Nikon D5200 DxOMark score: better than the D3200


DxOMark published their test results for the Nikon D5200 camera. The camera scored slightly better than the D3200, which is interesting since they both have a 24MP sensor manufactured by two different companies (Toshiba for the D5200 and Nikon for the D3200). Here is the overall ranking for Nikon DSLR cameras:



This entry was posted in Nikon D5200 and tagged . Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • AM

    Yawnnnn! Wake me up when the D400 arrives.

  • Eric Duminil

    The overall DXOMark score is utter bullshit when comparing different sensor formats or sensor producers.

    Give me a D3S sensor over a D5200 anytime of the day.
    Maybe it’s relevant for D5200 vs D3200.

    • Sensor + optics. It would be interesting to compare, say, a 33mm f1.2 lens on a D5200 with a 50mm f1.8 on a D3S (assuming similar quality lenses). The problem is the necessary lenses don’t exist so you never see a fair comparison (but you could get close with, say, that zany new demagnifier using the same 50mm lens on both).

    • rhlpetrus

      You need to read the graphs to understand the DxO Mark results. The D3s sensor is better than the D7000 sensor in high ISO, but not for pushing low ISO shadows. That’s why the D7000’s sensor scores well. We have to see if the D5200 is as good as the D7000 in that regard.

  • Eric Calabos

    I dunno why %50 of DSLR market should be in hand of Canon, with this miserable lag behind numbers. Why a company shouldnt be failed while has nothing better (or at least up to date) to offer?

    • dxosuck

      you and dxo don’t know anything about DSLR!

      • Eric Calabos

        Thanks, now my knowledge is at the level of one of the most respected brands in photography community

      • Tommy

        Cute. You’re a Canon owner, aren’t you?

      • rhlpetrus

        Good one! Try to get a job at DxO then.

    • jake

      maybe the AF , the AF of the 5D3 was incredibly good , I borrowed my partner’s one 3 times for a few weddings.
      but the biggest problem with the Canon 5D3 was the annoying magenta cast in all images.
      I think Canon’s sensor in real life is not as bad as DXO says but it is clearly behind the Toshiba sensor in the D5200 or the D600.
      btw, imo, the current best sensor in 135mm market is the D600 sensor and I am going to replace one of my three D800 bodies with it.
      so, I guess horses for courses always.

    • iamlucky13

      DxO numbers are of limited utility. Handy to some degree, but the whole package of body, sensor, processing, etc is what you’re buying and using, not the DxO chart.

      I’d bet if you take proper comparison shots, you won’t see a drastic difference. I’ve got a good feeling the D5200 results will still beat the 650D, but won’t blow it away.

      • rhlpetrus

        My D7000 performs, relatively to my older D80, exactly as DxO Mark predicts, in terms of both high ISO and base ISO DR. It’s very good at predicting IQ performance in my exp.

        • iamlucky13

          Naturally. I’m not questioning whether the measurements are accurate according to their test criteria.

          I’m questioning whether the test criteria are actually meaningful in terms of image quality. In particular, should the 650D be scoring so poorly compared to the D7000 or the D5200? DPreview also shows a moderate deficit in dynamic range for the Canon compared to the Nikons, but very closer performance in terms of noise. It’s honestly hard to pick between the two looking at their test samples.

  • When D3200 got DxO score you announced: “Nikon D3200 DxOMark score: second best APS-C camera ever”.

    D5200 is, now, the first best APS-C camera. Ever. Two points over the last champion, Pentax K-5.

    This should deserve a best announcement.

    • Jer

      The Pentax K-5 IIS will overtake the D5200 and be the king of APS-C once again after it’s been tested.

      • Mato34

        Why do you think it should be so? Anyway it has been tested, and its performance is just the same than K-5 and K-5 II, so D5200 beats it.


      • desmo

        you should post on Pentax boards

        maybe they might believe you

      • jake

        isn’t it already there listed in the list?

        the K52s is just the same camera as the K52.

        I used to own the K5 and it was a indeed very good camera but it has its own quirks such as extremely slow wrting speed and mode dial response.

  • LeGO

    The overall DxOMark average score is not very helpful. One has to look at the performance in terms of SNR, DR, Color and Tonality at different ISO setting to understand how each sensor performs. The DxOMark average score and thus the ranking based on this is misleading and almost useless.

