Nikon D3x vs. Sony A900

See for yourself: Nikon D3x vs. Sony A900 (ISO100 to ISO6400 ISO samples accompanied with 100% cropped samples). This should put an end to the "Nikon D3x sensor - made by Sony, designed by Nikon" discussion.

This entry was posted in Nikon D3x. Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • looking at the D3x, first of all, i dont see how there will be any comparison to the A900 (quite frankly i dont think Nikon would allow that poor image quality to come from a “Pro” line camera), second, im still not 100% convinced on the D3x, jumping back and forth between 200, 400, and 800 it seemed like there was a significant loss of detail in each step presumably due to some heavy NR… we shall see i guess…

  • good post! Thanks!

  • Bianco

    Hmm, amazing difference. But why are all the scratches gone? Heavy (-handed) noise reduction in the D3x?

    • Siletto

      The shots were probably taken on different occasions, one of which the color patches was scratched.

      • rhlpetrus

        Clearly the color patch was cleaned up for D3x images. The lighting is also very different, so no comparison of detail at low ISO is possible.

        Now, about high ISO jpegs …. wow, D3x is just way better, even preserving detail.

  • Justin

    While it sure seems like the images of the D3x are much more pleasing at higher ISOs, the differences in exposure and the marks on the color checker make me think that this test is not all that conclusive.

  • bigmouth

    it doesn’t really tell whether the d3x sensor is different from A900’s. It only show how Nikon’s image processing is so much more matured than Sony’s which is understandable. In fact, many people have been suspecting D3 and D700’s high iso performance was the result of very sophisticated image processing on the Expeed side rather than the sensor. In that sense, Nikon’s treat secrete is probably it’s ability to turn a less attractive photo into something amazing. Adobe, please consider merger with Nikon.

  • PiGgY

    Looks like Lots of noise reduction applied, even at Low Iso…
    Need to see whow that translates to in real world shooting…Loss of detail at expense of more “pleasing look”…???

  • PiGgY

    Maybe it could automatically remove spots from the face of acne ridden subjects…:P
    That would be great…:)

  • somebody

    The D3x look better here, BUT…

    the lighting was NOT constant / the same. This can lead to different noise which is hard to campare. Plus, you see nikons usual approach: heavy reduction of rgb-noise and leave grain so it has more detail left. Look great in this case. but what about real world objects? Why can’t we see the fur? If you NR the last pic it looks similarly blurry as the sonys. It’s not a big step sensor-wise in my view.. maybe some tweaks, nothing more.. (not justifying 5k difference)

  • Anonymous

    i am sure you get better postprocessing with the nikon…. but it could still be the same chip

  • Anonymous

    and i mean “in-camera” postprocessing before outout to raw

    • JR

      there is no such thing as processing before output to raw.

      yes the camera makes a small jpeg to accompany the raw
      data for playback.

      raw is raw.

      it is OBVIOUS the sensor is of much higher quality than the sony’s. This, and the many ergonomic/feature differences easily makes up 5k in my mind.

      • Sloaah

        Not true. The Expeed processor handles the data, and using a different variant will result in different noise characteristics. The two sensors can be identical, yet perform completely differently using Bionz than when using Expeed

  • I’m just waiting for the stream of posts from the guy who will say that they are both lousy and that instead he is going to buy a Canon 5D MkII Canon 5D MkII Canon 5D MkII rah rah rah. I mean, that camera is so awesome that it can capture the tiny particles of dark matter that make up 70% of the universe.

    There, I’ve saved you the trouble of posting…

    • d40

      or a D40….

      • Stephen

        Thank you once again for you insightful reviews Ken.

  • Anonymous

    Lighting isn’t the same … I believe that the D3x will produce superior images than the A900, but this test doesn’t illustrate that. It’s not an apples to apples comparison.

  • alex

    the test is completely idiotic

    the charts are not identical or there was very different lighting

    also jpeg at default camera settings may not show the best possible

    this is another test that confirms my idea that most so called “reviewers” or “testers” are retarded

    • alex

      oh and let’s assume a900 has the same quality (can’t really be due to increased noise, higher low iso and lack of gapless lenses) but however.

      d3x, like 1ds, tries to have all the best features. a900’s rest of the features are behind and far behind canon or nikon.

      it’ll take lot more time to develop a good AF, good metering, good flashes and so on.

      they also lack a set of lenses. there is no pro wide zoom for sony.

      anyway a900 will be hard competition for 5dmk2. that’s for sure.

    • rhlpetrus

      Lighting is definitely different, just check shadows. Looks like for A900 they used a direct flash and for D3x other type of lighting. So can’t use these for detail But, for high ISO noise, yes, it’sd valuable and there the D3x is just waaaaay better.

