Canon EOS 7D Mark II DxOMark test score: identical to the 5 years old Nikon D300s camera

Canon EOS 7D Mk II DxOMark score vs Nikon D300s camera
DxOMark published their test results for the new Canon EOS 7D Mk II camera - their main remark is that the camera's low ISO performance lags behind rivals. The 7D MkII overall test score is actually equal to the 5 years old Nikon D300s DSLR (dynamic range is even worse). Obviously the new 7D has several advancements not present in the D300s and many of you will question DxOMark's test practices, but here I just report their findings.

Here is the 7D Mark II vs. the Nikon D7100 comparison:

and the Canon EOS 7D Mark II vs. the first Canon 7D:

Canon EOS 7D Mk II DxOMark test score
DxOMark's conclusion:

"On paper, the Canon EOS 7D Mk II looks to be a solid choice for sports and action photographers, but its sensor performance is somewhat behind the best in class, at least at low ISOs. Relatively high noise, less discriminating color, and below-average DR at base ISO all continue to hold back Canon sensors against rivals, but that’s not the case at higher sensitivities. In fact, when light levels fall, the Canon EOS 7D Mk II performs competitively, even surpassing rivals slightly. If Canon could only address performance at base and low ISO, the EOS 7D Mk II would make a thoroughly convincing all-round choice, but in this category the Sony A77 II looks to be the more compelling option."

In contrast, the Canon PowerShot G7 X camera scored very well:

Canon PowerShot G7 X DxOMark test review
DxOMark's conclusion for the Canon G7 X:

"The G7 X has a high-speed lens with the widest range that sets it apart from the RX100 series. If the 24-70mm lens on the RX100 III is enough though the Sony is the better specified of the two, but then that's reflected in the steeper price. Although it was Nikon who was first to adopt and popularize the 1-inch sensor format, it was Sony who dominated this format with the RX100 series. Now with Panasonic and Canon adopting the sensor the field is opening up."

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

    Oh Canon’s fanboys will still buy this no matter what DXOMARK say.

    • Christopher

      I didnt. Im a Canon fan but i check the specs on this before the release and now I’m happy i bought a 6D instead for much cheaper.

    • Martin Kozák

      I bought Canon crap, should I feel sorry about it?

      • Martin you have a very beutiful website, could you please tell which CMS and template you used to ceate it? THANKS!

        • Sky

          It’s WordPress with
          HQ Photo Responsive WP Theme

      • Tom

        That is a beautiful folio, very dynamic!

      • Harry

        You surely know how to play with light mate. How did you light the bike in the wood in the bottom row, second from the left?

        • Martin Kozák

          There is one flash with red gel behind the rider and WB set to tungsten (which caused the blue tint)

          • Harry

            Love it, only one flash. And you were pretty close, you do know how to use a wide angle. I always have problems with those 🙂

          • Martin Kozák

            I used to have only one wireless transmitter, so I was limited and had to found a way how to live with it 🙂

          • Harry

            From left-behind, right? And it was pretty big coz everything had to be in focus, so you stopped down. Love it. I particularly like the colours with less yellow.

      • mikeswitz

        You succinctly show how stupid the fanboy arguments about cameras really are. It’s not the camera that takes great pictures, its the photographer.

      • Dee Melbihess

        Beautiful photos. But am I the only one having trouble with this site loading?

      • Pippo

        There is talk about sensors, not photography. Yours hard contrast photos on individual taste. If I want little bit softer, there will be problems with dinamic range 😉

    • Riccardo Andreaus

      That’s not about being a fanboy, but just using your scientific knowledge and judge the way DxOMark works. I mean:

      1) the deltaE is something you should always consider, in their tests the deltaE is so wide that every result should be inverted. If you’re skilled in electronics or studied science you’ll know what I mean.

      2) DxOMark is not testing the sensor quality, but only the quality of their own demosaicing algorithms with that sensor (and you know, that’s something we should not care…). If you want I can explain why it’s impossible to test the sensor as an hw piece (but if you look in their site they’re explaining that too).

      3) DxOMark started as an external software and testing lab for some big companies, among these companies there was Nikon. Now they are working for some imaging companies, here what you can found on their website “DxO Labs provides an image quality evaluation solution called DxO Analyzer for such manufacturers and organizations as NASA, HTC, Nikon, Kodak, Fujifilm, Panasonic, Olympus, Samsung, etc.”. Is it just a chance that the cameras from these manufacturers scores better than the others? I’m not telling they’re creating false results, but only that their software is made ad hoc for that cameras (algorithms).