    • timon_comment

      I ignored the DxOmark Scores, Camera sensor and Lens.

      Screen mode, SNR18, Dynamic, Tonal, ColorDepth
      d5100/ISO84, 38.0dB, 13.04EV, 8.51bit, 22.0bit
      d5200/ISO69, 36.2dB, 13.05EV, 8.19bit, 21.8bit

      Print (8MPix, actually is valueless, merely A4 sized),
      d5100/ISO84, 41.2dB, 13.56EV, 9.03bit, 23.5bit
      d5200/ISO69, 41.1dB, 13.86EV, 8.99bit, 24.2bit

      The d5200 is the measured ISO69, but the d5100 is ISO84, that is saying the screen mode ISO84 the d5200 to have a slightly lowered variable actually.

      about the d5200 and the 3200,
      the d5200 is 14bit RAW, but the d3200 is 12bit RAW, and the d3200 is the lowest level model, so both models have a measurable small differentia is no surprise.
      the d3200 sensor is Nikon-developed, (not manufacture). The d5200 employs Toshiba sensor (Copper process, 4 layers).

      the d5200 sensor is even smaller 3.8μm pixel pitch 24MP, would be more sensitive to vibration, but the d5200 is a faster fps the easier increased vibration. The pixels increased a maximal number (APS-c), but the d5200 is likewise no electronic 1st curtain mode, ( competitors have using some years).

      The handhold shaking influence also went up, a smaller pixel pitch.

      a smaller pixel pitch always has even harder of the imaging quality of the Border and Corners, it could only turn into worse but not better. The sensor micro-lenses will be continued at next 5 years or longer.

      Most of people have really demanded the 24MP in APS-c camera? The 16MP in an APS-c camera is not enough? Nikon is running out of tricks and an only last way is wildly heavily-packed-megapixels? Nikon is already no better way?

      I would rather Nikon go back a 4.8μm 16MP APS-c sensor.

      • timon_comment

        DxO tell us like below,

        DxO’s Dynamic range is defined as the ratio between the highest and lowest gray luminance a sensor can capture. However, the lowest gray luminance makes sense only if it is not drowned by noise, thus this lower boundary is defined as the gray luminance for which the SNR is larger than 1. The dynamic range is a ratio of gray luminance; it has no defined unit per se, but it can be expressed in Ev, or f-stops.

        “The (DxOmark) dynamic range is a ratio of gray luminance”. “SNR is larger than 1”, (that is saying SNR > 0dB).

        So, even if the Screen mode, DxO there are the same two DR, but you will be still incapable of knew highlight range’s difference in between the two cameras.

        d5100 SNR 0dB, Gray Scale 0.012%, ISO 84 measured
        d5200 SNR 0dB, Gray Scale 0.012%, ISO 69 measured
        d3200 SNR 0dB, Gray Scale 0.018%, ISO 71 measured
        d700 SNR 0dB, Gray Scale 0.027%, ISO 162 measured

        Nikon d700, (DxO) DR 11.85EV (screen), ISO 162,
        Sounds like the d700 (DxO) DR is lower than the d5200, actually is obviously unequal ISO, and base ISO is higher (shadow area) noise on SNR 0dB.

        In ISO 800,
        d5200 SNR 0dB, Gray Scale 0.056%, ISO 547 measured, (DxO) DR 10.81EV (screen)
        d700 SNR 0dB, Gray Scale 0.045%, ISO 651 measured, (DxO) DR 11.18EV (screen)

        This DxO DR problem is likewise appearing in other cameras.

        • rhlpetrus

          The way DxO Mark measures ISO settings is different from camera’s setting. They use saturation. It may mean that the distance from midtone gray (18%) to clipping in the D5200 is larger, meaning more HLs headroom, compared to D5100’s.

      • rhlpetrus

        @timon: You have it wrong. The Print mode equalizes size, so that’s what you should compare cameras. It’s not “A4” size, it’s normalized to 8MP. You may normalize to 16MP (as D5100’s) and print from there, the advantage will be in D5200’s favor (same relative difference as at 8MP, renormalized). And you will lose even more IQ if you upres the D5100’s image to 24MP to print larger.

        • timon_comment

          you ought to normalize to 4MP, not 8MP. you merely needed to print A5 and a cellphonecam. merely you copied from DxOmark, no else.

