DXO Mark Leica M10 Score

[onasj]

Advertisement (gone after registration)

I’m a huge Leica fan, and not a big fan of DXO’s pseudoscientific ratings. But since I’m also a career scientist I have to note the simple truth that the sensor on the Sony a7riii is noticeably better than the sensor on the M10. I am fortunate to own both cameras, as well as some of the best lenses both systems offer currently. Both cameras do just fine when not pushed hard, as some of the examples posted in this thread show. But starting around ISO 6400-12800+, in terms of sharpness (holding on to detail), DR, and low-light sensitivity, the a7riii is better, and not just by a little bit (maybe by 1-2 stops).

 

Again, I consider both cameras to be superb, with each excelling in its own way, and I don’t mean to rain on anyone’s parade. But it’s simply not the case that the Leica’s *sensor* is better in any substantial way that I can perceive over the Sony’s *sensor*.

 

Where the M10 is clearly better is:

- The best Leica lenses are MUCH smaller and MUCH lighter than the best Sony FE lenses

- The Leica M10 is a somewhat smaller in size than the A7riii

- The rangefinder experience is fun and involves the photographer in the process in a unique way compared to modern AF cameras.

- The Leica is much easier to use than the Sony (the actual full manual of the a7riii, which I have read in its entirety, is over 600 pages!)

 

Where the Sony is clearly better is:

- When the shoot demands autofocus (sports, moving subjects in low light, etc.)

- When cost is a significant consideration (for example, the a7riii plus three of its best lenses costs about the same as Leica’s 50 APO lens only)

- When sharpness and detail, rather than convenience of portability and ease of use, is the primary goal. A world-class, current-generation 42 MP BSI sensor with world-class DR is simply hard to beat with a good, if slightly dated already, 24 MP sensor.

- When the maddeningly abundant features (several 4K video settings, automatic near-eye focus, pixel shift technology, absolutely silent shooting at 10fps, and literally hundreds of other options) are actually necessary for a particular job.

 

I have no plans to sell either system, and thoroughly enjoy and recommend both to my colleagues and friends, depending on their goals and needs and budget.

 

For those curious: the Leica M lenses don’t fare particularly well on the Sony a7riii. Corner and edge softness is abundant, even at 50 mm and 90 mm focal lengths (somewhat to my surprise). The thickness of the Sony sensor stack, and other factors, ensures that each manufactruer’s lenses perform best on their native body, as expected.

 

Happy new year to all!

Edited December 30, 2017 by onasj

[Chaemono]

And then lets publish our new LUF score. DxO is useless.

Well, let’s not get carried away.

[Advertisement]

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

[jaapv]

I’m a huge Leica fan, and not a big fan of DXO’s pseudoscientific ratings. But since I’m also a career scientist I have to note the simple truth that the sensor on the Sony a7riii is noticeably better than the sensor on the M10. I am fortunate to own both cameras, as well as some of the best lenses both systems offer currently. Both cameras do just fine when not pushed hard, as some of the examples posted in this thread show. But starting around ISO 6400-12800+, in terms of sharpness (holding on to detail), DR, and low-light sensitivity, the a7riii is better, and not just by a little bit (maybe by 1-2 stops).

 

Again, I consider both cameras to be superb, with each excelling in its own way, and I don’t mean to rain on anyone’s parade. But it’s simply not the case that the Leica’s *sensor* is better in any substantial way that I can perceive over the Sony’s *sensor*.

 

Where the M10 is clearly better is:

- The best Leica lenses are MUCH smaller and MUCH lighter than the best Sony FE lenses

- The Leica M10 is a somewhat smaller in size than the A7riii

- The rangefinder experience is fun and involves the photographer in the process in a unique way compared to modern AF cameras.

- The Leica is much easier to use than the Sony (the actual full manual of the a7riii, which I have read in its entirety, is over 600 pages!)

 

Where the Sony is clearly better is:

- When the shoot demands autofocus (sports, moving subjects in low light, etc.)

- When cost is a significant consideration (for example, the a7riii plus three of its best lenses costs about the same as Leica’s 50 APO lens only)

- When sharpness and detail, rather than convenience of portability and ease of use, is the primary goal. A world-class, current-generation 42 MP BSI sensor with world-class DR is simply hard to beat with a good, if slightly dated already, 24 MP sensor.

- When the maddeningly abundant features (several 4K video settings, automatic near-eye focus, pixel shift technology, absolutely silent shooting at 10fps, and literally hundreds of other options) are actually necessary for a particular job.

 

I have no plans to sell either system, and thoroughly enjoy and recommend both to my colleagues and friends, depending on their goals and needs and budget.

 

For those curious: the Leica M lenses don’t fare particularly well on the Sony a7riii. Corner and edge softness is abundant, even at 50 mm and 90 mm focal lengths (somewhat to my surprise). The thickness of the Sony sensor stack, and other factors, ensures that each manufactruer’s lenses perform best on their native body, as expected.

 

Happy new year to all!

