DXO Mark Leica M10 Score

[Mark II]

Advertisement (gone after registration)

I think that this is slightly missing the point. Pretty much any digital camera for the last 10 years has been “good enough” for many purposes. However, cameras with greater dynamic range, higher/lower ISO options and higher resolution increase the shooting envelope - the number of scenarios in which the camera can be used without compromise.

 

As cameras approach the fundamental limits of physics (modern sensors are astonishingly efficient), the scope for such gains gets less and the benefits are increasingly edge cases that might be better served by different tools (eg medium format for landscapes, or small-sensor cameras for street photography).

 

However, given that Leica brands and prices itself as the best-of-the-best, it is obviously incongrous that this is not true of their sensor design. Clearly better technology exists and is widely deployed.

 

Personally, I do not care much about high ISO - but I do care a lot about low ISO for noise and subject isolation. It is therefore very dissapointing that ISO 100 in the M10 is not a true native ISO, but a partial software pull.

[FlashGordonPhotography]

I knew that going in.

 

Gordon

[Advertisement]

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

[rodf]

Having looked at DXO reports before I am convinced that the higher the Megapixels on the sensor the higher the score. It is purely a quantiative test  and doesn't consider the qualitative aspects of an image. Since I have had my M10 I am convinced that I can get better quality images from it than I could on my Canon 5Ds, sharper, better colour, less distortion etc. I have learnt to take DXO with a grain of salt

[M11 for me]

Having looked at DXO reports before I am convinced that the higher the Megapixels on the sensor the higher the score. It is purely a quantiative test  and doesn't consider the qualitative aspects of an image. Since I have had my M10 I am convinced that I can get better quality images from it than I could on my Canon 5Ds, sharper, better colour, less distortion etc. I have learnt to take DXO with a grain of salt

In #45 I wrote a bit about what is important about DxO. I can not really agree wizh you as you claim implicitely that DxO is marketing rather than science.

 

I can though support your statement about the comparison with your Canon hight pixel camera. By my own experience it is very often (not in all situations) the case that the M10 pictures look better and are sharper. I think that the Leica lenses you use make the difference. And a last thing: The Canon 5Ds sensor is not the latest. You should compare to the MkIV.

[pgk]

In #45 I wrote a bit about what is important about DxO. I can not really agree wizh you as you claim implicitely that DxO is marketing rather than science.

 

Firstly, compressing everything to a single number is marketing not science. So why do it?

 

Secondly, suggesting that full frame and cropped sensors are comparable is patently pointless and certainly not science either.

 

Sorry but this is technical information made into hype, like it or not.

[M11 for me]

Firstly, compressing everything to a single number is marketing not science. So why do it?

 

Secondly, suggesting that full frame and cropped sensors are comparable is patently pointless and certainly not science either.

 

Sorry but this is technical information made into hype, like it or not.

Pls inform yourself about the relevance of DxO. You will have to change your opinion.

[pgk]

Advertisement (gone after registration)

Pls inform yourself about the relevance of DxO.

 

Now there's the real question which needs answering . I have a background in photographic science and there is relevance in tests only when they are pertinent and useful to the application in question. As purely technical exercises which do not (because they cannot) take the end usage into account, they have little if any relevance to the real world, but are a way of determining which is current 'top of the tree' camera, so are only relevant to marketing if anything. I can actually see high levels of relevance for appropriate testing of equipment for specific application optimisation but certainly not for general photography. So no, highly unlikely I'll change my opinion I'm afraid.

 

I'll just add that it does amuse me that we utilise extremes of sophisticated testing which we then use to tell us how good a camera is for an unknown application . We should really be testing to find out if a system is 'fit for purpose' but this would be unpopular because most probably are.

Edited December 20, 2017 by pgk

[jdlaing]

DXO is about as relevant to the actual use of a Leica camera as black or silver.

[adan]

DXO is about as relevant to the actual use of a Leica camera as black or silver.

 

That's bad news - the "black or silver" thread is up to 260+ posts

Edited December 21, 2017 by adan

[M11 for me]

Now there's the real question which needs answering . I have a background in photographic science and there is relevance in tests only when they are pertinent and useful to the application in question. As purely technical exercises which do not (because they cannot) take the end usage into account, they have little if any relevance to the real world, but are a way of determining which is current 'top of the tree' camera, so are only relevant to marketing if anything. I can actually see high levels of relevance for appropriate testing of equipment for specific application optimisation but certainly not for general photography. So no, highly unlikely I'll change my opinion I'm afraid.

 

I'll just add that it does amuse me that we utilise extremes of sophisticated testing which we then use to tell us how good a camera is for an unknown application . We should really be testing to find out if a system is 'fit for purpose' but this would be unpopular because most probably are.

