DxOMark scores for 5DMkIII out - total score 81, 5DMkII had 79

[kozakm]

To be honest, I’m not that suprised. The biggest disappointment is banding and color blotches in the shadows even in ISO 100 and very small improvement compared to 5D II.

[bernardovaghi]

Maybe for huge dinamyc of the D800, and very good ISO, and bit depth evolution over his predecessor, D700.
Really the MK III it´s not so far from the MK II in those qualities.

But for me, it´s a superb piece o technology.

; )

I really like Canon stuff. I´m a Nikon, Canon user. The two brands have their advantages. But here in this topic, what i´m seeing is an attempt to "fire down" the DXO Lab. So i brought a link for you to know how scientific the tests are done there.

http://news.cnet.com/8301-11386_3-57415777-76/how-dxo-labs-tests-hot-cameras-like-canons-latest-slr/?part=rss&subj=news&tag=title

Dude, for studio photos, landscape, advertising, portraits, food, the D800 blows the MK III out of the waterm it´s a matter o fact. For weddings, video, sports, journalism, wildlife the 5D MK III blows the D800. Like the D700 and the MK II , but this time switching "sides".

Cheers!

Lots of people test the way DXO does, I know a few guys that have remarkable testing facilities in their basements.

Your second paragraph is the issue, the D800 and 5D3 have completely different strengths and weaknesses as far as IQ goes, so how is their such a big gap in the "total score"?

My issue is not the results, my issue is how they weight results to get their magical number.

[Wrathwilde]

And your arguments regarding the DX0 Mark tests, bias etc. If the shoe was on the other foot, you lot would be putting the boot into Nikon and praising DX0 Mark for their testing techniques.

  Exactly, I don't believe the results are biased at all... the data isn't collected for the purpose of giving sensor reviews, that's just a byproduct. The data is collected to form a basis of the camera sensor performance, so that DxO's software can try to exact the very best image quality from the known limits of the sensor. The RAW files the Nikons produce exhibit much less sensor noise, it has been said that Nikon cooks it RAW files to compensate for sensor noise, where as, Canon does not. This could be why the RAW tests seem to favor Nikon so heavily in DR at low ISOs.

That said, even my conservative estimates for the 5D3's DxoMark score were way off.

This is what I'm expecting, for the 5D3 vs D800

Overall Score - 84 ( Actual Score - 81)

Color Depth - 24 bits  (Actual Score - 24 bits) one of three, I'm on fire!!!

Dynamic Range - 12.4  Evs  (Actual Score - 11.7)

Low-Light ISO - 3200  (Actual Score - 2293)

I am actually disappointed that Canon hadn't made more strides in sensor performance in 4 years, don't get me wrong, it's still a great camera, but I was expecting four years of sensor development to yield more, and I was not expecting to lose ground in DR as compared to the 5DMk2.

[bp]

The DR result is no surprise.  The "sports" results just make me guffaw.  D800 better than the 5D3 eh?  OK

Hand these two cameras to any actual real-world sports shooter (or a wedding shooter in a typical reception), and see which one they set down on the ground.  This just shows how useless their test results are

[altenae]

Well even after the DXO numbers my 5D mark III is still just as good as before.

This DXO numbers issue will continue for some time on all forums.
I will be back when thinks are cooled down.

Just for the record:

All Nikon counterparts lenses are rated better then Canon ones.
I know for sure this is not true for al lot of lenses,because I have used them both.

I can assure all photographers that it's a bad idea to buy a camera and let the DXO numbers decide for you.

See you all later after the DXO storm

[atvinyard]

I think people are forgetting that this is a sensor score, not a score for the whole camera.  The D800 has a better sensor, but does it have better features, controls, autofocus capabilities, frame rates?  Are the lenses available for it better?

[psolberg]

So DXO is the forbidden term now? Should we spell it D..O?
First K..n Ro..ell and now D..O.

haha, no you can still talk about them. I just want people to know I don't support this type of marketing by DXO.

I realize some may disagree with me on this topic, just trying to "keep it real".

+1 this is almost too obvious that they are bias.  Come on saying the d800 is better rat low light then the mkiii!  I bet they never too a picture with either camera!

They normalize by resizing files to the same size. This is similar to what printing does. Thus the per pixel advantage of 100% crops does not help the lower resolution body. If you do a lot of 100% crops, then no doubt the 5D3 will be better. However I'm not sure 100% crops are what most people sell.