      4) If you still believe that DxOMark is a great test, you should read what they wrote about Canon EF 70-200 f/2.8L IS II. Everybody knows that’s the best 70-200 ever made, it’s way better than the previous version (that was good, but the new and more expensive one is another planet), and that’s way most of the pros that used to work with the old version sold it and bought the new one (a professional photo/videographer never change his gear if there’s no real advantage). DxO wrote that the new one is not worthing the change. WTF?!?!?

      • Tom

        Thanks, I did not know that. It always looked dodgy though, and their trial software did not work very well with my Canon cameras. Your post explains why.

      • I think no matter how strange their test is, the method is applied to all cameras equally, So YES it is not good source of absolute evalution BUT you can compare cameras and se how they relate one to another. And that relation is true, because (i reapet) same testing is applied to all of them.

        • ItIsThatSimple

          If you test multiple devices with the same pointless procedures, you are going to obtain equally pointless results.
          Non-documented procedures are pointless by definition. DXO does not document their procedures.

        • Sky

          “you can compare cameras and se how they relate one to another.” – problem is that: you can’t. DXO favours certain type of data processing which not all companies use. Their results are screwed in one, quite specific direction that got little to non relevance to the real life cases.

          • doinkclown

            let me guess – they are only skewed against Canon

          • Riccardo Andreaus

            No, not only. Maybe you missed the explanation… read some post above.

      • Arthur Nazarian

        1)2) Yeah, so? They also state that the outcome is not 100% true, but within margins. And I don’t see why there should be a problem in the relative comparison between the cameras/lenses since they do their tests equally.

        3) Nice speculation. You’re a scientist, right?

        4) That’s indeed a great lens, but have trouble finding DXO’s statement. If I compare the results, it seems perfectly accurate:

        • Sky

          1)2) It is a problem cause what you will get are final images developed by your darkroom software. What DXO does is complete abstraction irrelevant to any real-life use cases.

          3) It’s not a speculation. It’s a fact that DXO says themselves. They even offer services to the corporate customers for “improving the image output”. Few years back Sony had a contract with them – after that they released A550/500. (I don’t know if they still cooperate).

          4) DXO lens tests are a separate topic – these are complete and utter bollocks – with numerous examples where lenses people buy are completely different to whatever came out from DXO tests. And: I’m laughing hard at them giving a lens resolution in megapixels. What a joke. Anyway, to answer your question – statement Riccardo talks about comes from here:

          • Arthur Nazarian

            1)2) That’s the whole point of scientific tests, right? If you want real-life tests, check real-life tets.

            3) I’m referring to the allegedly bias of DXO because they have cooperation with other firms. If I state on my website that I photograph several major companies, that doesn’t mean that I prefer them and make better pictures for them.

            4) That’s indeed crazy, it contradicts with their own results (at first glance)

          • Sky

            1)2) point of scientific tests is to get relevant numbers. Not numbers for a sake of having numbers.
            3) Think about that: these companies got their cameras adjusted to produce results that DXO mark scores high. So in the end – their results are biased, you can argue if it’s intentional (companies essentially buying high scores) or not (just a pure coincidence) or that Canon could have bought their service just like most of the other companies did (keeping in mind how tech-geeks seems to like simple numbers that DXO provides – it doesn’t require anyone to use his brain on order to analyse anything), but don’t try to imply here that DXO sensor marks are some sort of ultimate objective judgement of the sensor quality cause they’re not. In fact: they are very far from that.

        • Riccardo Andreaus

          1)2) Maybe you need to read again what I wrote: they are measuring not the quality of the sensors, but the quality of their algorithms with that signal. I think it is a very simple statement. I suppose you work with digital photography, so maybe you noticed how a different algorithm can change image quality differently in different camera (for example Adobe Camera RAW during the passage from CS4 to CS5 drastically improved the noise quality of RAW of the 5DmkII, but not the quality of the RAW of the 7D…).

          3) I’m a skilled professional that uses mainly Canon gear, but not only. I don’t understand what’s the point? It’s not a speculation, it’s a fact. They’re developing the image algorithms for that companies, it’s obvious that their software will work better with that cameras…

          4) After a big mess in the comments by a lot of users they shortened the review, but you can still found silly conclusions here:

      • doinkclown

        just your Canon fanboy opinion – nothing more

        • Riccardo Andreaus

          I’m sorry, but I showed facts, you’re not adding nothing to this conversation. Troll?

