skip to Main Content

Olympus E-M5 gets DxOmarked. Best m43 senser ever made!


DxOmark finally published the full Olympus E-M5 sensor test (Click here to read it). And the conclusion couldn’t be any better: “These Olympus high-end compact hybrids achieve scores that approximate those of their APS-C rivals (such as the Sony NEX). Simply put, the OM-D E-M5 is equipped with the best sensor we have ever analyzed for a micro 4:3, and to date the closest to that of the Sony NEX.

Cameras like the Sony NEX-7 still have a small edge but than as usual you have to consider the lens quality too and the story will look even better for Olympus as we know that most E-mount lenses aren’t that good (with the only exception being the Zeiss 24mm lens). I had some fun trying the comparison tool and the E-M5 beats cameras like the Leica M9, Canon 7!

If you consider the quality of the sensor even the not spectacular new E-Pm2 and E-PL5 cameras do look much better than we initially thought they would be! So here are the store links to all three Olympus camera using the same 16 Megapixel Sony sensor:
E-M5 at Amazon, Adorama, B&H, Jessops, Amazon UK, Amazon Germany, Amazon France, Amazon Japan and Digitalrev.
E-PL5 at Amazon, Adorama, Bhphoto. In EU at Amazon Germany, Amazon UK, Amazon France,
E-PM2 at Amazon, Adorama, Bhphoto. In EU at Amazon Germany, Amazon UK, Amazon France,


E-M5 Case:
There is the official CS-36FBC Olympus case. A leather half case from Hong Kong. A full leather case with strap from Cina. Than we have half cases from two well known producers, Zelenpol, Kaza and Gariz.

E-M5 Batteries, and other things:
The Olympus GS-4 strap and the DSTE E-M5 batteries you have to use with an extra charger. There is an E-M5 car charger a new E-M5 LCD screen and a Front Cover+ Top Panel.

  • Jankoff

    True, but I’m disappointed.

    • jake

      dont be disappointed , who cares what this BS test says.

      it is always BS, DXO is biased , every rational peopple can figure it out.
      DXO even says Nikon plastic fantasy(d600) is better than the IQ180 from Phase One, who can trust this BS test.

      BTW, I test all my cameras myself and I believe the EM5 sensor is better than the NEX7 sensor.

      DXO is doomed , take a look at its best 10 camera ranking, almost all Nikon craps.

      • Jankoff

        DxOMark may be overating and downrating a bit but the problem in this case is that many people, including me, were led to believe – by the advertising campaign and the numerous rave reviews – in something almost impossible, viz., that the 43 sensor can easily beat APS-C. The lower ISO performance with respect to APS-C, as stated, is probably true. Let’s hope now that the not-so-high ISO performance plus the bad handling of the PENs and the EM5 will force Oly to rethink their strategy and admit what thousands of Oly owners have been telling them for a long time: that they shouldn’t have abandoned FT and that they must resume making E-650, or similar, with the new sensor and stabilization.

        • Bandit

          Couldn’t agree more.

        • Bart

          That is a silly thing to believe as it is technically absolute nonsense, and many people have been pointing out that this is the case, and that a larger sensor will always be somewhat better then a smaller one based on the same technology.

          Reviews? Some have been saying that to gain anything substantial above what the E-M5 sensor offers, you need to goto a much more expensive modern ‘full frame’ system, which still does not imply anywhere that the 4/3″ sized sensor is better then APS-C, rather, it implies that even the best APS-C sensors are not substantially better, and not worth upgrading to if you are an e-m5 user.

          I think it is a matter of what you want to believe, and how you then read what others say. One thing is for sure, ignore the advertisers and pay more attention to photographs.

        • Raist3d

          I am sorry but you are partially – at least – to blame for it. I kept saying the EM5 DID NOT match APS-C and I agree many reviews stated so. But you have to be skeptical enough to question the claim a little bit at least.

          I have shot the K-5 and EM5 side to side and I can tell you the K-5 sensor does have a notable advantage still over it, but that said the EM5 results are pretty good.

          • GuidoRTt

            @Raist my friend there are a group of Olympus fans who are so wrapped up in an Olympus comfort blanket that they have zero ability to look at any topic rationally. Kind of creepy actually. I have seen posts here claiming that the E-M5 which is barely a stop better than the GH2 and less at genuine ISO settings,is as good as the Nikon D800.Now I am not interested in FF and I love my E-M5 but I at least can accept that it is not that amazing

            • Raist3d

              +100 – exactly. And there isn’t any issues with that, the EM-5 sensor is great. Believe it or not my main deal breaker with the EM-5 for me personally is the ergonomics. I wouldn’t mind buying some of those fast primes and using it for my street night life work.

              I am eying the new Pen because it can fit in a pocket, but I am not thrilled they have that “Powershot wheel of death” on the back.