[neuroanatomist]

DXO is run by nikon fanboys take this with a very very very small pinch of salt!

Ahhh...and you have proof to support that serious allegation?

Honestly, everyone should take a step back and consider what the DxOMark scores mean - and don't mean.  They are quite transparent about their scoring metrics and methods, and there's no reason to suspect there's collusion or favoritism occurring.  At the same time, it's important to remember that the sensor score is just that - a score of the sensor itself, not a 'camera score'. 

Furthermore, keep in mind that all of their scores are based on reducing the image to 8 MP - thus, the greater the starting resolution the more downsampling, which means lower apparent noise.  That lower apparent noise means apparently better ISO performance and apparently more dynamic range (lower apparent noise lowers the 'floor' for the DR estimate).  If you like, you can substitute 'artifically inflated' for 'apparently'. 

They have a page describing the mathematics of the normalization to 8 MP, and on that page, there's the following statement:

What should be remembered is that doubling the resolution adds:
3dB to the normalized SNR
0.5 bit to the normalized DR
0.5 bit to the normalized TR
1.5 bit to the normalized CS.

So, compared to 8 MP the D800 is 4.5-fold higher (just over two doublings), whereas the 5DIII is 2.75-fold higher (just over one doubling).  What that means, mathematically, is that the D800 has approximately one extra doubling of resolution relative to DxOMark's 8 MP normalized value - that accounts for all of the differences in the scores for both ISO and Color Depth, and part of the difference in dynamic range. 

Note that DxOMark does provide the non-normalized data, they just don't use those data to calculate the overall scores, the rationale being that normalizing to 8 MP allows appropriate comparisons.  In one sense, it does - if you're going to print 8x10" images all the time, then their scores actually apply pretty well.

Let me give a specific example for color sensitivity, which is the basis for DxOMark's Portrait Score.  That's one area where the D800 with 25.3 bits 'beats' the 5DIII with 24-bits.  Below is the comparison without normalizing to 8 MP, where you can see that the D800's advantage pretty much entirely disappears. 

[tomscott]

Well if Canon did make such a mess of it go buy the Nikon whos stopping you?

Its amazing how much fire there is here! The D800 isnt good enough to trade in thousands of £ of lenses for, when the 5D MKII came out we were all amazed with 22mp! Not only does it provide fantastic IQ it provides a good file size. Its more than 90% of photographers need. How many who would use a DSLR print bigger than A1? If you do you need to go medium format for crisp images.

IMO Canon made the best all round camera you can buy! The differences we are talking are so minuscule, unless you need the resolution then the Canon is much better IMO. All this crop rubbish, if you need to crop that far with an image you need to go back to basics.

There are going to be bad reviews, and poor testing just calm down.

OMG 81% THIS CHANGES EVERYTHING!!!!!!!!!!!!!!!!!!!

[squarebox]

Let me give a specific example for color sensitivity, which is the basis for DxOMark's Portrait Score.  That's one area where the D800 with 25.3 bits 'beats' the 5DIII with 24-bits.  Below is the comparison without normalizing to 8 MP, where you can see that the D800's advantage pretty much entirely disappears.

As always thanks for the level headed explanation.  I was surprised that the canon scored so low, but it is good to know that these cameras aren't so night and day apart as far as the sensor goes. 

[Chewy734]

Thanks for the explanation neuro!

[EYEONE]

The SNR test is also very reveling once you move from "print" to "screen". The 5D3 jumps out ahead.

[steven63]

How many people here that are bashing DxO for the results would NOT be bashing their results had they come in higher than the D800?

None of this 'it's how they weight their tests' or 'different results from different software users' would even be mentioned on this site if the scores were more respectable.

This is classic 'kill the messenger' tactics. 

[Mike Miami]

I think people are forgetting that this is a sensor score, not a score for the whole camera.  The D800 has a better sensor, but does it have better features, controls, autofocus capabilities, frame rates?  Are the lenses available for it better?

^^^ Well said!! ^^^

[traveller]

It will be interesting to see just how the 1D X sensor fares when it is tested.  I've got a sneaking suspicion that Canon really haven't done an awful lot to the 21MP sensor design to produce the 22MP version in the 5D MkIII.  If true, one would wonder whether this is a deliberate ploy to protect the 1D X or simply concentrating resources on other projects.