          • MB

            You really have more than a point about what DXO is doing … they claim some scientific testing procedures but fail to provide any scientific proof of methodology used …
            But I am not so sure about your claim that electronic circuit (sensor) is impossible to test as a hw peace … you are saying that it is not possible to test for example signal to noise ratio for an electronic circuit, that would be testing it as a hw peace and it seams to me that this is probably as accurate and scientific way to test it as it can be.
            The thing is this test just isn’t relevant enough, but you are right that “method” used by DXO heavily depends on demosaicing algorithm so they are effectively testing their software ability to interpret the data from a particular sensor …
            And for DXO lens test they are as irrelevant as it can be because they are not actually testing the lens itself but the data from the sensor that is interpreted (for better or for worse) by their software demosaicing algorithm … and as a result the final lens scores are totally incomparable …

          • Riccardo Andreaus

            Yes, sure, we could test the signal to noise ratio of a sensor, but we will miss a lot of other aspects: how that signal is converted by the electronics (that can drastically change the quality) and how the hw-noise reduction on chip can affect that signal (posterization, loss of dynamic range etc). We could speak about it for hours, but I think that image technology nowadays is so good that the difference from one camera to another are so small in real world that nobody can see them actually 😉
            During reportage photography I shoot at 3200 ISO and up to 6400 ISO with no problem with a 5DmkIII, something I cannot even think more than 4 years ago.

  • Tags

    Yep, I’m a canon shooter and I’m sure all the fanboy apologists will come out saying how unreliable DXO is etc etc. I’ve shot both systems extensively and canon sensors are as crappy as dxo says. It’s not a gimmick, they suck! I switched to canon for love though, and every time I shoot landscape i cringe because I just know I have to merge images to get something similar to what I had with Nikon. High ISO on 6d is very impressive though. Landscape is abysmal considering what the camera costs

    • DXOLobotomist

      It looks like you cannot expose properly. What should the ccd-based mf camera shooters say?

      • Eno

        He can expose very well, but Canon DR at small ISO sucks, period! If you never shot a modern Nikon or Sony camera you don’t know what a good DR means and all the talk is for nothing.

        I read that Canon in planing to use a Sony sensor in their top end DSLR, this means great DR finally.

        • Tom

          A while back, you posted in this forum about your need to recover 4 stops exposure in some video. It seemed to be a normal procedure for you.
          I am sorry to say that whoever buys this camera should be above that level. They should manage to expose properly, and then some.
          It’s also a matter of investing one’s money wisely: if you spend that type of money on a camera, you want to get the basics right to make it worth the price.

          • Eno

            You are not welcome here TROLL, everybody’s sick of your misleading comments and lies!

          • Harry

            In your place, I would thank the poster for sound advice. Too many internet “photographers” lack basic preparation. Too many others are paid to do “scientific” tests to expose this or that. Needless to say, they rarely know what they talk about.
            Myself, I have been shooting Canon digital for a decade and I never had problems. And I am not alone, definitely not 🙂

          • Eno

            Tom, Harry, Viezuvure, Pro, etc, the same troll under different names.

            I bet you know nothing about photography, composition, the golden rule, the rule of the thirds, perspective, any exposure formulas, etc, etc. Heck, I bet you don’t even know what DR stands for, especially how it’s measured.

            You are just an irritating troll payed by Canon to spread lies on the internet.

          • Harry

            What a concentrate of internet commonplace, and angry to boot. You surely make a case.

          • questus

            DR (Dressing Room) 🙂

          • Chad

            Ignorance is bliss.

            I shot Canon for eight years before switching to a Sony A900 from the original 5D. The difference in color and dynamic range was a revelation and Sony (and Nikon by proxy) have continued to widen that gap while most Canon shooters are oblivious.

            You don’t need huge dynamic range to correct for poor exposing skill, you enjoy it for a more film like reproduction.

      • Tags

        I’ve been shooting professionally for almost a decade… I can expose a picture properly. I would find it alarming if someone who knows what good IQ looks like finds canon base ISO acceptable for the given price point of the camera. You honestly think their low iso images look good in comparison to the competition? Acceptable… yes, for most purposes. Ideal? Far from it!

        • Riccardo Andreaus

          You wasted your time if you found Canon’s image quality different from Nikon’s image quality… It can only mean that you don’t know anything about IQ.

          • Tags

            you’ve got to be kidding me. It is dramatically different AT LOW ISO. you seriously can’t see it? I dunno maybe I’m an IQ snob.. I’ve shot large and medium format film for a long time. I know damn well what good IQ looks like. Canon sensors are totally garbage for landscape. Can they do it? Sure, they can. Should any serious landscape shooter shoot Canon for full frame landscape if they have a choice and aren’t totally invested in Canon lenses? NO. You would be a total dimwit to give up the massive gain in IQ if you had the opportunity to pursue other brands. Yep, I just don’t know anything about IQ Riccardo! Or perhaps you just can magically change the laws of physics when you shoot your antiquated Canon at ISO 100 and get IDENTICAL results to a recent Nikon FF sensor. Ok buddy. what, are you going to tell me next its all the photographer and not the camera? BS. Good cameras give good photographers better results all things being equal. DR matters.