          • timon_comment

            Served as a DxOMark fanboy?
            you wanted with 8MP to print A1 or A0?

            DxOmark said below,
            “For DxOMark Sensor Overall Score and Metrics, we chose a reference resolution equal to 8 Megapixels, which is a bit less than a 12″ x 8″ print with a 300dpi printer”.

      • timon_comment

        Nikon d5100, ISO 84 measured,
        SNR 0dB, Gray Scale 0.012%, SNR 45.5dB, Gray Scale 100%,
        Nikon d5200, ISO 69 measured,
        SNR 0dB, Gray Scale 0.012%, SNR 41.9dB, Gray Scale 100%,
        Nikon d3200, ISO 71 measured,
        SNR 0dB, Gray Scale 0.018%, SNR 42.2dB, Gray Scale 100%.

        Nikon d5100, ISO 165 measured,
        SNR 0dB, Gray Scale 0.020%, SNR 42.0dB, Gray Scale 100%,
        Nikon d5200, ISO 138 measured,
        SNR 0dB, Gray Scale 0.018%, SNR 40.5dB, Gray Scale 100%,
        Nikon d3200, ISO 140 measured,
        SNR 0dB, Gray Scale 0.033%, SNR 40.0dB, Gray Scale 100%,
        Nikon d700, ISO 162 measured,
        SNR 0dB, Gray Scale 0.027%, SNR 44.8dB, Gray Scale 100%.

        Nikon d5100, ISO 673 measured,
        SNR 0dB, Gray Scale 0.073%, SNR 36.4dB, Gray Scale 100%,
        Nikon d5200, ISO 547 measured,
        SNR 0dB, Gray Scale 0.056%, SNR 35.9dB, Gray Scale 100%,
        Nikon d3200, ISO 573 measured,
        SNR 0dB, Gray Scale 0.119%, SNR 34.6dB, Gray Scale 100%,
        Nikon d700, ISO 651 measured,
        SNR 0dB, Gray Scale 0.045%, SNR 40.9dB, Gray Scale 100%.

        The d5200 the measured ISO sensitivity has a larger downward deviation, that is the d5200 performance to have a slightly lowered variation actually.

        Looks the d5100 has a quite well APS-c sensor, the highlight area have higher SNR, not only for shadow area.

        Nikon d4, ISO 75 measured,
        SNR 0dB, Gray Scale 0.016%, SNR 47.4dB, Gray Scale 100%,
        Nikon d600, ISO 79 measured,
        SNR 0dB, Gray Scale 0.010%, SNR 46.0dB, Gray Scale 100%,
        Nikon d800, ISO 74 measured,
        SNR 0dB, Gray Scale 0.010%, SNR 43.3dB, Gray Scale 100%,
        Eos 1dx, ISO 80 measured,
        SNR 0dB, Gray Scale 0.043%, SNR 45.1dB, Gray Scale 100%,
        Eos 6d, ISO 80 measured,
        SNR 0dB, Gray Scale 0.036%, SNR 44.4dB, Gray Scale 100%,
        Eos 5d3, ISO 80 measured,
        SNR 0dB, Gray Scale 0.050%, SNR 45.4dB, Gray Scale 100%.

        Nikon d4 is the best highlight range SNR. Nikon d600 has a very excellent sensor.
        In SNR 0dB the Canon sensors are not very ideal shadow range, as a true causation brought of Canon sensor’s DxO DR is lower than Nikon.

        People should also notice the two points,
        1. If camera maker overly reinforces camera’s internal front-end noise reduction, which can bring sharpness worse, though looks like higher DR and SNR.
        2. In testing period, a longer run time could bring the shadow SNR to turn into worse.

  • Camp

    Does it really matter who makes the sensor since they are all (presumably) designed by Nikon? I’m asking…

    • Just stating the facts, it doesn’t really matter. I am still curious why they stopped using Sony sensors.

      • mikegorton

        Probably the same reason they stopped using CCD sensors, cost. When it comes down to it, the sensor itself plays only one part in the IQ equation. Nikon’s processing technology and algorithms probably play just as big a role.

        • rkas

          Thought they stopped using CCD because CMOS is better..

          • thomasv99

            no they didn’t; because it’s cheaper; only recently have the CMOS sensors becoming to catch up; you’ll still find CCD in very expensive hardware.