      • HiFinut

        come on guys, stop trashing DxOmarked for their reports. Their overall score maybe flawed but the dynamic, tonal and color range graphs are pretty accurate. The curves of the dynamic, tonal and color range for EM-5 are very close to the curves sets by the best apsc sensors cameras (NEX-7, NEX-5 and D3200)and slightly surpasses the canon apsc cameras except in the lowest ISO setting. If Olympus could improve on their lowest ISO setting (below 100ISO)dynamic range, it will equally the best apsc overall scores.

        • Raist3d

          The amazing part is the EM-5 results are actually phenomenal, but people trash Dxo mark because besides the fact they don’t understand it, it didn’t put EM5 apparently at the top of any other camera ever made.

  • Ricky

    I hope the new PENs have the exact sensor performance.

  • metalaryeh

    Taking in account the new stabilization, the iso score can be looked at slightly differently.

    • Unless you shot moving subjects…

      • David

        Yes IS helps a lot when panning for moving objects!!!

        You can’t do that always with a tripod for various angles, and IS in critical. Use your camera instead of making silly comments.

      • Alan

        I love the look of some subject movement, but with rare exceptions (like panning), don’t like the image deterioration caused by camera movement.

  • caver3d

    And why does it have to always be stated that it’s the best m43 ever made? The E-M5 is now competitive with APS-C, it’s not just a great m43 camera. These “experts” (and that includes dpreview) need to start giving the OM-D series (and there will be more of these Oly cameras) its due. The APS-C bias continues. Ridiculous.

    • MJr

      “performs like a decent apsc”. Nah that doesn’t sounds better than the best m43 sensor everrr !

      • caver3d

        You miss the point.

        • MJr

          No you do :)

    • Lighting

      It doesn’t comparable with the D3200 APS-C sensor. The small sensor will away trail behind.

      • Ganec

        it is: look to the measurements and switch to “screen” (it is “per-pixel” result)
        You can see then that D3200 is little behind at higher ISOs

        • jake

          yeah ,and in real life low light performance , the EM5 is not a bit better but a lot better than the Nikon camera because to get the same dof , you must stop one full stop more with the D3200 or NEX7 or any APS-C or whatever Nikon call it,and thus, you must need to shoot at full stop higher ISO with the D3200 than with the EM5.

          so, unless you always shoot almost wide open or accept ridiculously shallow almost uselss dof at f1.4 or like that with you FF or whatever Nikon calls it, you wont get so-called low light advantage of FF.

          • Anonytrackball

            Are we confusing 135 attributes and APSC attributes?

    • jake

      exactly, it is not an honest test and these days less and less numbers of people find it interesting or even usable.

      DXO is doomed ,one of a few it’s main sponsors is Nikon.

      • Digifan

        I’ve always said to take DxO’s results with a big bag of salt.
        The E-M5 was already better than the Nikon and Canon stuff in real life. H#ll even the older Pens were a match but those just at lower ISO’s.
        Measurements and real life is worlds apart. Outside there are no controlled conditions and you need to make do with the available light, wind etc.
        m43 is unbeatable when you don’t have an assistant to carry your equippement (well I never shove my stuff to anyone else but me anyway, I’m not a slave driver).
        My E-5 and E-1 haven’t seen much light of day since the E-M5 arrived.
        And with the new lenses from Pana and Oly it’s getting practically impossible for Sony Canon and Nikon to keep up.

    • Mike

      The bias towards APS-C is simply because that’s what most people shoot. People getting serious about photography and leaving point and shoot will skip over m4/3 and yet can’t pony up for FF or larger. They settle with APS-C.

      To me APS-C offers no advantages and is simply too middle of the road. Equipment is too large for a sensor less than 1/2 area of 35mm. That’s why I never bought in. I waited ’til the IQ and performance from m4/3 reached a high level and use FF when I need better still.

  • bluebox10

    m43 user have always screamed that DxO is no good. Now the E-M5 gets high marks, and we’ll see the crowds are ready to embrace DxO.

    Time for the Sony guys to scream that DxO is no good… :-)

    I still take the DxO stuff with a grain of salt…

    • Kyle

      That’s true hahaha

    • kenneth

      Sony rules…again

      • Esa Tuunanen

        Indeed right now Sony is for sensors what Kodak was for film.

        • The Real Stig

          Fuji were/are better than Kodak for film.

    • Chris K

      Nope. DxO is still a worthless synthetic benchmark designed for camera collectors.

      • JimD


    • jake

      well, you might want to study sensor politics behind this shitty test and come back to tell us what you think about this crappy test and this really Nikon biased tiny software vender from France.

      if it is from an American company like Adobe , I might trust it but DXO?

      anyway, no thanks DXO, I do my own test with all my cameras.

