3 minutes ago, IkarusJohn said:

What exactly are you looking for?

What Scott & Thighslapper have shown pretty clearly is that, in isolation, it would be impossible to look at either image and say with anything more than a wild guess which image was taken with which camera.  They have also shown, side by side, the output is almost indistinguishable.

As Thighslapper comments above, these are excellent cameras, and any differences are minimal at best, and the output from 35 Summicron-SL looks fabulous.

The problem with the DXO conclusion about the APO Summicron-SL 35 is that it suggests that the lens falls well short of the designer’s claims.  DXO’s conclusions don’t stand up.  It has been a useful discussion.  I’m not in the market for this lens, as I’m not a huge fan of this focal length, and I have the pre-asph 35 Summilux-M and the XCD 45 P on order.

I am, however, holding out for the 28 Summicron-SL.  I hope it doesn’t come too soon ...

I don't have a dog in this race either. I'm not interested in the 35 Summicron-SL. I too prefer a 28mm.

[scott kirkpatrick]

I think that the detail is just a bit better in awake cat (SL2) than in sleeping cat (S1R), but I don't think it proves anything.

[Sohail]

1 minute ago, thighslapper said:

I'm sorry, but I am not blind, and viewed at 1:1 on a 5k 27" iMac screen there is no difference at all. 

You can prove scientifically that there is a difference but if you cannot actually see it on a screen or in a print then it is irrelevant. 

 

I don't dispute you're blind or that you can't see a difference in this single instance. It would be more meaningful though to do this across a wider range of shooting scenarios. We can all think of specific not-so challenging shooting scenarios where an iPhone image is no different from that of an SL2. It wouldn't prove anything. Neither does this. 

[thighslapper]

1 minute ago, Sohail said:

I don't dispute you're blind or that you can't see a difference in this single instance. It would be more meaningful though to do this across a wider range of shooting scenarios. We can all think of specific not-so challenging shooting scenarios where an iPhone image is no different from that of an SL2. It wouldn't prove anything. Neither does this. 

I've done all this before with various Leica's and posted multiple images in the past ..... but achieving exactly the same conditions in a variety of settings is very challenging ..... which is why everyone ends up using static well illuminated test charts (or vegetables in Sean Reids case) which are hardly representative of what we all take photos of. There are just too many variables, and if you have to compensate for them by different processing then the results are questionable. 

I stand by the assertion that if you take any random image with both of these cameras under the same conditions you will not find any appreciable differences that would show up on a high resolution screen or A1/A2 print. 

On another long running thread they have been hammering on about differences between the SL2 and X1D ... and there are indisputable differences which are inherent in the fact the pixels are bigger and therefore have a better dynamic range and light catching capability ..... but the ultimate difference in print and on screen is no longer as great as you might expect it to be. All this technology is fast reaching a point where the reproductive media are unable to show the differences in the inherent quality of the capture equipment.

Of course there is always going to be someone who feels the need to drastically crop their 1000 mpx image they have taken from half a kilometre away ..... but I am not one of the paparazzi after Harry and Megan so it's currently of no concern to me, even if DXO gives it a cracking score.  

[MarkinVan]

Well, I have prepaid for the SL 35mm F2 and have been told it should be here in about 2 weeks. The DOX report has not made me decide to cancel my order. I purchased it partially because of what the Leica lens designer said, for auto focus and smaller size on the SL, I find I consistently have my 35mm mounted on my M240 over the 50mm, the good reviews and positive comments from current owners and lastly some Leica Marketing on their website for the SL 35mm F2. 

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!

This last bit really made me hand over the credit card and I am day dreaming about the possibilities on my new SL2. 

I hope to have the lens for a long time. 

[Chaemono]

Using the S1R to test the APO 35 Summicron-SL doesn’t change the fact that there isn’t a built-in profile to instruct supporting RAW converters to correct for distortion as DxOMark claims.  They make this up, they make up the rest.  It doesn’t matter what camera they use.  The S1R gives them an alibi for their lies.

[Sohail]

5 hours ago, thighslapper said:

I've done all this before with various Leica's and posted multiple images in the past ..... but achieving exactly the same conditions in a variety of settings is very challenging ..... which is why everyone ends up using static well illuminated test charts (or vegetables in Sean Reids case) which are hardly representative of what we all take photos of. There are just too many variables, and if you have to compensate for them by different processing then the results are questionable. 

I stand by the assertion that if you take any random image with both of these cameras under the same conditions you will not find any appreciable differences that would show up on a high resolution screen or A1/A2 print. 

It's all very well to pooh-pooh Reid's findings (nothing is 100% conclusive) but there is at least some rigour in his tests and we all know how he goes about substantiating his claims imperfect though they may be. Failing that, what you're inviting us to do is to accept your "assertion" that there is no "appreciable difference" without really telling us in what specific sense: micro contrast, colour fidelity, chromatic aberrations, geometric distortions, vignetting? All we have is your cocksure untested assertion about any random image under the same conditions. Not particularly persuasive. That being said, let me be clear: I have no dog in this race. I'm willing to be persuaded by substantiated claims. Further, I think it's a reasonable objection to hold that DXO should have used an SL2 body for their tests.

Edited January 27, 2020 by Sohail

[Donzo98]

1 minute ago, Sohail said:

It's all very well to pooh-pooh Reid's findings (nothing is 100% conclusive) but there is at least some rigour in his tests and we all know how he goes about substantiating his claims imperfect though they may be. Failing that, what you're inviting us to do is to accept your "assertion" that there is no "appreciable difference" without really telling us in what specific sense: micro contrast, colour fidelity, chromatic aberrations, geometric distortions, vignetting? All we have is your cocksure untested assertion about any random image under the same conditions. Not particularly persuasive. That being said, let me be clear: I have no dog in this race. I'm willing to be persuaded by substantiated claims. Further, I think it's a reasonable objection to state that DXO should have used an SL2 body for their tests.

I’m in Thighslapper’s camp...

SL2 vs S1R... same lens, same image. He finds no difference on a 5K monitor... 

That’s good enough for me... 

Not sure why it’s not good enough for you.

[Advertisement]

Hallo,

Look here: Leica L-Mount

[Mr.Q]

DXOMark claims that the images from the 35SL aren't as sharp as the Sony/Canon lenses in the outer zones and corners from F4 to F8.

The best way to test this would to take a landscape shot at infinity, with the horizon tilted to the corners.