DXO Review

[ohnri]

Advertisement (gone after registration)

Ever since I had my M9 adjusted for proper focus (the sensor needed alignment) it has been my most used and favorite camera. I am primarily using the f/1 Noct and getting terrific results.

 

Although the DXO Mark results are fun to look at, anyone wanting to use a rangefinder camera will be exceedingly happy with the prints a well adjusted M9 produces.

 

Pixel peepers may find cause for disappointment but serious photographers will continue to produce beautiful images with the lovely M9 sensor and M-mount lenses for years to come.

 

Having said that, the other night I needed to shoot a party in very low light and I did choose to use my backup camera, a Nikon D3x which I equipped with my 85 f/1.4. Most of the shots were wide open at ISO 3200 to 6400. Horses for courses.

 

Best,

 

Bill

[yanidel]

Pixel peepers may find cause for disappointment but serious photographers will continue to produce beautiful images with the lovely M9 sensor and M-mount lenses for years to come.

I think that sums it up pretty well.

[Jerry_R]

Story here is if you want/need a Canon, get a Canon. If you want/need a Leica, get a Leica. Nobody's forcing anybody to buy/use something they shouldn't. End of story.

Wow! Will there be a day when I will be surprised postively? Almost any day, any thread, any criticism of Lecia and... ALWAYS someone is advising to sell it...

Come one, is it GOD figure to praise to?

 

I used to have 5D II which is ages ahead in many aspects of M9. I sold it.

I am more happy with M9. But it doesn't change the fact, that M9 is behing most of dSLRs in aspects we can compare.

 

* * *

Another person asked if we can see its test result in our pictures. This seems to be misunderstanding. Of course I see noise in my M9 raws compareable to other cameras tested there.

DxO reviews SENSORS. What we see is mostly results of glass.

 

A comparison of the DxO charts for the Leica M9 and M8 and the Canon EOS 5D and 5D Mark II shows that these four cameras perform similary with regard to noise (as seen in the SNR graph).

Not at all.

First - RAW from M9 is not RAW, it is smeared RAW. But OK, let it be.

M9 dot for manufacturer ISO 1280 = 5D II dot for manufacturer ISO 1600.

And 5D II is better on 1600, than M9 on 1200.

 

I am not technically savy, but what I love about the IQ of the M9 more than anything is the quality of the colors it produces relative to other high end caremas I have used. I find that they are more natural and very difficult to match in pp from raw files of other cameras.

No ;-) Not.

Strange colors from M9 is first aspect I observed. You can only observe it when compare the same sample with other cameras. One important comment - I am talking about RAWs opened in LR. I did compare results with u43 E-P1 and GH1. Both u43 cameras, especially E-P1 - produce more natural results. Results that remind reality.

 

M9 colors (LR RAW) are shifted. Even when I set white ballance carefully and manually. They may be more pleasing to viewer, no doubts. Like more contrast photo is more pleasing to many people too, but it doesn't mean it is natural.

 

And today I saw I am not the only one. Ken Rockwell and Erwin Putts also see it.

Ken Rockwell's Updates

(go to 20 March 2010) (yellow tinge to the people)

 

M9, part2

(have a look at M9 vs Nikon)

 

PS: It is worth mentioning, that when you open RAWs in LR - and want to see natural results - you must change tone curve from medium to linear.

Then you see picture more corresponding to reality. With any other camera doing this will give you dull results. Not with M9. But OK, it is probably Adobe approach to M contrast lenses.

As Steve Huff observed, just after opening RAWs - Aperture gives nicer results (without any playing in developer).

[pascal_meheut]

Strange colors from M9 is first aspect I observed. You can only observe it when compare the same sample with other cameras. One important comment - I am talking about RAWs opened in LR. I did compare results with u43 E-P1 and GH1. Both u43 cameras, especially E-P1 - produce more natural results. Results that remind reality.

 

M9 colors (LR RAW) are shifted. Even when I set white ballance carefully and manually. They may be more pleasing to viewer, no doubts. Like more contrast photo is more pleasing to many people too, but it doesn't mean it is natural.

 

And today I saw I am not the only one. Ken Rockwell and Erwin Putts also see it.

Ken Rockwell's Updates

(go to 20 March 2010) (yellow tinge to the people)

 

M9, part2

(have a look at M9 vs Nikon)

 

Tried another RAW converter than LR? Because since the DMR days, Adobe product have given strange results when it comes to colors and Capture One has been much better.

[ho_co]

... M9 colors (LR RAW) are shifted. ...

Which camera preset are you using for the M9 when you say that?

[DFV]

I am not really that concerned to be honest with you. When you compare the M9 with the Canon 5D mkII at ISO 160 the results are very close. I rarely use anything else than ISO 160 and if I go above it is because I shoot B&W and look for a grain effect on purpose.