      • Bart

        Testing your own cameras, and looking at in-depth tests are good ideas, going for an overall score of one specific test is not good.

        I don’t see what France or the USA have to do with this, like American companies never lie..

  • G_C

    beats the M9 and 1D3 but loses to the K-r haha fun!


    lets enjoy take photo

    • avds

      Interesting catch. The K-r is less than 1% better than the EM5 for both color depth and dynamic range while almost 10% worse for ISO, and yet its overall rating is 1% better :) Based on this, I’d say their weightings for the overall score don’t seem useful at all (as with all synthetic benchmarks, perhaps).

      • Duarte Bruno

        Their final score is as useful to tell how a football performed over a season has their final score, ie: shit.

        DxOMark is the best tool around to tell which are the weak and strong points of a sensor and to let you know where your camera performs.

        The only problem with DxOMark is that a lot of people only look at their final number because they’re too mentally challenged to dig deeper.

        • jim


        • jake

          >The only problem with DxOMark is that a lot of people only look at their final number because they’re too mentally challenged to dig deeper.

          not mentally challenged to dig it deeper but it is a important point.

          most people just see final scores at print mode not screen mode.

          but even after that has been taken into serious consideration , DXO is still doomed and the fact it is sponsored by Nikon never changes.
          so, to me it is a worthless test.


    Lol…DXO takes months to tell us what we’ve already known for months. I just had to look at the photos produced by this amazing camera.

    Why did they wait till now to publish?? Seems weird…

    • julesvern

      because all of you forcing DXO to show the result.. :D

    • elflord

      just some speculation — they probably wanted to make sure they were right before publishing results that said that the manufacturer ISO was off by more than a stop.

    • Raist3d

      It’s not weird. They explained when they first did the test it looked like the results were off, so they took their time and due diligence to make sure there were no mistakes on their part.

      Oh yes, this is part of the methodology (some people think that the fact this happened points to a flawed methodology- if they didn’t double check then they would indeed have such flaw).

      As for what “you already knew” look around- apparently some are surprised that the EM5 does not match APS-C.

  • Jankoff

    How about buying a EM5 when D600, twice the price, has rather more than twice higher ISO score?

  • ISO and IS are two totally different things. IS won’t freeze a running athlete in it’s tracks. High ISO will. IS only counters camera instability. And if you use a pod, it’s pretty much pointless. I’d take ISO performance over high ISO any day.

  • Of course its the best, its made by SONY! LOL

  • Mike

    No surprise here. I’ve been tremendously pleased with the IQ from my OM-D.

  • Steve

    Looks like the E-M5 actually has a measured base ISO of 100.

    • Marck

      that’s a good point! They measured from ISO100 to ISO12800, shouldn’t everything be moved forward by one stop, thus giving a better result to the E-M5’s mark?

      • Irresistable

        Wait a minute.

        The iso base of E-M5 is at ISO200. How possibly did they measure at ISO100?

        • Steve

          DXO is saying the ISO200 on the E-M5 is actually ISO100. All of the charts at DXO are based on measured ISO, not manufacturer ISO.

          • ED

            When you are metering using a light meter in a studio do you use 100 or 200??

            Another (unrelated) question lol Does anyone know if you can trigger strobes and studio flash with the OMD without the use of additional triggers such as pocketwizards etc..???

      • Steve

        Yes the results would be better if you believe the Camera ISOs are correct.

        • David

          I actually just calculated these from the Dpreview article, because I thought DXO was wrong.

          But DXO is not that wrong. You can get the Shutter speed values, fstop and ISO from each image in Dpreview. My E3 is rated one of the very few to list ISO correctly ie 100 = 100.
          For the E3 shutter speeds were:
          100, 200, 400, 800, 1600, 3200
          40, 80, 160, 320, 640, 1250.
          Interestingly for the EM5:
          200, 400, 800, 1600, 3200, 6400, 12800
          50, 100, 200, 400, 800, 1600, 3200
          These means the actual Iso for the EM5 is 2/3rds off.
          200, 400, 800, 1600, 3200, 6400, 12800
          125, 250, 500, 1000, 2000, 4000, 8000

          Both cameras on Dpreview were shot with 50mm F2 lens at F 6.3. So this was easy to calculate as only the camera was different.

          • Not so fast!
            DPReview’s lighting setup has changed over the years, and definitely between the E3 and E-M5.

          • Marck

            I did this kind of testing at home by myself with an E-M5 and an E-5, using the same lens (Olympus 12-60 2.8-4).
            I shot the exact same scene in “A” mode, letting the camera decide the shutter speed. The two cameras shot at the exact same shutter speeds and the RAW pictures show the same amount of light.

            So my direct expenrience tells me that E-M5 and E-5 have the same exact ISO behaviour.

            • David

              Thats very interesting!