The interesting thing here is that Leica do not make their own sensors. It is undisclosed who makes the CL sensor, but it may well be Sony, as all previous APS-C "Leica" sensors were off-the-shelf Sony ones.

It is said that the only reason that Leica did not choose Sony for the  240, Q and SL ( and most likely M10) sensor was that Sony could (would?) not match Leica's specifications.

[M11 for me]

I’m a huge Leica fan, and not a big fan of DXO’s pseudoscientific ratings. But since I’m also a career scientist I have to note the simple truth that the sensor on the Sony a7riii is noticeably better than the sensor on the M10. I am fortunate to own both cameras, as well as some of the best lenses both systems offer currently. Both cameras do just fine when not pushed hard, as some of the examples posted in this thread show. But starting around ISO 6400-12800+, in terms of sharpness (holding on to detail), DR, and low-light sensitivity, the a7riii is better, and not just by a little bit (maybe by 1-2 stops).

 

 

This is exactly what the pseudoscientific ratings (as you call them) show. If you consider the ISO deviation you are just there.

[onasj]

Yes, I’m not saying DXO’s conclusions are necessarily wrong on this case— rather, my objections to their methodology are the lack of transparency in how they arrive at their numeric scores, and the fact that their weighting of different factors in calculating their overall scores (sharpness vs CA vs distortion vs low-light) appears to be inconsistently applied. The conclusion that the a7riii sensor is better in many respects to the sensor in the M10 is consistent with my observations.

 

That said, unless I anticipate very low light or moving subjects or need to shoot video or silently, I take the Leica bag instead of the Sony bag.

[jdlaing]

I have a question:

 

What difference does a DXO score make?

[M11 for me]

Advertisement (gone after registration)

The interesting thing here is that Leica do not make their own sensors. It is undisclosed who makes the CL sensor, but it may well be Sony, as all previous APS-C "Leica" sensors were off-the-shelf Sony ones.

It is said that the only reason that Leica did not choose Sony for the  240, Q and SL ( and most likely M10) sensor was that Sony could (would?) not match Leica's specifications.

And as I understand it is the DESIGN of a sensor that represents the technical knowhow. The building of the sensor is then common technology. And does not Leica design its sensors or make at least part of the design? And as far as I know there are then several suppliers adding parts to the sensor.

[Mark II]

Leica certainly will not design the sensor - it is far too specialised for the company.

 

I strongly suspect that the M10 uses exactly the same sensor that was used in the M240 - which I think was from CMOSIS (I could be wrong about this). Leica's architecture puts the ADC stages off-sensor, and the performance gains at high ISO and lower banding probably come from better conversion electronics as part of the Maestro II update. There may also be some improved high-ISO NR that is applied to the RAW files (my M262 clearly is applying some NR to the DNG files at higher ISOs, as line edges are softer if I select ISO 3200 rather than selecting ISO 200 and then pushing 4 stops in post processing - which I can do given that the sensor is almost ISO-less).

 

To get substantial performance gains needs a significant change of sensor architecture. The latest Sony designs use on-sensor ADC stages for lower noise and faster readout, as well as back-side illumination. The latter could help reduce colour shift at the edges of the frame with some Leica lenses. Another big gain would be to use a sensor design with a global reset, which would improve live view lag without the pain needed to implement a full global shutter or super-fast readout (Olympus).

 

It is not clear why Leica does not switch to Sony sensors. It is possible that - aside from commercial considerations - they have too much IP invested in the Maestro (Fujitsu) architecture to make a change viable without excessive engineering work. But there is no inherent reason why any of the current Sony sensors could not be paired with an equivalent thin cover stack and offset micro-lenses.

[M11 for me]

. . . and CMOSIS was bought by AMS . . . in November 2015

Edited December 30, 2017 by Alex U.

[jaapv]

Leica certainly will not design the sensor - it is far too specialised for the company.

 

I strongly suspect that the M10 uses exactly the same sensor that was used in the M240 - which I think was from CMOSIS (I could be wrong about this). Leica's architecture puts the ADC stages off-sensor, and the performance gains at high ISO and lower banding probably come from better conversion electronics as part of the Maestro II update. There may also be some improved high-ISO NR that is applied to the RAW files (my M262 clearly is applying some NR to the DNG files at higher ISOs, as line edges are softer if I select ISO 3200 rather than selecting ISO 200 and then pushing 4 stops in post processing - which I can do given that the sensor is almost ISO-less).

 

To get substantial performance gains needs a significant change of sensor architecture. The latest Sony designs use on-sensor ADC stages for lower noise and faster readout, as well as back-side illumination. The latter could help reduce colour shift at the edges of the frame with some Leica lenses. Another big gain would be to use a sensor design with a global reset, which would improve live view lag without the pain needed to implement a full global shutter or super-fast readout (Olympus).

 

It is not clear why Leica does not switch to Sony sensors. It is possible that - aside from commercial considerations - they have too much IP invested in the Maestro (Fujitsu) architecture to make a change viable without excessive engineering work. But there is no inherent reason why any of the current Sony sensors could not be paired with an equivalent thin cover stack and offset micro-lenses.