  • It looks like the D3x has a pretty clear advantage over the A900. But, I still don’t have $8000.

  • faterikcartman

    Sure the Nikon looked better at high ISO, but if that is the same colour chart, the Nikon performance is 100% UNACCEPTABLE!

    I was shocked how much extra detail the Sony produced and how much over-processing was exhibited by the Nikon.

    I need a camera that captures the subject. If heavy-handed processing is going to be done, it is going to be done by me as post-processing.

    Please tell me those were different colour charts — and if they were, than that test tells us little regarding a comparison.

    • rhlpetrus

      Chart was cleaned up, the detail is all there, check eyes of rabbit and deffects around chats squares. I’m amazed people are not seeing that, to me it was obvious the first time I saw it the other day.

  • Lawfarm

    They have to be different charts, or the same chart after going through a war. But I think we can all safely assume, and agree, that for each camera, the chart was the same at the time that that camera’s pics were taken (i.e. all D3x pics were taken at the same time, with the same chart, and we can safely compare the D3x pics to each other). What amazed me, with those assumptions in mind was: 1) how great the 3200 pics were on the D3x, and how markedly crappier they got at 6400. What an amazing decline!

    • faterikcartman

      Good post.

    • Sloaah

      The same seems to be the case with the 5d mkii; once you have to extend the ISO range above its normal max ISO (6400), noise gets a lot worse, very quickly. There is also a very large difference between 3200 and 6400, in comparison to 1600 to 3200.

  • OK, I think the DSLR MP wall has been hit by the D3x and A900. Frankly, if your presently shooting a D3 or D-700, Your really going to be hard pressed to have any need for a 24 MP sensor. The $8000 price tag is just a little to high compared to image quality people already are getting. That, unless your a well heeled, must have the latest new thing, look what I got kind of person. Or if your just wealthy photographer. But then, your already into High MP, MF cameras anyway and Haven’t been waiting for this body.
    I don’t see many pro shooters and landscape artist complaining about the D3’s lack of image quality. Are you ready to stitch 3 or 4 24MP files?
    I really feel the D3x’s won’t be flying off the shelves until
    D3’s begin to wear out, maybe. Also, By that time a D-800 will be out, and the thought of lugging a new brick will give people pause before they buy a body with with no real improvements except for the sensor. But I could be wrong about all this. I’m no expert.

  • JPR

    Re the question as to whether the charts are the same, and the scratches you can see witth the Sony but not with the Nikon… there are images on the same web site of what is obviously the same chart taken with the Canon 5D mark II..
    …exactly the same scratches are visible as you see on the A900 images. So where have they gone on the Nikon? The patterns of the chipped edges tell you they are the same charts… Are the Nikon images out of focus??

    I own a Nikon – but if you asked me which camera I would buy out of those three on the basis of these images, I’d have to say the Canon – particularly given the price difference.

    • rhlpetrus

      C’mon, it’s obvious the chart was cleaned up, the detail is all there.

      • JPR

        And why would they want to go to the trouble to “clean” the chart if they were making comparisons? Unless they didn’t like the first results? Lighting differences seems more likely – scratches show more in oblique lighting – but then it’s strange that the shots with the Canon taken 3 weeks earlier show the same scratches? Why would they change the lighting just for the Nikon??

  • Justin

    Looks to me like the chart got ‘washed’ in between the two shoots. The D3x detail on the black part of the chart would be smeared away by any noise reduction that would get rid of those light colored marks that are on the A900 shots. So I don’t think it’s an issue of noise reduction, but a difference in subject.

    • rhlpetrus

      They just cleaned the chart.

  • jeff

    does anyone else think the ISO 50 shot looks MUCH better than all the others? The difference between 50 and 100 seems huge to me… in detail and especially colour. There seems to be less of a difference between even 100-800 (except maybe more noise).

    • rhlpetrus

      I think you need a better monitor or new glasses.

      • Jeff

        okay, so we know rhlpetrus doesnt agree. By the way my monitor is good and I dont wear (or need) glasses
        Anyone else?

        • rthomas

          Well, I do wear and need glasses, but my monitor is very high resolution and is calibrated (I’m currently doing archival scanning for a living). To my eyes, the ISO 50 shot seems to have quite a bit more color saturation. For some people, this will be a great reason to drop $8000… I’m not one of those people, my D300 is fine.

  • rhlpetrus

    Similar tests for D3x and 5DII, D3x looks good IMO, but again lighting is different and lens on 5DII is not same level (24-105, compared to 24-70 on D3x).