              I heard from users of the E5, that it under reports by 1/3 a stop compared to the E3. This was tested as you say, and with users shooting Manual and needing to push 1/3 EV up to equalize exposures between shots.

              This is good news as I want the Higher ISO to get faster shutter speeds than my E3. If only 1/3 stop off then this is very good.

              SO in this case DXO would be Wrong!

    • photomic

      ..that would explain, why i always (mostlyx) set the correction “-1,5”, to get a not so bright picture with more details in the shadows..

      i think, olympus wanted to stretch the “high iso” performance, instead of starting the realistic value of 100 iso.. hasn`t the rx 1 ISO 100 – ISO 25.600??

  • safaridon

    Why oh why are you showing a comparison with the G3 and not the G5 which at 11.6 evs is only 7% less than that of the EM5?

    Seems there is way too much hype on dynamic range which is only one aspect of IQ when the differences are numerically relatively small in fact? First we had an emphasis on MPs, then noise, then high ISO, sensor size, and now evs. With the advent of very good HDR programs and fast performance evs should be less of an issue. However as manufactures and reviewers encourage the hype people are tempted to think the latest cameras are “much better” than what they currently use and enjoy.

    • Llamaman

      Perhaps because (as many suspected), dXo wrote the report weeks ago before the G5 was released, and have only put it up now the PEN annoucement made their delay look ridiculous.

    • Neonart

      12.3 EVs to 11.6 is not just 7%. Each EV is a big step, not just a numerical digit, much like F Stops or decibels.

      f8 is 2x (100%) brighter that f11, but 11 is only 27% less than 8.

      The highest dynamic range DxO has ever tested is only like 14.4, so it’s not like a 0-100 thing.

    • Will

      Dynamic range is not a hyped point my friend, it’s the main determining factor in IQ for “large” sensors. Why do you think people still use Full Frame, or Medium Format or Large Format? all for the DoF? Nearly all large format photography is landscape with as much DoF as possible, it’s the dynamic range they’re interested in.

  • mooboy

    Was there any mention of what’s so surprising about it and why took so long to be reviewed?

  • Adriaantie

    Low raking for tiny sensor. Seems fair. And no m43 can not compete with modern apsc.

    • You are so obsessed with size that one starts wondering about your own shortcomings. :-P

    • Gianluca

      in fact it’s better than all canon aps-c camera sensors…

    • Jan

      I think you cannot directly compare DxO ISO results of sensors with different sizes. I guess you would have to multiply the result with the square of the sensor crop factor (4 for m43, 2.25 for APS-C, 2.56 for Canon APS-C). Otherwise the old full frame sensors could not be better than the current smaller sensors.

      • Huh? The DxO print normalizes for output. They are the signal to noise ratio of the final image. You don’t multiply anything by anything. Yes, the E-M5 is better at high ISO than older APS-C and some full frame sensors. Having owned a lot of them, I can say that this is absolutely true, and no multiplying need be done.

      • DXO downsamples all images to 8 mpix before analyse. I think they have a quality assured process and results are relevant.
        One thing: I do not know how many DXO-points it takes to see a difference viewing an A4-size print.
        Second thing: All cameras we discuss here are capable of excellent photos! (maybe we as photographers are not all at that level – no camera is better than its user :-) )

    • GreyOwl

      Is a Leica S sensor large enough for you, or perhaps something with even more (expensive) area? :-)

    • Richard

      Tiny EM-5 sensor DR 12.3 EV, BIG Nikon D700 sensor DR 12.2 EV, oups!

    • The Real Stig

      Low IQ of tiny, insecure troll.

    • JF

      Lol, thank you for all your funny comments !

  • Can’t people just enjoy taking photos without these measurements? Do you need DxO to tell you how good or bad your images are?

    • Irresistable

      DXOMARK is considered as the worldclass reference for sensor performance database. We all have to believe in their test.

      • JimD

        And Politicians tell the truth?
        If the camera take the pictures you want then the photographer is good.
        If the camera does not take the pictures you want then, we all know the answer buy a more expensive camera. The psychology will make the camera great, until a more expensive one comes along.
        DxO has nothing to do with world reality or using cameras, its all ‘eye’ or phsycology, the marketing people sell to the latter, the former, use feel and result, and a little ‘the last one was good’. DxO what is that, a washing powder?

  • Ab

    I never put too much weight on DxO and still dont. With all the images out there, and Exif data available real photographers have far better tools to evaluate a camera.


  • The results look realistic to me. No-one expected it to be the best thing on there. But it’s certainly better than most APSC cameras, only the very best APS-C cameras beat it.

    Happy that it’s all done now anyway. It’s nice to have a verified reference to the performance to point people who think m4/3 will always be crap too. I hope the PENs match it, as with their price range it will embarrass a lot of Canon owners.

