DxOMark Redo: 5D Mark III Sensor Better than the D800?

Nikon D800 vs Canon 5D Mark III

Shortly after the Canon 5D Mark III and the Nikon D800 were released last year, DxO Labs published the results of the sensor performance of both cameras. The initial tests showed much better results for the D800’s sensor than the 5D Mark III.

Now, the DxO team is singing a different song after further comparisons between the cameras using high quality lenses.

Mounting a high quality lens makes all the difference. From our database, the resulting P-Mpix scores from various body/lens combinations reveals the D800 and Canon EOS 5D Mark III may be closer in resolving detail than expected.

When comparing the huge volume of data accumulated over measuring 147 lenses, one very surprising result was revealed. The average sharpness scores of the Canon EOS 5D Mark III matched the Nikon D800 and if the results were based solely on the mean average, the Canon actually out-performed the Nikon.

Either way, the Nikon D800 sensor simply isn’t as adept at resolving detail, pixel-by-pixel, as the Canon EOS 5D Mark III. There’s another surprise as well.

[DxOMark via Planet 5D]



  1. Corey says

    Anything to try to stir up more drama and generate some sort of revenue. I own both these cameras and I’m done Numerous tests and well they’re both great cameras in different aspects, the D800 Resolves much more detail that’s a plain fact.

    And as far as the D800E is concerned that camera outshines both the D800 and 5D Mark III

  2. says

    What is most impressive to me about the D800/E sensor has nothing to do with resolution. It’s the dynamic range. That’s what I covet.

    • says

      As a Canon user I am surprised to hear how well the 5D3 resolves detail compared to the D800…but quite honestly the more important spec for me is the greater dynamic range the D800 can produce. By the way…no one ‘bashed’ Canon…..the D800 is simply an all round better sensor, so quit being such a fan boi.

  3. says

    If it takes computer software to truly tell the difference between a photograph taken on a Nikon and the same photo taken on a Canon, does it really matter which camera one gets then? Obviously the human eye can’t objectively notice the difference so the argument is rather moot, no?

    Also, there may be an underlying problem with this test: was the same lens used for both cameras? If so, then one is truly comparing sensor to sensor and the test is valid. But if the test uses a Canon lens on a Canon body and Nikon lens on a Nikon body, then you are testing and comparing a combination of the 2 pieces of hardware and not just the sensor.

    It also points out one obvious but commonly overlooked fact when it comes to comparing cameras: it’s the lens that really matters and not the sensor. Any photographer worth their salt knows this.

    • says

      All good points TJ.

      I would like to see DxO take an approach to use a lens with interchangeable mounts (e.g., a Zeiss CP.2) when comparing two cameras like this. However, I think your first point is spot on concerning the need for a computer to tell us the difference between two cameras…

  4. Edgar Siscar says

    It depends on which camera brand you believe in. For those who believe in one, no persuasion is necessary; and for those don’t, no persuasion is enough. Just shoot them all right!

  5. John Marc Green says

    Most newspaper photographers I have known have shot Nikon, because that’s what the companies bought for them, and they are mostly shooting hard news, sports, politics, events, etc. So I see a lot of news and sports guys with Nikons. On the other hand, most wedding photographers and commercial photogs I see with Canons. Portrait shooters are a toss up. People that never do video are likely to be Nikon or Canon shooters, but people who do primarily video or a mix of video and photo tend to shoot Canon from what I have seen.

    I think Canon adding the All-I codec to the 5D mark III probably cinched it for video until Nikon provide something similar; the biggest weakness in DSLR video besides moire is probably the H.264 compression, and the All-I codec is not only less compressed, it’s a lot easier to work with straight out of the camera. No ProRes conversion needed really, unless you’re still laboring with FCP 7. And with these new results, it’s clear there’s not enough of a quality bump to warrant ditching Canon for Nikon, which always seems just a step or two behind Canon in the video department.

  6. Eugene Powers says

    I also own both of these cameras and there is no question that D800e is much better at resolving than 5D3 and dynamic range. But 5D3 is much more adept to what I am shooting. So in reality they compliment each other.