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

[LetTheRightLensIn]

Re: DxO results out for 5D3

In summary 5D3 vs D4:

They have the same SNR (D4 much better at 100% view but that is not fair, normalized they are the same).

The D4 has very much better low ISO dynamic range (and they don't use Exmor, proving that Canon has no patent excuse here) and noticeably better hugh ISO DR. D4 simply has way better DR than the 5D3 across the board.

The D4 has much better color sensitivy.

The D4 has much better metamerism index.

The D4 has noticeably less MP.

 

[LetTheRightLensIn]

Re: DxO results out for 5D3

In summary overall:

Basically the D800,D4,D3s,5D3 are all tied for the best SNR ever (other than the D3s has higher base so it doesn't top out as well as the others at lowest ISO).

The D4 has the best non-Exmor DR ever by a large margin.

The D4 and D3s have the best high ISO DR of any cameras ever by a large margin.

The D800 has the best maximum DR ever, although some other Exmors are close, even some that are only APS-C.

The Canon DSLRs have the worst low ISO DR of all modern cameras.

The D4/D3s have the best color sensitivity ever.

The D800 has the most MP ever for DSLR. And then come the D3x/sony, 5D3/5D2.

The 24MP DX has most reach ever and then come stuff like the 7D.

For all the talk, the D3s really never had the most amazing low ISO though, lacking ISO100 and with a low MP count. At lowest ISO for each, it has DR only modestly better than the 5D3, many less MP, 2/3 stop worse SNR, a trace worse color sensitivity.

Modern cameras are all becoming more and more color-blind as they cheat more and more with color filter arrays, it seems like the 5D3 sinks to a new low for color-blindness, adding some blue-green to the red-green color-blindness (OTOH the R does need less boost than the 5D2 so the red/red-ornages might be a bit better than on the 5D2 which might make it a touch better for say fall foliage or the red-orange type? whereas maybe people don't care about telling fine blue apart as much?). It may be how they managed to tie the top Nikons for best SNR ever while still using somewhat older technology for the sensor, perhaps?? This whole issue is very complex and I'd really need to dig into all the detail much more.


overall (sensor only, not talking about actually shooting with them and fps/AF/user interface/usability , etc. just sensor and nothing else):
For super duper high ISO I'd rank them D4/D3s then 5D3 then D800 then 5D2.

For high ISO I'd rank them D4 then D800/D3s then 5D3 then 5D2.

For medium high ISO I'd rank them D800 then 5D3 then D4 then D3s/5D2.

For lowest ISO I'd rank them D800 then D4 or 5D3 (depends whether you need MP or DR more) then 5D2 (can go above D4 if MP are all that matter) then D3s.

(overall ratings are dangerous, since sometimes DR and sometimes SNR and sometimes color and sometimes MP matter the most)

 

[LetTheRightLensIn]

Re: DxO results out for 5D3

And finally, especially since DxO does not count MP count as something contributing to a score, I absolutely fail to see how the D800 scores considerably higher for low-light sports overall rating on DxO than the 5D3. How? Where did they come up with that??? Honestly, well something doesn't seem honest about that almost. Very odd.

 

[JR]

I was a nikonian in film era equipped with N90S, N 8008S bodies + SB25 speedlight + (were) high end Nikkor lenses. Because of situation, I have left phography since 2003. However, my eyes still have been following the progress from SLR to DSLR. I have a little changed my view from Nikon to Canon since the first time I had a chance physically seeing a Nikon and a Canon DSLR bodies sitting side by side in a camera store years ago.
In my point of view, the Nikon looked cheap because of the orange stripe on left front. And I started loosing my love with Nikon.
Now, my situation has changed and I have opprtunity to come back with my love photogaphy. I was thinking and reading a lot of reviews online to choose among Nikon or Canon or Sony as it was an expensive investment in hobby (I was a Pro but have not been earning in photogaphy since the situation told above). After months of considering, I dropped Sony and paid attention on 2 new comers for my come back: D800 and 5D Mk iii
Finally, last night, I placed oderd 5d Mk iii and 2 L lenses. Why?:
- D800 looks cheap because the orange stripe as my view of aesthetics.
- I do not like the pop-up flash. It does not help much but is annoying, and looks cheap as well. As a Pro, external speedlight was always at my side.
- My composition used to be tight and in purpose. I do not shoot randomly then crop a lot later so I do not need extreme high MP. If 36 MP were good, it would be in D4.
- I used to take pictures in early morning and late afternoon and evening, so low light is a issue to me. 5D Mk iii gives me the convenience. I can rarely use speedlight.
- I am not biased by reviews online. I judged on my own eyes, especially the comparison on www.imaging-resource.com.
I am sad to leave Nikon and happy to join Canon.