  • You guys should never look at the overall score just by itself… look at the graphic charts… in SNR and Tonal Range it has half of stop of advantage over GH1 and G5, it has around 1 stop to 1/2 stop of advantage on DR and Color Sensivity over the entire ISO range.

    Though, D3200 beats E-M5 (but not for much) while costing a lot less… as for me, I’m keeping my ‘ancient’ GH1 …

    Remember that this is about RAW performance… there is a lot to account in a camera aside from raw image quality… Now … if they made a m43 camera with a proper sensor (E-M5 looks proper to me) and reasonably priced (around 500€ or 700€ with a nice kit lens)…

    • Jorginho

      EPM2 or EPl5 you mean? They are just around the corner.

    • Casadilla

      See E-PM2/PL5, Olympus. :)

    • JF

      I switched from GH1 to E-M5 and I can tell you it is really worth it. The main problem of GH1 (at least the sample I have) is banding which dramatically reduces the performances and the possibility of DR and noise…E-M5 has a more random noise. In lightroom I can push the shadows to +100 while with GH1 at +40 you start seeing the banding…same for exposure…E-M5 is not just better on paper and in Dxo results, you see it immediately in real life !

  • ArKersaint

    Come on Adriantee, stop this maniac trolling habit on our nice site and come back to your big boys community !

  • Dummy00001

    > Cameras like the Sony NEX-7 still have a small edge […]

    To me that’s another good example of uselessness of the DxO. Because no way I would rate IQ of E-M5/NEX-5N below the NEX-7 (and all other cams based on the sensor). The NEX-5N sensor to my eyes is clearly superior to the NEX-7’s.

    • Esa Tuunanen

      DxO’s method is biased to more lower quality pixels being better than less higher quality pixels because it’s blind to decrease of actual image details and quality per pixel.
      “Our noise and dynamic range chart, is supplied by DXO and it’s made up of a series of pure optical glass filters with a range of neutral densities mounted in a thick plastic plate. The chart is held on a purpose-made light box that produces diffuse, uniform illumination.

      Using a glass target avoids the lighting problems and inconsistencies that are often associated with printed charts for assessing dynamic range. The glass also removes the issue of media texture that can create issues for accurate noise measurement with printed targets.”

      So it doesn’t matter any if increasing noise per pixel lowers details in actual image because there’s no details to start with.
      As a matter of fact Bayer sensor’s demosaicing makes higher noise in single sensor pixel to have higher effect also in surrounding pixels of resulting RGB image. That’s no doubt the reason why higher ISO noise in images tends to become larger and blotchier instead of being fine grained.

    • Frog

      Agreed, the Nex 5n is a much better camera the the Nex 7 for imaging and handling. That is what I found out last summer. The new Nex 6 (well at least for me, I like a viewfinder) should be much in all respects then both the Nex 7 and Nex 5n. Sometimes more money for a camera is not better.

      Dx0 testing is just a measurement. Not a full test. Though the new Pens utilizing the OM-D M5 sensor will be a great image for the price and size.

    • You dont understand the standardized test.

      • JimD

        It would seem that he is not alone. DxO as well as many in the rest of the world do not understand the test. BUT the results are hallowed and not to be argued with! NUTS.
        God gave you eyes. Use them to evaluate real photos.
        The world does not need every little thing to be analysed and qualified within certain bands (or totally misused such as the apsc boys claiming magical 135 properties for their cameras, when they are closer to 4/3 than 135).
        The world is bigger than that.

        Quick vote.
        The ‘eyes’ have it

  • Jens

    Where do I find the results for resolution? When I compare for example the D700 with D800 (and the OM-D) most results of D700 and D800 are equal (SNR, e.g.) except of the higher start value at ISO min. Both cameras have the same sensor size, but different resolution (3x). So the new D800 achieves the same high ISO quality, but with 3 times the details. Comparing the OM-D with the PEN we have the same thing. Higher res and higher high iso, but res is not mentioned, is it?

    I am looking for something like linepairs. (“Linienpaare”)

    • Javier

      Well, resolution is highly dependant on the lens.
      DR, color depth, Noise; these all depend on the sensor and have nothing to do with resolution.

      By normalizing all images to the same printing size you are comparing how overall noise, DR, etc will look on the same printing media, not just pixel peeping.

    • Resolution is taken into account when viewing the ‘print’ comparison. This compares noise for each camera at the same print size. The ‘Screen’ comparison shows differences visible at 100% on a screen (so at the pixel level). ‘Print’ will give you the actual difference with all factors considered.

    • Bronica

      It depends on the lens. In this amazing test the E-PL2 beats with its 12MP Sensor and the famous M Zuiko 45mm the Nikon D7000 (it beats also a D300:

  • Aim9X

    Thanks you Sony for this great sensor!