          • Riccardo Andreaus

            You’re just trolling in my opinion. There’re no difference in DR at low ISO from my 5DmkIII and a Nikon D810 (tested with a friend of mine) in landscape, but my Canon was able to retain more information in the highlights. D810 was sharper obviously, but that is not about DR.

          • HF

            Not true. Clearly visible if you need to ETTR and boost shadows in scenes of high dynamic range,
            a simple google search shows some examples:

          • Riccardo Andreaus

            That’s quite true for shadows. What I noticed is that D810 retain more information in shadows, but lacks informations in highlights, that’s quite important in landscape photography because with my 5DmkIII was easier to ETTR. Using Lightroom to convert both RAW isn’t a guarantee of “same testing condition”. If you’re skilled in editing software you’ll know that each RAW converter acts differently on different cameras, Lightroom change drastically image quality when you choose different “camera profile”, speaking about how it performs with Canon 5DmkIII (that’s the camera I know more because I use it everyday) you can easily notice that in its default profile it changes the gamma curve in a dramatic way (and that profile lowers the quality when you need to recover highlight, but mainly shadows). That’s a different approach compared to RAW profile when converting a Nikon file (D810 is my test camera for .NEF), because the “standard” profile doesn’t change so drastically the gamma curve and yes, you can work better on highlights and shadows. Try it! I’m sure of what I tested and I’m sharing my conclusion here.

          • peterw

            … because of the arbitrary tonal profiles that can be chosen, it is better to measure ‘dynamic range’ of the sensor instead of compare photos shot with an arbitrary iso value… in an arbtirary rawconvertor.

            Hey, that’s what DxO does…
            It doesn’t matter if you recover highlights (Canonwise) or correct shades (the Nikon way). It matters how much information you can get from the sensor (and lens).

          • Riccardo Andreaus

            No, they use an arbitrary RAW converter (their algorithm), that’s why it is not scientific 😉

          • peterw

            Indeed, technically it is possibe that their algorithm is doing less well for Canon. You have a point.

            However, please look at my comment at Neopulse reference (3 days back) to a DxO-diagram showing developpement in sensors starting from 2003. You can see developpement in all APS-C sensors going about twice as fast as in Canons. I think this indicates that it is not a problem of DxOs algorithmes, but it is something about the Canon APS-C sensors.

            (It is a rather a sad thing, considering all the enthousiasm around the rumored new type of sensor which was hoped to be part of this camera.)

          • Tom

            I think this indicates nothing. If it does, it probably indicates that Canon sensors and DXO software have strayed apart. I don’t use their software anyway, that’s probably why I don’t see those fabulous differences.

          • PeteD

            Tags , don’t bother, you’re wasting time with zealotry. There is a huge difference and the workarounds Canon shooters had to go through are ridiculous. I wouldn’t expect this kind of hassle for the amount of money Canon charges for their products but here we are. The people that need IQ, buy A7/R/S and metabones adapter, or they buy a Nikon D750 or D810 if their work is landscapes, architecture or product shots. There is no debate, only denial and cognitive dissonance. You have no idea the amount of photographers i personally know that added Sony to their gear, Nikon, or just entirely dumped their Canon gear and moved to Nikon.

          • you are useless

            Nobody needed this extra bit of marketing, but thanks for the giggles herr Useless.

    • Guest

      Funny who many Canon apologists are chasing you! They are definitely hurt b/cs their secret is now open for everybody to see.

    • Dada Kaka

      Canon apologists are too furious now! Btw, everybody knows your secret now! There is no hiding for bad sensors.

  • Siva Kumar D

    10 fps and fast AF is what matters BTW I own D610!! and I see a need for speed!!

    • Eno

      I wold be very glad with only 4-5 fps but a much better frame coverage AF.

      • Eno

        You (like most keyboard warriors) don’t need new gear or new features, you need to learn the basics. Practise practise practise.

        • Eno

          Crawl into the cave from which you came from TROLL!

    • tim sims

      awesome ill go buy a sony a77ll then

  • stormwatch

    Canon sucks big time.

  • Also keeping it real

    Not a hint of prejudice in the Nikon Leica fanboy admin is there? Bet your rumor reports are just as biased.

    • Global

      The Admin is a Canon user. Try again.