    • JR

      yet again, another reason to buy into the Nikon system.

      Canon’s 24-whatever lens is nowhere near the quality of Nikon’s. Same with the 14-24.

      If you are a pro, and not some rich dentist, these things matter.

  • Photo Guy

    Hi NR — Do you have a FORUM at which we (your readers) can write and talk together? No need to make it 10 rooms worth of space. Just a rumors section, a nikon talk, and an other cameras section would probably be fine.

    • I’ve been going back and forth with this idea – I am not a big fan of forums just look at dpreview, but lately I’ve been getting a lot of comments and I think I should offer a forum.
      I will try to put something together and see how it will go. Just give me some time.

      • JPR

        You are doing fine as it is. Many thanks for hosting what is an excellent site, with (generally) sensible and thoughtful comments. Don’t change a thing!

      • faterikcartman

        We do okay here don’t we, with a few thoughtful people posting comments just like a forum. But it isn’t really advertised and never reminds me of DPReview — which has its good and bad, but there is bad and perhaps one is enough?

      • STFU


        NR rocks, but every time I read the name in its abbreviated form, my head says ‘Noise Reduction’… Shows what kind of a techno freak I am 🙂

        Anyway, great post, pics look amazing despite all the idiots who spend too much time nitpicking every pixel. FFS, D3x has 24 Megapixels!!! Any noise would be expected, yet these charts show it is far better then A900. Like it or not, it is what it is, so stop complaining!

        P.S. Can you please add a preview button on the boards?

        • I am sorry please clarify – what kind of preview button? Do you mean for the comments?

  • pete

    this is a flawed test. “All shots were taken using Program mode with auto white balance”

    they should have used a grey card to set WB and controlled exposure in manual mode. far too many variables coming into play here.

  • Eric B

    This is a ridiculously telling test!

    People who are moaning about the color sample thing being different are bering amazingly short-sighted. People who are moaning about the low ISO are being amazingly short-sighted.

    For all we know, one test was taken a month ago and the other today. The temperature of the sensor might have been hugely different, the lighting is different, the D3x image might be slightly OOF etc. People talking about the D3x being heavy handed with NR are obviously not looking at the Sony images because the latter look like they were subjected to a media filter with a 5px radius.

    Anyway, the D3x is obviously much, much, much cleaner than the sony. I don’t care about 1 sec exposures at ISO 100… At those settings it is more due to lens quality and temperature than sensor quality. Look at ISO 6400, it doesn’t look much noisier than the D3!

    Personally I am stunned and impressed. I thought the sony was bad but not god-awful… but now, in comparison, I have to say Sony is still two year behind. What amazing quality from the D3x sensor! Not only that, the D3x produced files with less NR and a more pleasing noise dynamic, almost like grain.

    • Eric B

      Also it is worth noting the the D3x displays shockingly little color noise. Even pegged out at 6400 there are no color splotches to be seen. amazing!

  • Stephen

    So, I am looking very carefully at the detail on both charts, and they are the same. If you look at the pink square, about 1/4 of the way down on the right hand side, you notice the thread in both images. There are lots of other details that are the same. My guess is that Nikon is doing a whole lot of NR and we are not seeing the scratches.

    • Eric B

      What about ISO 100!!!! the D3x isn’t applying NR at ISO 100. It simply isn’t. The test cart is obviously different. I can say with certainty that the chart for the A900 is different (scratched) from the one used for the D3x. Nikon is applying LESS NR at the high ISO if you look at the texture of the area between the swatches this becomes apparent.

      • Stephen

        You can’t reproduce those details. They are incidental. Maybe the scratches got there on the second shot, but they are the same card, taken at similar times. A thread moves with the slightest brushing. So you are not likely to get the scratches out without moving it. Look at all the boarder threads. They are all in exactly the same place. I still think nikon is doing some NR even at low ISO.

  • ennan hamill

    this shows the sample images for A900 were taken in october (look at the dates) Given that the D3x has only just been announced I think it’s fair to say that the testing conditions were quite different. I believe that the scratched panels have been replaced for the D3x test which is why they havent shown up on the sample images. There is no way in hell nikon would have been that aggressive in their noise reduction as to completely destroy such obvious detail.
    Sony I’m afriad dont even come close to matching the Nikon.
    I would like to see a comparison taken at the same time under tighter conditions if only to shut the naysayers and canon fanboy detritus up.
    Still dont think its worth the price they’re asking though…

  • Interesting test. Although I’d like to see the results in low light – especially in shadow detail and IQ. These colored blocks were taken in relatively bright conditions. I’m more interested in what goes on when the sun is down.