At the introduction of the M240 it was made clear that they did approach Sony first and only went through CMOSIS after they could not come to a satisfactory conclusion. The next supplier was Towerjazz, presumably because of the Panasonic connection The X Series runs Sony sensors. Take your pick...

[Keith_W]

I have a question:

 

What difference does a DXO score make?

 

It can make a lot of difference when it comes to marketing. If it affects the camera you own, and if enough people care about that score, then it will affect your resale value. 

[jaapv]

I don't think the average customer knows about DXO, or even  cares. It is a typical site for pixel-peeping geeks like us, and even then most of us are knowledgeable enough to apply the required pinch of salt to the findings.

[Bill W]

I don't think the average customer knows about DXO, or even  cares. It is a typical site for pixel-peeping geeks like us, and even then most of us are knowledgeable enough to apply the required pinch of salt to the findings.

 

I guess I am average because I knew nothing about DXO till it appeared here. I started to read it and it was giving me a headache so I quit. Took a sip of my Martini and went on to another topic. 

[Chaemono]

Enough people look at it to affect the “image” of the M10. “It’s like a nice, little APS-C camera. But it comes with a red dot.”

[M11 for me]

I have a question:

What difference does a DXO score make?

Lets make the silly comparison with cars. When I evaluate a french car I often see that they have less hp with the same size of engine and might use more petrol than a german car. Now one might ask what difference this makes: None; yes and no. There are still plenty french people buying french cars and even without knowing the strength of your cars engine you can comfortably drive from A to B. I just think that it does not harm to know a few comparable details. So you will find out that BMW has its latest technology in the cars branded BMW and not in the cars branded MINI. You need not know that either. Actually you need know nothing at all: You can even stop reading the LUF :-)

 

People who know more have a more interesting lifes and might know more interesting people that bring back more interesting information. Yes there are cars and art and theatre and cooking and history and newspaper and politics and economy and business and management and hiking and travelling and photography. Who cares. Lets having breakfast now and talking to my wife. She is actually a woman with extremely broad interests and knows a lot of interesting people but does not even know the term DxO. And I don‘t mind.

[Bill W]

Lets make the silly comparison with cars. When I evaluate a french car I often see that they have less hp with the same size of engine and might use more petrol than a german car. Now one might ask what difference this makes: None; yes and no. There are still plenty french people buying french cars and even without knowing the strength of your cars engine you can comfortably drive from A to B. I just think that it does not harm to know a few comparable details. So you will find out that BMW has its latest technology in the cars branded BMW and not in the cars branded MINI. You need not know that either. Actually you need know nothing at all: You can even stop reading the LUF :-)

 

People who know more have a more interesting lifes and might know more interesting people that bring back more interesting information. Yes there are cars and art and theatre and cooking and history and newspaper and politics and economy and business and management and hiking and travelling and photography. Who cares. Lets having breakfast now and talking to my wife. She is actually a woman with extremely broad interests and knows a lot of interesting people but does not even know the term DxO. And I don‘t mind.

 

That sums it up.....

[Chaemono]

That sums it up.....

Why in such a rush? There is more, including ISO 10000/ISO 12800.

[rodf]

Why in such a rush? There is more, including ISO 10000/ISO 12800.

Well I am looking forward to it even though I never shoot at ISO 10,000

Edited January 1, 2018 by rodf

[lucerne]

It takes such a short time to set the camera and take a test shot. Why do we need this continual flow of low quality images to convince us of anything. If you haven't shot at 10000 or 16000 iso on an M10 you are missing out. As they say.... Do Your Own Research.

 

Chaemono. Take a break. Go outdoors and take some real photos.

 

Happy New Year to Leica Forum photographers.

Edited January 1, 2018 by lucerne

[Advertisement]

Hallo,

Look here: Leica M

[Chaemono]

Ok, here at ISO 3200/4000. ISO 10000/12800 later. 50/1.4 Zeiss Planar FE on the α7R III and Noctilux on the M10. Both cameras do a good job but I liked the Sony colors in this instance a bit better after adjusting WB, even though they are not necessarily more accurate. So, I tried to adjust the colors on the M10 to match the α7R III output here. On the Sony, I only adjusted WB to try to match, lifted the Shadows by +100, NR +60, Sharpening +40. On the M10, I did all that plus adjusted colors a bit, Whites, Saturation, and Vibrance to turn the more subdued Adobe profile into something a bit more punchy that comes closer to the Sony. Oh, and by the way, I do upsize the M10 JPEGs to 42.5 MPx in order to level the playing field also as far as noise is concerned. This works against the M10's possible noise advantage. Raw files can be provided .

 

Less compressed JPEGs here: https://www.smugmug.com/gallery/n-Jfdr66/

 

α7R III, from ARW with AWB by the camera and opened in and exported from LR, not touched. 

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!

ISO 4000 f/2.8 @1/60 sec.

 

 

M10, from DNG with AWB by the camera and opened in and exported from LR, not touched. 

ISO 3200 f/2.8 @1/60 sec.

Edited January 1, 2018 by Chaemono