      • Ufupuw

        Admin is Nikon fanboy and runs

        • Neopulse

          Admin run more than one website that relates to other camera brands like Photo- and Leicarumors for example. Try again

          • Sky

            And… owning Leica camera prevents you from being Nikon fanboy? What is this nonsense? You just discovered a magical pill that cures people from being Nikon fanboys?

            We know that admin runs 3 websites, not just one. It doesn’t change a thing. Read some of his posts – it’s quite obvious which brand he is favouring.

          • Neopulse

            Apparently Ufupuw didn’t know. Why are you speaking on his/her behalf?

      • Sky

        Yea, I’m sure he does, and like every other Canon user – he has in-depth knowledge about Nikon, spending most of his time on discussing Nikon products. lol
        You could just as well say that he is a Samsung user.

    • I was very clear in my post – I just reported DxOMark’s findings. Not sure where you got the prejudice part from? Want to see real biased reporting – go to any other rumors website.

      • nwcs

        Biased reporting? Nah, more like targeted affiliate links and some repeat of forum conjecture.

  • tim sims

    so did canon just use a sensor form 2011 for the 7d mill or?

  • Arthur Nazarian

    Why question DXO? They have the best, most thorough, and most scientific tests from what I’ve seen. If you question them, give arguments why their testing equipment or procedures are not reliable. They publicly show you how they test cameras/lenses, and I don’t see any errors.

    Of course, just a sensor doesn’t make a camera good or bad. But the sensor is about the only thing that can be scientifically tested. Other things are of personal opinion and preference. And that’s why I love DxO – they give you the indisputable results of the sensor, and the other features are for me to judge.

    • Riccardo Andreaus

      I don’t want to post again everything, so, if you want to know why their tests are not scientific and don’t reflects real results, read my post some lines above.

      • doinkclown

        nothing by a sour canon fanboy

        • Riccardo Andreaus

          Troll? Please read above and try to understand what I wrote.

  • ItIsThatSimple

    DXO don’t describe their tests in any detail. They simply fail to explain why their tests are good.
    As a consequence, we are entitled to think they are bad.
    As a corollary, I also think DXO are shills.
    As a further corollary, I think whoever believes their data is either incompetent or a shill.

    • If I compare DxOMark scores between Nikon camera models, the test results make complete sense to me. I think the problem comes when you start comparing between different camera manufacturers.

      • Riccardo Andreaus

        That’s (partially) the point! Comparing different cameras from the same manufacturer can bring a valid result, but that’s not sure actually: different cameras/sensors use different technology and different algorithm, so you cannot be shure that a different measure means a difference in hardware, maybe it’s only a better/worse behaviour by their (DxO) software. It’s always “software related”.

        • Harry

          And, if you make cameras and don’t use their service, the consequences will be visible in the internet. Got it, it makes perfect sense, photo-wise and commercially, too.
          Thanks again for posting that, it’s a good defense against the usual nincompoops (I don’t mean the admin of course, he seems very balanced).

        • HF

          Irrespective of software you use, the DR difference at base ISO is easily visible with different RAW converters in real life shots. I provided you three random links to show it. There are other reviewers comparing the DR, for example or They all show consistent results. Until proven otherwise I think their individual measurement curves to be quite reliable. If it is only software related, why hasn’t CANON provided a raw converter to give much better results?

          • Riccardo Andreaus

            I replied in the other post 😉
            sensorgen use DxO datas, so obviously there’s no difference… I have to check the first one because I don’t have time to read the way they test the cameras.

    • doinkclown

      they rate the G7 above the Sony RX100iii

      are they still shills?

      • Riccardo Andreaus

        You did not get the point. For me it’s not a way of testing cameras, their procedure is not scientific. I think that their wrong approach reflects in all their test (pro and against Canon).

  • johny

    Only manufacturers can measure their sensors, because you don’t know what has been done inside JPG or RAW.

    • Spy Black

      Take the blue pill…

  • Don`t trust DXO scores.

  • neonspark

    well so much for the “new canon sensor technology”
    I’m really curious if somebody with tech knowledge really looks into the canon hardware and verify that canon doesn’t make false claims as to their 14 bit processing engine. I recall when 14bit was new, canon had a rather easy time getting high fps while nikon lagged, even as they were both using similar chip processes. Eventually as chips got faster, nikon matched their fps but canon isn’t where the more’s law would put them had they really had this advantage to begin with.
    makes me wonder if in order to improve performance canon reduced the amount of data internally which is why dxo keeps giving them these ratings. Basically a canon 14 bit file is really a les than 14 bit file internally and the information is saved to 14 bits but the original data was lost, clipped, or merged in poor ways.
    why else does canon always seem to need to pack dual chips just to keep up, but doesn’t get 2x the performance but their raws have terrible low ISO noise in shadows and other strange artifacts of bit clipping. something is going on and likely they can’t fix it without a new digic engine from scratch that matches the work nikon/sony really do: true 14 bit raws.

    well just a theory anyway.