 

I was a bit surprised at the result of the Canon 5DmkII and the Nikon D700 which is clearly superior according to them at any ISO. When I "pixel peep" the results of my D700 and my M9 I really can't claim that the D700 is superior. In fact, to me I would take the M9 over the D700. This could be because of the CCD sensor compared to the CMOS. Even if grainier at high ISO the pictures seem to look closer to real film to me where the CMOS sensor's results still look digital to me. I thank Leica to favor the final image and use CCD (for now) rather than going CMOS because the industry says so (or is actually more comfortable using it). Statistic and data are a tricky business that the Japanese play very well. They do this in the car industry all the time. They copy the specs of European cars on paper and sell it to you at half or a quarter of the price. Obviously in the end you get what you paid for but on paper they fooled you. Nothing against Japanese cars (owned several and am most likely going for another soon) I am actually glad that they are giving the Europeans a good load of competition but one must be aware that regardless if Lexus or Infinity might copy or even improve Mercedes, BMW or AUDI on paper they will never feel like them when behind the wheel.

 

I do not know if it is the lenses or the sensors but I have not had results with the D300, D700 or 5DmkII that where nearly as good as the M8.2 or M9. In the end this is the test that really counts.

[Driver8]

Advertisement (gone after registration)

Not so nice reading. What do You think?

 

DxOMark review for the Leica M9

I can't say that I'm surprised. In terms of noise, the Canon 5D and Nikon D3X blow the M9 away for ISOs above 800. But I still prefer the look of the M9's images at ISO 160 and 320 compared to those two cameras.

 

By the way, is it just me or has anyone noticed that the contrast and colors of M9 images look similar to old Kodachrome 25? I especially see this in how it renders shadows and skin tones.

[k-hawinkler]

FWIW, I've been very happy with my D700, provided I use my Leica glass...

 

... on the D700? I may have misunderstood completely, but let me ask anyway: What Leica glass, what adapter do you use on the D700? Which M lenses would let you focus to infinity on the Nikon?

 

So far, I have used the 135mm f/4 Tele-Elmar lens head with 16464 universal focus mount and LEICA m-AI adapter on my D3, as well as the Telyt 400mm f/5.6. The Tele-Elmar does a spectacular good job, I think. Both configurations focus on infinity.

 

Thanks for a reply, K-H.

[andybarton]

I have no idea whether one can mount M lenses onto a Nikon or how they would work, but don't forget that there is/was a whole other wonderful world of Leica lenses out there.

 

I use my Leica R lenses, orphaned when the "R10" was canned. I use Leitax mounts.

 

There are lots of my shots taken with this combination on here. To name but three...

 

http://www.l-camera-forum.com/leica-forum/landscape-travel/119846-tuscan-hilltop-farm.html

 

http://www.l-camera-forum.com/leica-forum/landscape-travel/117550-world-heritage-site.html

 

http://www.l-camera-forum.com/leica-forum/landscape-travel/115427-hadrians-wall.html

[snaggs]

Not so nice reading. What do You think?

 

DxOMark review for the Leica M9

 

I think this is stupid. Measuring the M9 based on its high-ISO performance is not telling us anything we don't know. I had a 5D (and remember when people were amazed how low the noise is).

 

The files out of the M9 are out of this world. I knew it would be noiser than a D3, but it has the CCD look that CMOS sensor's don't have.

 

I also remember when Nikon got criticised for noise, whilst Canon was producing images that made everyone look like plastic dolls. Does its the high-ISO king.

 

Would I like another stop out of the M9? Yes, but ISO 1600 is enough to work with, and given the fast lenses and lack os SLR mirror, I'm not sure I'll never need anything more than ISO 3200.

 

Daniel.

 

PS. DxO are not impartial either, if everyone made their lenses like Leica, not needing correcting, they'd be out of business!

[k-hawinkler]

I have no idea whether one can mount M lenses onto a Nikon or how they would work, but don't forget that there is/was a whole other wonderful world of Leica lenses out there.

 

I use my Leica R lenses, orphaned when the "R10" was canned. I use Leitax mounts.

 

There are lots of my shots taken with this combination on here. To name but three...

 

http://www.l-camera-forum.com/leica-forum/landscape-travel/119846-tuscan-hilltop-farm.html

 

http://www.l-camera-forum.com/leica-forum/landscape-travel/117550-world-heritage-site.html

 

http://www.l-camera-forum.com/leica-forum/landscape-travel/115427-hadrians-wall.html

 

Andy,

 

Thank you. I knew the R lenses could be adapted to a Nikon. The M lenses just work in macro mode. I just confirmed that one more time. Except the lens heads of some of the 65, 90, and 135 mm M lenses can focus to infinity on a Nikon. Also the Telyt 200 mm f/4 with M39 adapter should work on a Nikon too.

 

WOW. Very impressive shots. Thank you for the pointers.

 

K-H.

[mjh]

First - RAW from M9 is not RAW, it is smeared RAW.

Huh? Could you elaborate?