We have a problem understanding DxO. The data is absolutely usefull. First of all it is just a data base. You can compare DR and SNR18 of many many sensors. Thats great TO ME. It is another question wheather or not it is relevant to your photography if your sensor renders 15 or 13 EVs dynamic range or whatever. I certainly do not mind thant you do not mind this. Everybody can like what he/she wants to like. Still facts remain facts. A database is nothing more than a data base. You could say something else: You could claim that the measured data is wrong. You could claim that the sensor of the M10 has 15 EVs. This would be a very nice discussion. But what I believe is not very good: Just to say that DxO is stupid and funny because they show bad results for our camera. Sorry: It is a fact that the DR is lower than the best in the market. So is the SNR18. A fact as well.

 

Imagine if the reults for our sensor would have been the top of the cream . . . And please: In the year 2005 good photographers made wonderful images with the first Canon 5D. That camera sold like fresh bread. Today that sensor is fully outdated compared to the MkIV. Still one can go to buy a second hand MkI. Of course. Everyone according to his/her own preferences. I like technology. It makes me a bit mad when the best camera has a mediocre sensor.

 

If the M10 would be the latest model of Mercedes consuming petrol like a car 10 years ago and the air condition blowing into your face . . . Some would not mind. I would change to Jaguar.

Edited December 21, 2017 by Alex U.

[pgk]

But what I believe is not very good: Just to say that DxO is stupid and funny because they show bad results for our camera. Sorry: It is a fact that the DR is lower than the best in the market. So is the SNR18. A fact as well.

 

I did not say stupid. I referred to its irrelevance and the fact that this is now marketing. If someone can show me an image which is inadequate because of marginal SNRs or a miniscule variance in DR in undefined general photography as practised by most then I will be impressed. As I have said before in other threads we are now discussing minutiae whether its to do with lenses, sensors, whatever. Of far more importance is the camera and systems, its ergonomics, carriability and last but by far not least, the way the photographer works with it. I'm sure that Leica will put new sensors in their cameras eventually but the current models are superbly good and arguing about nuances because of some small differences is about marketing not reality.

 

Actually the same is true about cars, the vast majority of which outperform their user's requirements, given that they are mostly used for travel and not the racetrack where nuances become important.

[M11 for me]

You say that having a better or a worse sensor does not matter to you. And you explain why you think this. That is fully acceptable to me. Thank you very much for your openess.

 

And you are right in an other point. If the sensor was the only point that would have mattered when I bout my M10 I would have waited for test results as DxO and others before making my decision. And indeed I am mire than happy with the result I can achieve with it.

 

We see that we are not too far from each other.

[pgk]

You say that having a better or a worse sensor does not matter to you. And you explain why you think this. That is fully acceptable to me. Thank you very much for your openess.

 

I'm still running M9s! But I do see prints from a lot of cameras and often these are large prints too. My experience is that cameras have been fit for purpose for general photography for a good few years. The marginal differences now only really matter if you have a specialist application and the rest is largely hype - in that it often takes careful examination to assess. So to me we are, and have been for some time, at a point where 'absolute specification' has little effect on most images. We should be concentrating on systems that we feel work well for us and that we enjoy using, stop obsessing about specification (though there's no problem in appreciating nuances, they have to be understood to be that) and concentrate of the image content, its composition and lighting - the three key elements of a photograph. Personally I enjoy using my Leicas and no longer worry about their technical specs as long as they produce the images I want.

[Pjay]

At least the conversation here is civil. The discussion after the DXO report on the M10 is full of obnoxious personal attacks, really in need of some serious moderation!

[jaapv]

We have a problem understanding DxO. The data is absolutely usefull. First of all it is just a data base. You can compare DR and SNR18 of many many sensors. Thats great TO ME. It is another question wheather or not it is relevant to your photography if your sensor renders 15 or 13 EVs dynamic range or whatever. I certainly do not mind thant you do not mind this. Everybody can like what he/she wants to like. Still facts remain facts. A database is nothing more than a data base. You could say something else: You could claim that the measured data is wrong. You could claim that the sensor of the M10 has 15 EVs. This would be a very nice discussion. But what I believe is not very good: Just to say that DxO is stupid and funny because they show bad results for our camera. Sorry: It is a fact that the DR is lower than the best in the market. So is the SNR18. A fact as well.

 

Imagine if the reults for our sensor would have been the top of the cream . . . And please: In the year 2005 good photographers made wonderful images with the first Canon 5D. That camera sold like fresh bread. Today that sensor is fully outdated compared to the MkIV. Still one can go to buy a second hand MkI. Of course. Everyone according to his/her own preferences. I like technology. It makes me a bit mad when the best camera has a mediocre sensor.

 

If the M10 would be the latest model of Mercedes consuming petrol like a car 10 years ago and the air condition blowing into your face . . . Some would not mind. I would change to Jaguar.

Well, the main problem I have with DXO is that they measure camera output, pre-processed and all. In the end that distorts the sensor data. Yes- I realize that it would be quite difficult to measure pure sensor output.

[adan]

I find the DxO measurements to be mildly interesting, in explaining some "observed phenomena" of the M10.