    • CookingTheRaw

      Here is another theory: Canon don’t cook their raws to match certain parameters. If you check their raws with those parameters, they will look bad.
      By the way, Sony is known to cripple their raws, at least in the A7 series. Never heard about C or N doing the same, but they are also in a different league.

      Theories aside, processed raw pics from Canon are very similar to Nikon’s. Properly exposed and properly processed raws, that is.

  • Kynikos

    To borrow from Jerry Rice:

    New camera, same old sensor.
    Same old sorry-ass*d sensor.

    • Andrzej Lukowiec

      For m4/3 haters those numbers, of course, mean nothing 😉

    • J Shill

      Oh please, leave that junk alone…

  • Marcelo Tezza

    Well, as nikon and sony user, but more than that, as an enthusiast photographer.
    DXo really has something against canon, but not in the scientific tests. The words they choose to talk about canon products are lame. Read the reviews and is easy to spot.
    First, a camera is a camera, not a sensor. Usability must not be forgotten too.
    Second, every camera needs a lens to work at its best, canon has some of the best lens on market, this is undeniable.
    Third, DXo rating lacks information and highlights some facts as the only ones you should pay attention, missleading.
    This camera has 65 cross-type AF-sensors in a DX body, covers a very wide area of the viewfinder; The AF-works better than d7100, for sure, about A77II or A6000, i have not seen any review that really can show it.
    So, if you can’t focus well image quality doesn’t matter. This camera is aimed to wildlife and sports photographers, who don’t want or don’t have money to spend on top end camera.
    This camera as DXO shows has better High ISO Dynamic Range and noise performance than sony and nikon. This seems to be forgotten by many people. The High ISO performance is key for the public aimed for canon.
    This camera is a winner for people who knows what they want.
    If i haven’t a camera today this would be on my top list without a doubt.
    Lets not be fanboys, this is just bad for everyone who is looking for the best tool for the job, not the best brand…

    • doinkclown

      they rate the G7 over the RX100iii

      your point sucks.

      • Sky

        Exception doesn’t make a rule.

    • peterw

      You state:
      “This camera as DXO shows has better High ISO Dynamic Range and noise
      performance than sony and nikon. This seems to be forgotten by many
      people. The High ISO performance is key for the public aimed for canon.”
      It doesn’t. In color rendition and in dynamic range D7100 are better, all other sensor aspects are equal (according to DxO ‘measurements’).

      Being a bird photographer, if starting from scratch, I’d choose the 7Dii – from these two cameras. But not because of the sensor.

  • Spy Black

    Not sure if I should laugh, or just shake my head in disbelief.

    • Harry

      Just chop it off with a tiny but sharp sensor 🙂

  • Hendrik Mintarno

    make sense, because canon aps-c has smaller size sensor compared to sony and nikon. but a camera review is not complete when it just end in sensor quality, you have to take into account all the features (weather sealing, af performance, fps, viewfinder magnification, button layout, menu system, etc)

  • Bungle

    People who comment that DXO favour Nikon speak out of their ass. You do realise that Sony have had a contract to supply Nikon with sensors for the last 2 years (with the exception of the D4s and Df i think); hence why the megapixel counts are the same for Nikon/Sony. So how come Nikon still beat Sony in tests. It basically comes down to the fact that Nikon have far superior in-camera processing than Canon and Sony.

    I maybe wrong, but I assume DXO consider actual image sensor processing from the camera rather than Photoshop/Lightroom etc.

    The DXO test is standardised and has been for some years. So why can’t Cannon pick-up their game? Surely they know how the test works and criteria to match Nikon and Sony.

    Plus you should also take into account that Canon Aps-c senors are smaller that Nikon/Sony; which in my opinion compounds Canon’s problems.

    Canon simply need to invest in their sensor development. They have become the Apple of the camera world in my eyes. People will automatically buy Canon.

    Don’t get me wrong, if you have been using Canon for many years and have good expensive glass, then thats fine with me. But newbies buying their 1st camera etc. are the ones getting conned by Canon.

    And yes, i am a Nikon user (D610) before you mention it. I did have a Sony A77 too before i moved over to Nikon. It was by far the best Aps-c camera for the money at the time of purchase. Never had a Canon because they are simply overpriced for what you get (my opinion).