  • MikeH


    Why are you comparing this against the G3? The G5 has been out for some time now and costs less than the EM5. If you are going to have the G3 in there you might as well compare it against some of the old PENs while you are at it.

  • kesztio

    …and the GH3 seems to be at least on par with E-M5.

    Gongratulations, Micro Four Thirds!

    • hlbt

      Where did you hear that? Pana folks giving interviews at Photokina suggest otherwise.

  • Javier

    For me is incredible the way ISO is off in the E-M5, almost 1EV off

    Manufacturer ISO: 200 Measured iso: 107
    Manufacturer ISO: 400 Measured iso: 214
    Manufacturer ISO: 800 Measured iso: 394
    Manufacturer ISO:1600 Measured iso: 782

    In isos over 400 it is over 1EV off, worst I have ever seen. That is why iso 1600 looks clean, because it is not even iso 800

    Anyway, seems a good 4/3 sensor but I find this iso thing very anoying

    If you compare this to the GF3 or GF5 which are very well calibrated it would show very easy

    Same shot in same lighting conditions with same exposure time, same aperture and same iso the E-M5 will always be 1EV underexpossed

    For me this lying to the customers; I guess is better to say you have an amaizing iso 1600 than to tell the truth and say it is just iso 800

    • Eric

      yeah i always had a suspicion that this was happening too, my e-m5 raws were always darker than my canon raws. and if you saved a pic in raw + jpeg mode, you’ll see that they just take a dark raw and up the expose in the jpeg.

      • Steve

        I guess is how they can preserve the highlights in JPEG. Seems reasonable for what you gain.

        • Javier

          This has nothing to do with highlights, one thing is DR and other ISO. The thing is they they labeled Isos 1EV above they real value ( it reads ISO 400 when should read ISO 200)

          • Steve

            It does so make sense to do this way. By underexposing the RAW you help to preserve highlights. When creating the JPEG Oly just increases the exposure for the shadows and midtones by 1EV, while leaving the highlights as is.

            • Agent00soul

              Right. This is the only way to roll off the highlights, to avoid abrupt clipping. It has nothing to do with cheating.

    • kesztio

      I think you are absolutely wrong.

      Measured ISO just means that thew RAW ISO is lower, but the JPEG engine will make some exposure compensation (brightening the whole image) to match the correct exposure, thus by pushing up the darker zones which does mean additional noise. So technically the ISO1600 from the E-M5 IS REALLY ISO1600.

    • Esa Tuunanen

      And for me that means you’re an idiot… and surely at least not photographer but someone who prints out sheets of numeric RAW data instead of image.

      Photographically meaningfull sensitivity defined in international standard is certain brightness of image produced when shot at certain fixed combination of aperture and shutter speed in certain illumination.
      Which is exactly what every PHOTOGRAPHER has to keep in mind: Every stop lower ISO setting means need to increase either aperture or exposure time by one stop and every stop higher ISO needs decrease of either aperture or exposure time by one stop to produce correctly exposed image. And change in illumination causes need to change at least one of those three things.
      In manual camera era they even had tables available which listed value combinations of these three variables for producing correctly exposed (=brightness) image for certain level of illumination.

      Digital sensor itself doesn’t have any sensitivity, except maybe some ratio between outputted signal level of pixel per certain amount of photons received.
      That signal is then amplified, digitized (RAW data comes from this) and processed in various stages to get correctly exposed (brightness) RGB image for set ISO.
      Which is basically same what happened in film era: That certain sensitivity film needed correct development process to produce image with correct brightness.

      Numeric raw data really doesn’t matter for as long as image has correct brightness for used combination of ISO, aperture and exposure time.
      Sensor being run at lower “sensitivity” means that numeric data simply needs to be processed so that brightness of resulting image is correct.
      And that isn’t any miracle advance giving cheat: While it means there’s more dynamic range for highlights its downside is that also noise gets pushed up!

      Similar was done already in film era:
      For example ISO100 film frame was shot at aperture and shutter speed combination correct for ISO200 film and frame was then “pushed” in development to produce image with brightness equivalent of ISO200. Method was used for example if you had to take shot with faster shutter speed but had slower film in camera.
      Also there was opposite process of pulling exposure downwards during development, which sacrificed highlights. (both had their downsides in image quality)

  • Neut

    Exactly as expected.

    Since DPreview refuses to correct for it, we might as well get use to ISO inflation.

    • Steve

      Dpreview noticed only a 1/3 stop difference in exposure but they only test for JPEG where is seems Olympus pushes the RAW. When comparing RAW it makes sense to always compare 1 stop higher with the E-M5.

  • SLOtographer

    I don’t look at the score. That’s an arbitrary rating system. I look at the measurements that are relevant to my style of photography. Some people want DR at base ISO (landscapes). Others want a good signal to noise ratio a higher ISOs (people, event). Do your own due diligence and pick the right tool based on the available data and features.