well welcome former nikonian!

 

[JR]

That's it, I'm sending my 5DIII back. I really like this camera, it produces amazing images, and the build quality and ergonomics are superb, but obviously these DxO scores are more important than my hand-on impressions of using it in the field.

humm...think you need the Canon AF system to track all your fast moving cars V8 beast. Thank god we all know you are being sarcastic! lol

 

[Mike Ca]

In the D800 36MP Exmor sensor has an additional problem. You cannot read the whole sensor at 30 fps to do HD video. Nikon is using some kind of pixel skipping to reduce the amount of data that has to be read off the sensor. This means the D800 is not using the whole sensor area, even the whole area within the HD image, to generate its video signal.

I hate to break it to you but no camera does, not even the 5DmkIII. as you say, has to do with data size. 36MP is a LOT of data to read in one pass. The D800 actually resolves more detail in video than the 5DmkIII as has been documented by EOSHD and other sites at the expense of moire. It's method also allows for less rolling shutter. It's a tradeoff.

Canon has developed on sensor circuits that allow the analog signals of multiple pixels to be mixed for downs ampling to HD video resolution. Canon uses analog signals from the whole sensor area, at least the part that is in the HD aspect image. This allows the Canon sensors to give better high ISO video.

similar technology is no doubt present in sony video sensors. canon doesn't read the full sensor and downscales because of the amount of computational power required is beyond the abilities of their cameras. instead they pixel bin. pixel binning comes at a trade off in resolution as you can see.
http://www.eoshd.com/content/7631/panasonic-gh2-vs-5d-mark-iii

neither canon nor nikon/sony have perfect video on their dslrs. Neither camera is even remotely close to a 1080p downsample of a full sensor readout. everything we have today is a compromise in one way or another.

Canon does its down sampling by combining analog signals on the sensor chip before the ADC. This allows them to use signals from all the area illuminated by the HD image. The D800 is apparently using pixel skipping to down sample, which uses only part of the sensor area illuminated by the HD image. This may give the D800 slightly better resolution at low ISO video, but at high ISO the 5DIII video is clearly better.

 

[Fishnose]

Finally, last night, I placed oderd 5d Mk iii and 2 L lenses. Why?:
- D800 looks cheap because the orange stripe as my view of aesthetics.
- I do not like the pop-up flash. It does not help much but is annoying, and looks cheap as well. As a Pro, external speedlight was always at my side.
- My composition used to be tight and in purpose. I do not shoot randomly then crop a lot later so I do not need extreme high MP. If 36 MP were good, it would be in D4.
- I used to take pictures in early morning and late afternoon and evening, so low light is a issue to me. 5D Mk iii gives me the convenience. I can rarely use speedlight.
- I am not biased by reviews online. I judged on my own eyes, especially the comparison on www.imaging-resource.com.
I am sad to leave Nikon and happy to join Canon.