 

I don't think a higher score from DxO translates to "a better sensor," because DxO doesn't measure two very important performance factors - when high-ISO banding starts to occur, and color accuracy.

 

A sensor that produces a better or equal S/N ratio across the board, but starts showing shadow banding at ISO 2500 (M240, using DxO's "corrected" ISO values), is substandard to a sensor that has a lower or equal S/N ratio, but shows no banding until ISO 6400 (M10, "12800" corrected by a stop).

 

A sensor that is so "red" that dark greens render as brown or gray (my experience with the M240) is substandard to a sensor that renders those greens as green (M10).

 

Although as Jaap says, those may be a function of pre-processing, not raw sensor capability. In which case DxO needs to make it clear they are not actually measuring "sensors," but "camera output."

 

An analogy. If company "A" makes a drug that is very effective in treating headaches, and company "B" makes a drug that is less effective, but does treat cancer with some success - and I have cancer, then obviously the "less effective" drug is still the "best" drug.

 

I shoot manual-focus rangefinders. I'm very interested in which sensor in a rangefinder camera is the most effective. Which seems to be the M10 sensor (if sometimes only by a narrow margin), in both practical shooting, and in DxO's limited measurements.

 

I would not change to an SLR or EVF-only camera even if their sensors were five times "better." They treat the wrong illness.

 

Finally - Leica has not taken the lead in technological development since about, oh, 1959, when the Nikon F system came out. (Some might argue 1932, when the technologically more-advanced if sometimes less reliable Contax came out). Except, of course, in certain areas of optics. If that makes one "mad" - one has presumably had an anger issue for a long, long time.

[robgo2]

I mean that Leica enthousiast appreciate their gear, whether from the latest generation or not.

Nikon/Canon people tend to think they can only make nice pics with the latest stuff.

Giving you the benefit of the doubt, I think that you must realize that this is an unfair generalization.

 

Rob

[robgo2]

It is correct. Of course you can compare any camera to any camera but when you open the page you are presented with a comparison to two high megapixel cameras. Why not two current 20-28MP cameras. Would we be saying the sensor is poor if the cameras DXO chose when you open the page were the Nikon D5 or Canon 1DX2? Probably not because those sensors are actually close to the M10. Yet somehow the M10 sensor is shit and they are just fine? I don't think so. "Ah, but those cameras aren't designed to perform like an A7R3", you say. Well neither is the M10. It's a specialist tool for M lenses. Not just the newest ones but also the last 50 years of M glass. It isn't designed to operate like the A7R3 either.

 

It's simple marketing to choose the A7R3 as a direct competitor to the M10 on that web page. Same as when DPR put an SL body in the hands of a young woman to make the body look bigger than it actually is. Putting the Sony's score up creates a perception about the Leica. Put the Nikon's D5 score next to the M10 and the perception is entirely different. Now we see a sensor that's competitive with Nikon's flagship. As does just testing the sensor. They say that the Sony sensor is better. Not with a 50mm Summilux on the front of it, it isn't. No mention of the specialist microlenses or that unlike most Canikons it has no AA filter, etc.

 

DXO needs people on their website so they can sell their software. The cameras they chose to headline against the M10 were for dramatic effect and all of the major camera websites just blindly used exactly the same comparison cameras. Sure you can change that. But no one did. The truth but not the whole truth, as usual.

 

Gordon

Perhaps DxO chose the A7Riii as a comparator, because it and the M10 are the most relevant cameras for those who are trying to decide on which high end smallish mirrorless camera to purchase. Nevertheless, your point about the Leica being a better camera for using M lenses is indisputable. And for those who don’t want or need autofocus, it is also much more pleasant to use. As for sensor differences, there is no doubt that Sony is at the top of the heap, but it only matters at the margins of picture taking, which means that the A7Riii will produce more usable files under extreme shooting conditions. This is a trade off that buyers must understand when making purchasing decisions. Which camera would they prefer for the kind of usage they will give it. Finally, super-high sensor resolution is largely a marketing gimmick, IMO, except for those who print very large or crop severely. How many of us actually do that?

 

Rob

Edited December 22, 2017 by robgo2

[paulsydaus]

Does DXO test the ability for the sensor/processing to distinguish almost identical tones?

 

 

Sent from my iPhone using Tapatalk

[Advertisement]

Hallo,

Look here: Leica M

[Chaemono]

I’m not sure. They measure color sensitivity as ISO increases. The Sony α7R II/III sensor is impressive in terms of DR and the way it captures the tones, at least at lower ISO values, even though I don’t find them quite accurate at times. Also, it has very little noise at higher ISO values. We all know this. But I think Adam has a point when he talks about color accuracy. I think the Sony suffers in this respect as ISO increases even while noise stays low. Anyway, I just got a hold of a Sony α7R III and I’ll compare it to the M10. I’m very impressed with the way the M10 maintains color accuracy as ISO increases. Noise is about the same as the Sony. Hopefully, I will have some pictures to post in a couple of weeks .