    • Riccardo Andreaus

      No, that’s all wrong. They cannot test the quality of a sensor, they need to use their algorithms, and there is where all the troubles resides: they’re testing their algorithm, not the sensor. Nikon use (pay) DxO labs to develop and test their software/algorithm, so it’s obvious why they score better (nno because they pay, but because DxO works better with an algorithm that they developed)… and that’s also why Sony gets worse score…

      • Neopulse

        With so many replies you keep posting about the CxO scores being false, why not send a letter to answer all your questions pertaining to why Canon sensors aren’t doing well in comparison to the Nikon/Sony sensors.

        • Riccardo Andreaus


      • peterw

        lets try Capture One Pro and Adobe then and do these measurements ourselves…
        Still better, use Canon’s and Nikon’s own algorithms.
        I’m affraid I’m no longer capable of setting up such a test, but many of us can.

    • MonkeySpanner

      Nikon also sources sensors from Toshiba (D7100). And I don’t think the sensor size difference between Canon and Nikon is really that large, should not be large enough to make a noticeable noise level difference – if all other things we equal (but of course they are not).
      I think Canon has gone down some strange rabbit hole with the dual pixel thingy they were trying to develop, and have lost there way somewhat in sensor image quality performance.
      What say ye Riccardo – are you not entertained?

    • bonem

      Well said. I’ve always felt that way about the Canon Apple comparison.
      I bet if (big if) Canon was on top Ricco would be singing a different tune for DXO. He’s clearly expressed butthurt by the amount of defensive responses. Give it up man. It is what it is.

  • CW

    It’s not as if DXO is the only source claiming that Canon’s cameras have less dynamic range at base ISO. There is plenty of other evidence besides DXO to corroborate the claim that Canon cameras have less dynamic range at base ISO than Nikon or Sony Cameras. For example, DPReview typically does tests for new cameras in which they lift the shadows in raw files by 4 or 5 stops in ACR or Lightroom and see how much noise is produced. These consistently show that the latest Nikon and Sony cameras do better under these conditions.

    Another example from a different source comparing Nikon and Canon full frame sensors:

    In both cases, same exposure, same base ISO, same raw converter, but the Canon’s shadow areas fall apart when lifted too far.

    This also isn’t a matter of “exposing properly”. If you’re trying to maximize DR in a scene with lots of range and exposing to avoid blowing out the highlights, the degree to which you can lift the shadows without producing serious noise is absolutely relevant and represents the practical limit on DR for a given scene. On that basis, the current Canon cameras simply aren’t as good. They’re competitive on high ISO but aren’t as good on DR. Not terrible, but not state of the art.

    Given how widespread this view is, you can conclude that either there is a vast conspiracy against the MOST POPULAR CAMERA MAKER IN THE WORLD by EVERYONE ELSE, or that what they say about Canon’s base ISO DR is actually true.

    • Riccardo Andreaus

      Mmm… Don’t think so. Let’s go and take a look at dpreview and their Dynamic Range Comparison, most cameras by Canon have got more DR than any Nikon in the same price range (with HTP off and ADL off). When DLO and HTP are active you can see similar DR but more DR in the highlights in the Canons and more DR in the shadows in the Nikons. So they’re different, but the DR is almost the same. Remember that using the same raw converter doesn’t mean that you’re applying the same condition.

  • Joe

    I saw some sample shots on various independent reviews (action/low light settings). They look just fine to me.

  • Neopulse

    I honestly don’t think it’s that biased as you make it seem Riccardo Andreaus. In 2008, the Canon 5D Mark II was rated at 79 points in the DxO mark in September 2008 while the Nikon D700 was marked with 80 points in July 2008. It barely beat the Canon sensor in that time period.

    – The sensor that was top at that time in was the Phase One P65 Plus (July 2008)

    – The Canon sensor was tied at the time with the Sony Alpha 900 (September 2008) at 79 points also.

    – So in that time frame 6 years ago, Canon was in the top 5 sensors at the time.

    I don’t think they are that biased if you look at it in reference of time and the way technology advances. Will pass the link and you’ll see the marks along with the time frame at the bottom of the graph.

    • peterw

      I found this interesting diagram also.
      When selecting Nikon (and focus on DX / APS-C), you see a rather steep linear increase from 2004 to 2011. The rate goes from 40~50 (D2h~D70) to about 80 (D7000).
      Canon had the best results in 2005 (10D) but shows a much lower improvement rate untill now: from 55 to 70.
      The devellopement of Pentax (APS-C) is following (or heading) the line of Nikon.

      Following Riccardo Andreaus reasoning, DxO stopped developping proper algorithms to proces Canon APS-C sensors in 2005.
      I think it is more plausible from these data to assume that Canon did not improve their sensors as much as Nikon (and others) since 2005.
      Both theories can be tested.