Fascinating.
- The red stripe on the Nikon is VERY important to you....
- The popup flash on the D800 is NOT meant for use as a flash per se. It is meant to be used as master for controlling other flashes in the CLS (Creative Lighting System), which beats the hell out of Pocket Wizard and other primitive 'on or off' systems.
- You seem to think high MP is only for cropping, if you have 36MP one must crop, downsizing is impossible. Or maybe one gets sloppy and lazy from having SO many pixels.... what exactly is your point here?
The D4 doesn't have 36MP because it is impossible to have a ultra-fast camera with so much data to be shifted for every frame. Believe me, if it was possible, they would have 36MP in the D4. Maybe the D5 or D6 will get there. Depends on how successful R&D at Sony and Nikon are at resolving issues with shifting all that data in and out of buffers and onto cards.
- D800 handles low light as well as the MkIII. The difference is in fact so small that you have to pixel peep to see a difference.
- No, don't let online tests and reviews influence you. After all, the red stripe is the biggest issue, for sure.... :

 

[revup67]

- The popup flash on the D800 is NOT meant for use as a flash per se. It is meant to be used as master for controlling other flashes in the CLS (Creative Lighting System), which beats the hell out of Pocket Wizard and other primitive 'on or off' systems.

This is one reason why I love the 7D..having this built in feature to be able to control an external master and slave(s) flash (A+B+C) as one example. I rarely use the built in flash of the 7D but use it's wireless feature to control the 580 EXII. I do wish for the $3499 Canon opted to include a built in flash not necessarily for it's built in use, but to avoid buying another device to trigger the external flashes.

 

[JackM]

It doesn't make sense to rate a camera based on base iso performance and 8MP. Who shoots only at base iso and always downsamples to 8MP? What's the point in having 36MP or 22MP then? And the 2 point improvement of the Mk3 sensor compared to the Mk2 is ridiculous, the Mk3 is much better. Sensorscore is 560/781 (M2/M3), sounds more like it.

 

[dichiaras]

I went to read the nikonrumors forum (for the first time, so don't crucify me), and I was expecting a wild party (Cartagena security team style, just to be clear), instead was surprised to see rather moderate comments. Somebody (two, at least that's what I found) was deriding Canon gear owners, others were mentioning that here at canonrumors the DxO conspiracy theorists were getting wild (I partially agree with that), but more in general they were toning down the importance of the ratings. Actually someone was finding flaws in the D800 compared to the 5D. I guess some people are never happy: they just want to own the perfect toy, but they never will.

 

[Kernuak]

What I find rather arbitrary, is the use of a reduction in DR as part of the sports shooting score. The reality is, if you're shooting in low light, the dynamic range is typically very small, so looking to see when the DR drops below 9 EV is totally irrelevant. I'd much rather see R&D concentrate on low noise, than worry about DR at high ISO. At ISO 1600, DR does have some relevance if you have to use fast shutterspeeds (so the light is still relatively high), but beyond ISO 3200, it isn't generally a problem,

 

[wockawocka]

DXO scores are only any good to DXO. Their testing methods are bunk tbh.
They scored the D800 higher than a IQ180, morons.

For those that don't know, that's scoring a 14bit back higher than a 16bit back with twice the photon real estate and sensor size. Hahaha.

 

[LetTheRightLensIn]

Re: DxO results out for 5D3

Anyway it is good to see that the SNR did, in fact, go up a full 2/3 of a stop.

So we do get improved noise and high ISO so it's not only all just AF/fps type stuff that was improved in the 5D3, the sensor is better to a degree that will be noticeable, even if only at upper mid and high iso really.

The SNR was upped by as much as could have been at all hoped for.

 

[dtaylor]

Re: DxO results out for 5D3

OK Jrista (and a few others), despite all the incredible bashing and grief you gave me (really us, since some others on DPR also did the same things and got the same results and as I said then, I was hardly alone in this) and all the being called incompetent trolls without a clue who were using faulty methods and how there was no reason to trust us (and worse), etc. etc. well it turns out that DxO matches virtually exactly my (and a number of others) predictions from weeks and weeks ago

You would have gotten more respect had you said that Bozo the Clown replicated your results while riding in a VW Bug with 20 other clowns.

DxO is a joke. Their testing methodology is so bad that it results in low end APS-C DSLRs having a higher IQ ranking than medium format digital backs. Worse, the people performing the tests are so clueless when it comes to photography that they actually published those results rather than realizing the obvious, going back to the drawing board, and fixing the tests. Such people shouldn't be trusted to test a Kodak disposable camera.