            • rhlpetrus

              Really, which CCD is better than present CMOS, or even the D300’s sensor?

            • thomasv99

              research the subject sir. Do you even know what CCD is? I’m not only talking about DSLR here. But to prove my point: uses a CCD sensor and you’ll find them in high end (medical for example) applications. Besides I said CMOS sensors can have similar performance nowadays.

      • Mato34

        I agree. Not a big thing, but just for the sake of curiosity, why this change existing two 24 MP sensors?

        Nevertheless, results from DxO might provide a plausible technical answer without looking into another ones reasons (economics, brand…), because it seems that at high ISOs performance is clearly better regarding dynamic range and color sensitivity.

        Perhaps nothing terrific, but it also bests the 16 MP sensor here.

        So real advance in performance could be the reason, if further tests support this results from DxO.


      • desmo

        I suspect Nikon wants 2 things for D7000/D300s replacement
        24 MP and better sensor performance
        low noise hi iso than D3200
        the D5200 is trial to see if they can obtain this result in a DX sensor
        it may be bigger challenge than many suspect

      • gomoku

        Well, there could be many reasons for that. It is always good to have options so that you do not depend too much on one supplier and you can get a better price and a more reliable production capacity. Sony is also a competitor.

      • jake

        Toshiba is a much better company as a sensor manufacture than Sony.
        But the real reason why Nikon is trying to find a good new reliable sensor supplier is that Nikon thinks Sony is becoming a big threat to them in this DX/FX market segment.
        as we all know Toshiba is not interested in selling its own brand camera like Sony and thus, it is a bit safer and more reliable partner to Nikon.

      • rhlpetrus

        They are not all designed by Nikon, who said that? Nikon usually puts some sauce into the final product, but Sony are Sony, Nikon are Nikon, Aptina are Aptina and now this is a Toshiba. It matters fro many reasons. Nikon does not depend, any more, on Sony for a high DR, linear sensor, like the ones in D800 and D600. They could make a 54MP FF sensor with the density of this Toshiba w/o having to rely on Sony. Sony, if one looks back, usually releases the sensors they use afterwards to Nikon. There are exceptions, like the D600’s. But it happened with D7000 and before.

      • Spy Black

        As I mentioned elsewhere, it may simply be a matter of demand. Sony may simply not be able to meet the production requirements.

  • DW

    DxO methods are stupid. They are just testing noise floors after boosting shadows so of course sensors with less shadow noise will do better. Of course the D5200 is still likely one of the best cameras in its class but obviously the DxO tells almost nothing. As if the D800E & D600 could beat the D4 for low light ISO sports

    • DW

      I tested the D600 to the 5D3 and the dynamic range of the D600 blew the Canon out of the water but the 5D3 was better in low light. Shadow noise in Dynamic range is all DxO is basically testing and my D600 shadows look amazing compared to Canon:

      • jake

        thanks for the link ,appreciate your time and kind words here.

      • Spy Black

        Why are you comparing it to the Mk III? If you’re going to compare the Mk III, Canon’s top of their line camera, to a comparable Nikon, that would have to be the D4.

        • CSPP

          I thought the 1D was Canon’s flagship and the 5D was more on par with the D800 except for FPS and MP.
          I own a D4 and a D800. Pops bought the D600 but it is not in the same league as my 2 beasts. I just laugh when I see it rated better than the D4 on DX0. As mentioned don’t go by these results.

    • rkas

      Uh, yeah those shadows are a party of the dynamic range, so why wouldnt they mesure that noise?

  • Myself

    Wasn’t the the sensor was built by toshiba, but outsourced by sony/nikon? Eg. Still same sony sensor just built in a not sony fab?

    • rhlpetrus

      No, it has a Toshiba mark on it. When Nikon outsources their own sensors, they carry a Nikon mark (like D3s, D4). The D800, D600, D7000 sensors carry a Sony mark.

      • thomasv99

        Do you have any proof for these statements? nikon could just as well have designed it and outsourced production to the toshiba fab.

        For the sensor inside the D600 I would tend to agree because sony uses the same sensor inside some of their DSLR.

  • jones

    i think nikon 1 cameras will catch canon apsc soon, canon sensors are pathetic

  • Back to top