 

M9 dot for manufacturer ISO 1280 = 5D II dot for manufacturer ISO 1600.

I suppose you did notice that the 5D Mark II’s ISO 1600 is actually less than the M9’s ISO 1280? The DxO graphs illustrate that clearly. Yes, SNR is lightly lower from ISO 640 onwards, but the differences are small. Just from looking at those graphs, one wouldn’t see a reason to worry about noise.

[jaapv]

No ;-) Not.

Strange colors from M9 is first aspect I observed. You can only observe it when compare the same sample with other cameras. One important comment - I am talking about RAWs opened in LR. I did compare results with u43 E-P1 and GH1. Both u43 cameras, especially E-P1 - produce more natural results. Results that remind reality.

 

M9 colors (LR RAW) are shifted. Even when I set white ballance carefully and manually. They may be more pleasing to viewer, no doubts. Like more contrast photo is more pleasing to many people too, but it doesn't mean it is natural.

This is nonsense. The colors of each and every camera when converting RAW are interpreted by the profiles of the RAW converter. If you see a shift (did you measure a Greta-Macbeth card - your post suggests you used the most inaccurate instrument of all- your eyes-.) it just means you were using the wrong profile, or the RAW software did not provide the right profile (yet).

[pgk]

So who is going to explain to me what 68.6 actually means in practice? Or is it just another number which has technical meaning but no correlation with the real world:D.

 

But seriously, for any quantification to be meaningful it MUST translate into something actually visible otherwise, sadly, it is simply meaningless and of no use to anyone other than the technocrats who like numbers, isn't it?

[jaapv]

It is on the scale of the PPI and indicates the amount of white noise the review is going to generate in Internet forums. (PPI=PixelPeepers Index)

[pgk]

Jaapv. Thank you. The most lucid explanation I have as yet come across:D.

[Hacker]

Nikon D700 beats them all on noise - which is why I'd like to see an M9H with 12 Mpixels on full-frame.

 

And to think the Nikon D3s beats the D700 on noise.

[zlatkob]

The DxO numbers are meaningful, not just technical. If the M9 produced higher DxO numbers, we could SEE the practical effects as lower noise in high ISO photos and as greater dynamic range. And if that were to happen, everyone would congratulate Leica for achieving such great DxO results.

 

At the same time, DxO can only assign a numerical result for what they measure, and they don't measure everything about a camera. Cameras like the Nikon D3X achieve superlative DxO numbers, but DxO doesn't measure the discomfort of carrying a D3X for ten hours (if that's what your photography requires). The problem with some cameras is that you have to carry them. I think that that a realistic assessment of DxO numbers is that they appear to be accurate for the specific things they measure, but they don't measure everything.

[glenerrolrd]

I have found the DxO reports to be very helpful in understanding sensor performance and limitations. I have M8/M9/D700/D3x so I guess I can compare real examples to the tests and in every case they reasonably explain the sensor performance .

 

This is information you can use ...just look at the D3X verse the D700 . On the surface (if you just read the summary chart) the D3x seems to completely out distance the D700. At the detail level you can see that D3X big advantage in DR range appears at the base ISO and after that its similar to the D700/D3 . You can also see that the D3X overall performance starts to die at ISO800.....and in fact the D700 /D3 is better by 1 1/2EV . The D3 compares to the D3x at 1600+ its better than the D3X at 800.

 

In general all the sensors tend to fall off in a linear pattern until they reach "their practical limit" then performance falls off (for me too much).

 

Personally I define "practical working ISO" as where I want to be most of the time . For the M s I want ISO 160 until I can t get the necessary shutter speed or depth of field. If you always use 320 or 640 you are giving up some performance. (maybe you don t need it or have such low light to make 160 less practical).

 

As many have reported ..yes you can get good results at 1000 ..if you have even light (doesn t require maximum DR) and you nail the exposure to less than 1/2 EV .

 

The limitations of the DxO test results are really in two areas(1) sensors of different sizes don t seem to give credit for size ..so comparing MF to FF to APSC doesn t represent the impact of sensor size (2) the rest of the imaging system isn t factored in ..so lack of AA filter and great glass isn t considered.

 

I am sure I missed a few other weaknesses but their are some interesting insights in the details . The apparent presence of in camera noise reduction on the M9 above ISO 160 was interesting.

 

I guess you can complain about the presumption of "best" but some interesting insights can be found in the details.

[Advertisement]

Hallo,

Look here: Leica M

[Marty]

If the M9 produced higher DxO numbers, we could SEE the practical effects as lower noise in high ISO photos and as greater dynamic range.

This may be true in some cases e.g. noise but can you really see it all the time, with your eyes? Or only in lab conditions? And how much credit is due to Leica for the sensor itself?

sensors of different sizes don t seem to give credit for size ..so comparing MF to FF to APSC doesn t represent the impact of sensor size

Well according to DxO, comparing M8 and M9: "This DxOMark score improvement can be explained by the increase in sensor surface size alone"