    • I did earn money over the years with pictures of my paintings. Only sold a handful repro, but several originals with the help of repros.. In that regard, I really should aim for DR at base ISO as you say. If at all, move to Foveon stuff.

      But really, what I have is good enough not to start splitting hairs. I still like to read those stuff, sometimes too much for what it’s worth…

      • David

        I own and use the SD14. I would say if you are photographing your art work, then don’t use a foveon sensor, as the colors are off. And off by a lot. The sigmas are really good for true color sharpness. The mythical 3D effect, you will read about. But for true color accuracy you will need to look elsewhere. My Kodak SLR/n also has different colours than reality, but I do like them better. The shift for it is not that dramatic, just like the old Kodak film.

  • Dave

    Wow, doesn’t even come close the the Fujifilm X100, but it definitely kicks the arse out of the rest of the advanced point & shoots. Probably not the Sony RX100, though. Olympus is getting there, slowly.

    • avds

      “Doesn’t even come close”, uh? Have an eye check. I’m afraid you are getting there, slowly.

    • Ben

      What you smokin’?|0/%28brand%29/Olympus/%28appareil2%29/695|0/%28brand2%29/Fujifilm

      E-M5 / X100
      Overall Score: 71 / 73
      Color Depth: 22.8 / 22.9
      DR: 12.3 / 12.4
      ISO: 826 / 1001

      That seems to be VERY close.

      • Dave

        Like I said, not bad… especially for a small sensor camera. I was more focused on the ISO difference than anything else, actually. Plus, the X100 is simply a superior camera.

        • The Real Stig


          There are times when you just can’t mince words.

        • The difference in ISO is about 1/4 of a stop. Not enough to make a difference that you’d notice outside a testing lab.

          • That, and also easily made up by faster m43 glass, of which there is considerable choice for comparable FOV’s, versus Fuji’s only fixed f/2.

    • Nawaf

      I’ll have what you’re smoking!

      BTW it’s not all about the DxO scores. Other factors should be taken into consideration too.

    • the RX100 is crippled by its slow fixed zoom lens. That’s the price you pay for having a compact form factor AND a large sensor AND a zoom lens. Since you seem interested in low light performance — the OM-D with the 45mm f/1.8 trounces the RX100 which is f/4.9 at the long end — that’s an extra 3 stops or so of light

  • PannyMan

    Hey i have a question, how long do you think those 3.000 special edition black lens will last? anywhere near April of 2013?

  • Compared to the RX100, same level on color and DR but EM-5 much better on noise. For sure these sensors are close relatives.
    For some this is not as close to the polar star as they whished for, but DXO clearly shows the EM-5 is an excellent camera. It is so close to APS-C IQ that normally you will not be able to tell the difference by looking at images. Add the new macro and the 75mm, 45mm and you have a very strong alternative to NEX-6/7 and others.

  • Jerome LaPlume

    I love what E-m5 produce in terms of images, i like it more than Aps-c rivals(except Fuji), i have choose Olympus for that.
    I don’t care of numbers(positive or negative), because they don’t tell me nothing about CHARACTER of equipment.
    Photography isn’t a science, every value become meaningless against FEELINGS.
    It’s my opinion, luckily the camera industries offers solutions almost for everyone.

  • Dobbler

    So where are the ‘true ISO’ numbers for the G5? It sounds like from reading the comments here that there isn’t much difference in the noise levels given the same exposure parameters between the G5 and EM-5 when you would normalize the ISO to the _standard_.

  • Peet

    Again, and even the test looks good for the OM-D, this test is not the real life. Forget the ranking, make good pictures with your phenomenalkamera and your great lenses.
    Of course you like your OM-D, because it is a superb kamera.
    No DxOquark test is interesting, just look at the ranking.
    And no one of you can print or see the real quality of am OM-D or D700 because there is no chance to display it. Not even a 100% crop at your tiny display.

  • HappyVan

    Terrible!!!! I’m tearing out my hair.

    After seeing Robin Wong’s photos, I thought that the OMD was a miracle!

    Instead, it is only equivalent to the Nikon d300 (2007) which can be bought today for $600.

    • no, take a look at the curves. The OM-D has more dynamic range than the D300

  • dg

    Congratulations Olympus!

    • ashwaniman

      It should be congratulations Sony for that sensor

      • Digifan

        Yeah and for Olympus to recognise that and choose the best.

  • JP

    I love it, finally numbers to back up the images that clearly are as good as the best APS-C cameras and better than all Panasonic MFT cameras. The OM-D with prime lenses attached can’t be fucked with.

  • moof

    You’ll whine if they don’t do it. You’ll whine if they do. You’ll whine if they say the sensor is great. You’ll whine if they say the sensor is bad.