  • cv


    there is almost no detail in the testshots. Simply big squares with one color.

    The A900 shots are overexposed compared to the D3X shots.

  • towert7

    That’s why we buy Nikon.

  • someone

    These shots were picked up from Japanese famous site where review new products.
    (In Japanese Only)

  • Michael

    Testing show the limits of actual sensor technology.
    But what about the future?
    Silicon based sensors capture 15% of incoming light – nanotechnology promises 80%.
    Available, but not here implemented, black silicon technology may icrease this ratio to some 50%.
    Therefore the D3x seems overpriced, according to expected technological developments. Otherwise, 24 megapixels and 3200 ISO is nice for non-professional users.
    Michael – Germany

    • steve

      You’re kidding right? You’re making decisions on whether this is a fair price for the D3x based on technology that’s 10 years down the road? How many photo ops are you going to miss between now and then??

      Problem with that is in 10 years, there will be some other technology 10 years away that will make buying then a bad decision.

      Seems it wasn’t too long ago people were lining up to buy a 1Ds @less than 12MP for $8000, now you can get double that in a much better body with much better ISOs for the same price. Heck, when the first 1-2MP DSLRs came out there were in the $20,000 to $30,000 range. This is a deal and a half by comparison.

  • Ralf

    Nikon really pissed me off with the pricing of the D3x so while I still have my FE2 I have sold my D200 and related lenses. The D3x was going to cost $12,000 in Oz so I decided to go for the Canon 5D MkII. I had to get some new lenses for a D3x anyway. I ended up getting the 5DMkII, 17-40, 24-70, 70-200and 50mm lenses for less that the D3x alone. I opted for the f4 70-200 something I have been waiting for Nikon to produce for 10 yrs. I will sadly miss my AIS lenses that I still have but I may get an adaptor ring for the Canon and I may also get the Nikon 14-24 using the adaptor ring.

    D3x had better be a great camera coz the Canon is pretty damn good and nearly a 1/3 the price. The video stuff is cool if in all likelihood unused by me.

    I wanted extra res for landscapes – I do little else so I could care less about high ISO performance after all in the past I shot with Velvia at 40 ISO.

    • Lou

      You are all kidding. The D3X is overpriced. Now concerning the pictures on the jpwatch site. I downed the pictures made with the D3X and the 1DSmk2. I compared them and I do not see where Nikon has any advantage. Yes, there is some more per pixel detail, but just slightly, then the colors are Nikon flat, like usual.
      Concerning the capture device. It is the same capture device as in the Sony A900, but, the way Nikon handles it is the same as if you compare the D200 device to all other same sensors in other brand cameras including Nikon consumer dslr’s. Nikon handles this sensor with the known multichannel processing technology we knew in the D200.
      In so far, the sensor produced for Nikon is unique and even being basicly the same as the Sony A900, the technology as such is different. In so far Nikon says right that it is a personal for nikon developed sensor.
      I do not go for tests and comparing of downed shots. Most are made on different days and no one can prove that none of them have been reworked.
      Overpriced, yes, the D3X is overpriced. The D3 body was ready made to receive the larger capture device and even if one analyzes the content in electronic components, there is less in it as in the D3. So, can anyone now tell me where those 3000 bucks go then. We can not compare old times to new times in pricing. Technology is known and development goes far faster then in those related times. Units are easier to produce and here, especially here it’s just a question of 2 different print-boards to exchange compared to the D3.
      If one compares the 5Dmk2 to it’s big brother the 1DSmk3, we are in front of a new concept, a completely different body, a new layout with unfortunately an overdue AF system. Despite of that, the picture it makes is better then the one of the big bro and the price is just half of it. There was really no reason for Nikon to overprice the D3X and if I compare the pictures to those of the A900 or the 1DSmk3 and the 5dmk2, there are no 3000$ hidden nowhere and that could justify that price. Nikon sucks and what they have done here is the best publicity for Canon and Sony.
      Sony’s A900 can have many features less then the D3X, the use I would make out of it is shooting pictures and just any of those cameras do this and quiet well.
      Let’s now wait for dpreview’s compared tests. They are not there yet, but they will certainly come soon.
      In the mean time stop kidding and come back to the floor of reality.

  • Lou

    the test are here, now get me in detail where the 3000 bucks are hidden …sic

  • Jabs

    From Dx0 web site.

    Nikon D3X = 88
    Canon 5DMK2 = 79
    Sony A900 = 78.9

    ANY further claims at to the Nikon and Sony sensors being the same – ANYMORE?


    D3X blows everyone away!

  • Back to top