      (I happen to use a Nikon camera myself. Most fine pictures I see on the internet are made with Canon cameras, since Canon is market leader. If I’d be using Canon gear, I’d choose a 5D iii. I’ld be extremely happy with it, and I’ld try to make good pictures.)

      • Neopulse

        Most photos though I have to admit that I’ve seen shot from a Canon were between ISO 100-800 (with/without lights) at the most and it’s possible to pull good looking photos from that range with little post effort. The 5D III I also had the pleasure of using very shortly before selling it. Only wish I could have used the 70-200mm II with it.

        • peterw

          If I’m not mistaken, the main difference between the sensors is rather in color rendition at any iso, and in dynamic range at low iso.

  • TheInconvenientRuth

    Kannun stahp… waht r u doing….

  • roos

    There has always been lots of different kinds of photographers. One of the aspects where we differ from each other is whether to use a copist or lab or to have once own dark room.

    If you are the lab kind of guy you can as well shoot fomapan. It’s cheap and gives enough dynamic range for a nice automaticly created copy.

    If you have your own semi-pro dark room you probobly know that if you pick a bit more expensive film, you can work quite a bit more from that film than the foma. You might even know that different kinds of film gives good results in different lighting and looks better for different motives and how to transer that difference from film to paper.

    If you are a landscape photographer with enough sales hire a full time copist from time to time. He will tell you that if you shoot iso100 instead of 400 he will deliver quite better copies for you. And when you have sent in your first iso 25 after standing there in the cold all night and got the 10′ by 4′ panorama back. The shuttering of your teeth turns to happy bliss.

    I never could afford a full time copist. But ill tell you what, when i first shot with a digital camera equiped with a sony exmore sensor, I knew that there was never any going back for me. I few days later i got one, and by switching cameras I had gotten back a lot of what is possible in my dark room with deasent film in digital camera. I didnt feel crippled every time i shot digital anymore and the lost half of what i regard as photography was suddensly back.

    However, if your work as a photographer ends with the jpg out of the camera a canon sensor will work as well as a sony one. Probobly if you just let the raw converter use one of the presets too, at least most of the time.

  • Jonathan

    I find this disappointing. However, I also don’t think it reflects the camera’s capabilities accurately (whether or not the test is accurate). I say this because I own a D300. From the images I’ve seen, this 7D Mark II has better color, better high ISO, and better DR at High ISO, than my camera. Now It may be that it’s DR may not be as good at Base ISO, but I would have to play with RAW files to know for sure. Either way, I would buy one of these in a heart beat if I were a Canon guy. Actually, if it were not for sharing equipment with another Nikon photographer, I would get one right now and switch brands.

  • Clint

    I bought the original 7D when it came out and loved everything about the camera except for the sensor….images just didn’t have that crisp punch to them. Too bad that al this time later the 7D2 doesn’t seem to be much of an improvement.

  • Vignes

    Looked through the chart at Dxomark (Canon EOS 7D Mark II vs Sony SLT
    Alpha 77 II vs Nikon D7100:

    – ISO, The measured ISO along the plot of D7100 and A77Mk2 are off compared to 7DMk2.

    – SNR 18%, the all-3 plots are almost the same at 100 ISO (7DMk2 starts from 100 ISO where else the D7100 and A77Mk2 starts before that).

    – DR, A77Mk2 and D7100 are good at low ISO but 7DMk2 plot crosses A77Mk2 at near 400 ISO and crosses D7100 at >6400 ISO.

    – Tonal range, almost the same from 100 ISO but 7DMk2 extends to 51K

    – Color sensitivity, 7DMk2 only crosses A77Mk2 at >800 ISO but D7100 leads at the way till sub 25K ISO.

    The 7DMk2 plots improves well above 400 ISO. It seems to do better once above 1600 ISO. These are practical useful ISO ranges. Looks like the designers were focusing on these ISO. For me, these are more useful ISO ranges than sub 400 ISO. Not sure why DXO recommended A77Mk2 unless they meant specific for sub 400 ISO usage. Now how many of us use Sub 400, 100 ISO for shots other than in controlled environment (lab, studio). Studio guys would be using MF/FF stuff.

    Use DXOmark to compare Canon EOS 7D Mark II versus Sony SLT Alpha 77 II versus Nikon D7100. This has more use than the overall %. the overall % looks at the highest score (sub 100 ISO).

  • Pawel

    You can check here, the quality of DR of the Canon&Nikon sensors. It seems, that they really have a problem with low ISO, but reach Nikon when at High ISO.,D600,D750

    PS: I selected mine d200 to compare it to d600 and d750 😀 you can freely choose on the right the camera you like.

  • Back to top