Nikon D4s Beats Canon’s 1D X in DxOMark Tests, Falls Short of the Df in Low Light


The Nikon D4s is only a marginal improvement on the D4, but that didn’t stop Nikon from touting it as significantly better in low light and faster to boot. The question is, do those claims hold up when you put the camera through its paces?

Well, the DxOMark team did just that, and the results may or may not surprise you depending on how much stock you put in press release boasting.

When you put the three similarly-equipped Nikon full-frame shooters side-by-side, the differences are slight. The D4s only just beat out the D4 in low-light performance — and by “only just” DxOMark means “not that you’d notice in real word use” — while still falling short of the current low-light king: the Nikon Df.


The other main comparison DxOMark did was to pit the D4s against its primary rival: the Canon 1D X. And although Canon lovers might cringe to hear it, Nikon’s flagship full-frame is still outperforming Canon’s in every one of DxO’s categories.

Most impressively, they note that the D4s boasts “a close to +2 stops improvement in dynamic range at base ISO,” over its Canon competition.


A single sentence in the DxOMark conclusion pretty much says it all when it comes to summarizing these results. “Although some slight gains have been made minimizing shadow noise and expanding the dynamic range, sensor performance is largely unchanged.” And that, folks, is why press releases — much like patents and rumors — ought to be taken with a grain of salt.

(via Imaging Resource and The Phoblographer)

  • Peter “Pots”

    This is kind of like Mensa scores…common sense always rules.

  • lord eels

    Clearly fuji was a winner if it was tested …stfu canon and nikon

  • Steven Wade

    Lol, please. I would like to see you shoot sports with a Fuji and their terrible AF system and really high frame rate.

  • lord eels

    fuji xt1 is the answer my dear, its only a matter of time before canikon fall behind… really 1dx or d4s? overpriced toys

  • Roger

    DxO test are absurd… lab tests rarely reflect real world shooting. I shoot and love Nikon, but it’s pretty well known among us professionals that while 1Dx files may exhibit VERY slightly noisier files at higher ISO values, the Nikon does this at the advantage of detail in the image. D4s is supposed to help here but these tests don’t reflect that in the slightest.

  • Adam Cross

    point me towards a full frame fuji sports and heavy journo camera?… oh wait.

  • Bart Simpson

    I must not feed the troll.I must not feed the troll.
    I must not feed the troll.
    I must not feed the troll.
    I must not feed the troll.

  • Bart Simpson

    Please, don’t feed the trolls.

  • Daz

    sounds like someone’s jealous?

  • Steven Wade

    Again, lol please.

  • harumph

    What happened? You traded in all your pro Nikon gear for an XT-1, and now you’re going to start trolling Nikon threads with your nonsense?

  • Rune Stenseth

    I am still waiting for the D4X, more pixels would be nice for us large format printers.

  • Mike

    How many Fuji, Sony, Olympus, etc. cams did you see at the Olympics? It was Canon and Nikon, that’s it. Pros choose the best tools, they are not really that brand loyal.

  • Pokeken

    Wow a person who owns and uses camera gear in the real world. It’s so rare in forums these days. For those who only ever shoot randomly in yr town centre or on your tour guided holiday, his key point is that each camera has their own strengths and weaknesses so you need to know what you need from your type of photography in choosing these great cameras.

  • harumph


  • Sean Walsh

    Yeah, I got crapped on for stating that the Df’s ergonomics aren’t bad at all after I shot a full wedding day on mine, while I left the D700 to relax in the bag. Trolls and pixel-counters hate it when people who make a living using cameras in real situations debunk the myths others have conjured.

  • Nick Bedford

    In the real world, the difference between IQ on those cameras will make so little difference to your pictures that you are only really making a choice between secondary features and ergonomics.

  • David Vaughn

    I wish there was a way to measure standard deviation in these tests. Like, ‘x’ camera has 13 stops of dynamic, give or take half a stop because of discrepancies in the specific copy/test.

  • Stevo

    I’m sorry, but you’re comparing a 1300 to a 6,800 and 6,500 camera system. Wow. These bodies cost so much for a reason. And no, you cannot achieve the same results if you’re comparing between any Fuji and 1dx-d4s.

  • lord eels

    dream on

  • Richard Jones

    I still love my old canon 5d mk 1 simple, straightforward and easy to use :-)

  • Edward

    OK, I am quite happy that I probably (?) own the “low-light king” (Df), but does anybody have a reasonable explanation how “The D4s only just beat out the D4 in low-light performance ….. while still falling short of the current low-light king: the Nikon Df.” because it doesn’t make sense to me. They all have the same sensor right?

  • Csaba

    Love browsing these results, but the low noise at high ISOs is really irrelevant there as the same ISO values are producing way different sensor sensitivity results with the two major brands.

  • clifflwms

    I agree, I’m a D4 user, and I prefer it to the 1Dx, but I think the DxO scores are more friendly to Nikons, so comparisons between the brands become basically irrelevant.

  • Syuaip

    kudos for you. i dream to have one also.

  • Steven Solidarios

    D800s is coming….

  • Rune Stenseth

    Sure, I use a D800e alot, the e model gives much nicer detail, but I would like the added dynamic range of the D4 with the extra pixels, and some cleaner high iso images than the D800 pair gives. Hasselblad and Leica does not have that filter the D800 have, which I think is not needed, you never get any moire trouble anyway with that resolution and with the software removing any if you get some.

  • Graxxor Anandro Vidhelssen

    That’s because he’s now unemployed and has too much spare time on his hands.

  • Graxxor Anandro Vidhelssen

    hear hear! Oh, and someone above said “common sense.” It’s quite clear that photo technique, variation in position and lighting between shots and just general skill is far more likely to make a difference than taking, say, identical photos with two different bodies.
    I’d even wager, though I’ve never done it, that one could take two very similar shots with two different bodies from different brands if one were to use the same sensor, lens and settings.

  • Graxxor Anandro Vidhelssen

    I suspect that the D800 sensor is about as good as is physically possible currently. And of course, while a “price-no-object” object (!) might be able to improve a little on the 800E, we are starting to brush up against the quantum limits of sensor efficiency and low-light, random photon count variance, which obviously puts a minimum threshold on SNR for a fixed sensor pixel density.
    I think that most benefit would be found from moving up to medium format digital in terms of quality.

  • Graxxor Anandro Vidhelssen

    Then they would have to test more samples of each, rather than just take an average of three devices each. To be honest, DxO should be used as a general guideline rather than “scientific proof.” If one is so concerned about performance, I recommend just purchasing the said device and giving it a thorough burn-in test in any case and statistics-be-damned.

  • Graxxor Anandro Vidhelssen

    Sensors are analogue at heart and support circuitry includes complex analogue components (voltage regulators etc) which can affect the performance. Then there is the DSP on the output.
    Although considered digital, these cameras are so complex that their sensor implementations are no longer just black and white ;-)