I can say with confidence that if DxO has ever published a correct DR measurement, it was by accident. Their DR results are dead wrong for every camera I'm familiar with and/or have tested (properly) with a transmission step wedge.

I had said (and I got like maybe 30 smites in one day ;D for it, even though I said that I hated to have to say this) that the D800 would trounce the 5D3 for DR at low ISO by over 2.5 stops and more like 3 usable and that the 5D3,

It's the Internet. You can say whatever you want. But until you can back up your claims with real shots of transmission step wedges and actual, real world scenes, it's a waste of time for anybody to pay any attention to what you (or DxO) claim.

We will see what the real DR differences are when DPReview publishes their tests. Until then people should remember to not feed trolls.

 

[Bosman]

Re: DxO results out for 5D3

I'm with Daniel, their lab scores mean nothing to my real output.

 

[SpartanWarrior]

Re: DxO results out for 5D3

Numbers mean nothing to me and DxoMark are a bunch of clowns just like the Nikon trolls, trust your eye's and not BS numbers
http://luminous-landscape.com/essays/eyes-vs-numbers.shtml

 

[gmrza]

Re: DxO results out for 5D3

I'm with Daniel, their lab scores mean nothing to my real output.

To add to that, I don't think it is just about the sensor any more. Sensors have generally become so good (it doesn't matter which brand you go to) that really you have to look at the overall package, and then decide which one meets your needs.

with the 5DII, Canon could still get away with putting a great sensor in an otherwise mediocre package. People complained about that. In general, however, 5DII owners were not complaining about the sensor.

It will be interesting to see how success in the market plays itself out. I think Canon has answered the concerns of wedding shooters, who represent one of the biggest markets for the 5D series.

 

[V8Beast]

DXO scores are only any good to DXO. Their testing methods are bunk tbh.
They scored the D800 higher than a IQ180, morons.

For those that don't know, that's scoring a 14bit back higher than a 16bit back with twice the photon real estate and sensor size. Hahaha.

Haven't you heard? All the medium format guys are now switching to Nikon.

 

[cliffwang]

Fascinating.
- The red stripe on the Nikon is VERY important to you....
- The popup flash on the D800 is NOT meant for use as a flash per se. It is meant to be used as master for controlling other flashes in the CLS (Creative Lighting System), which beats the hell out of Pocket Wizard and other primitive 'on or off' systems.
- You seem to think high MP is only for cropping, if you have 36MP one must crop, downsizing is impossible. Or maybe one gets sloppy and lazy from having SO many pixels.... what exactly is your point here?
The D4 doesn't have 36MP because it is impossible to have a ultra-fast camera with so much data to be shifted for every frame. Believe me, if it was possible, they would have 36MP in the D4. Maybe the D5 or D6 will get there. Depends on how successful R&D at Sony and Nikon are at resolving issues with shifting all that data in and out of buffers and onto cards.
- D800 handles low light as well as the MkIII. The difference is in fact so small that you have to pixel peep to see a difference.
- No, don't let online tests and reviews influence you. After all, the red stripe is the biggest issue, for sure.... :

Good points. I don't have 5D3 and D800, so I cannot tell which one is better. However, many websites give D800 really good reviews. People have to think about why only people here believe 5D3 is better D800.
Honestly if I don't have many Canon gears, I might jump to Nikon. It's too late for me to switch to Nikon.

 

[Fishnose]

DXO scores are only any good to DXO. Their testing methods are bunk tbh.
They scored the D800 higher than a IQ180, morons.

For those that don't know, that's scoring a 14bit back higher than a 16bit back with twice the photon real estate and sensor size. Hahaha.

Uh, no... the IQ180 beats the D800 handily on Color Depth (of course), is pretty much equal on DR and falls way behind on ISO (of course).

The total Sensor Score is nothing more an average of these 3 factors. Which of the 3 one has highest priority for you defines what camera (or type of camera) you should use.