  • a

    these are great numbers, and anyone interested in buying into m43 now should be extremely excited about the epm2 and epl5. arguably, the best combination of size/iq/price on the milc market.

    • happy1

      “The OM-D with prime lenses attached can’t be fucked with.” The comment to end all comments here. lol

  • Honestly speaking, it is not surprised to me. I have D3200, NEX7, and EM5. D3200 and NEX7 are a bit better in terms of raw outputs.

    • Digifan

      I totally Disagree. The D3200 and Nex7 don’t come close to the E-M5 in real life.

  • Anonymous

    People never seem to get that the ISO rating is entirely left to manufacturer nowadays, and that even DxO confesses of using only ONE of the five definitions available.

    But what struck me is the comparison with the E-P3, and its sensor. By definition base ISO is not a factor, and yet the E-M5 is better not only in DR (2 stops!) but also in colour depth.

    Considering tht I am still very happy with my E-P2 and learning it, I tell myself that sometimes in the future I will get an E-M5 and be even more happy. it is what builds trust in a brand. Good image ideas can come to you even with a toy camera, but it’s nice to have a proof of reliability in a brand.

    DxO also insists that the gap with APS is closed, one by one the old argument against 4/3 crumble. It is as if the format was getting adult. So to the naysayers we can just say that we have a top perfomer for a fraction of the weight.

    Add the second part of the equation – some v. good lenses for resolution – and m4/3 becomes unassailable. It won’t replace a meme like brand recognition, but at least we can spread the word.

  • Bob S.

    Well….it makes sense. What’s wrong with these numbers? Do you expect something higher than NEX5 or NEX7? It’s simply impossible.

  • Irresistable

    This test probably implies that the best chance of m4/3 sensor is getting near to APS-C. It cannot be superior to bigger sensor.

    • lazybeans

      It may not be superior to a larger aps-c sensor, but it comes at least close enough that it won’t matter anymore. And yes, it does have plenty shallow DOF (if that’s your thing).

    • JimD

      They are both good enough to provide the quality images that 99.9+% of people want. Most want to show on web sites and such, for those, any pretence of anything better than m43 or apsc is just that pretence In print the same 99.9+% could not tell if an A3 photo was taken with a M43, APSC or 135. If comparison A3 shots were shown side by side a few may see a difference in the image but still not tell which was which. The vast majority would comment on the shade of pink in the dress or the brown of the dry grass, or the ability to see strands of hair and leave it at that.

      The difference may be measured using a scientific methodology. But who cares, when the visual differences are beyond normal eye view, understanding and satisfaction.

    • In terms of ISO rating Sony and Olympus must have really outdone themselves with this one, as it only seems natural to expect a 4/3 sensor to be about 1 stop behind APS-C. But not so for the dynamic range, which must not be affected by sensor size.

  • come on pana

    Nikon D800 beats them

    • E-1

      In weight?

      • Richard

        And price?

  • Anonymous

    I am really tired ho hear the argument that the ISO reading is flawed. It is right by DXO standard *own* definition, which has to do with the correct brightness achieved at base ISO. Then you read the SNR curve at the various ISO readings and can compare it with other camera.

    In this respect, compared tothe FF Leica the E-M5 is quite impressive, but this had been already noted by other bloggers.

    I also did a comparison with the E-P2 (12 Mpx sensor) and the results are exactly the same of what Pekka Potka had discovered months ago. So there is no misunderstanding at all.

    • JimD

      And by my *own* standard I use an old fashioned thing called an eyes. These have been tested and found to be fine if I use some shaped glass in front of them. This is fine as it something I can see making a difference.

      DxO on the other hand trys to tell me what my eyes are seeing. I don’t need them to tell me this I can see it. I don’t want to know that something out of my eyesight range (which is great) is measured as good, bad whatever.
      Show me a picture and I will tell you if I like it. DxO will not tell me if the expression on the subjects face is magical, or if the light through the hair is just right or the trees on the left balance the river twisting on the right, with the bird just leading the eye into the shot.
      What has DxO got to do with photography? If you like the camera and it does the above (or whatever) for you its the right camera. If it does not then are you the photographer you thought you were? Of course you can buy another camera, but that will be done on feel and seeing results from others mostly. DxO is highly unlikely to enter into it.

  • amalric

    I am really tired ho hear the argument that the ISO reading is flawed. It is right by DXO standard *own* definition, which has to do with the correct brightness achieved at base ISO. Then you read the SNR curve at the various ISO readings and can compare it with other camera.

    In this respect, compared tothe FF Leica the E-M5 is quite impressive, but this had been already noted by other bloggers.

    I also did a comparison with the E-P2 (12 Mpx sensor) and the results are exactly the same of what Pekka Potka had discovered months ago. So there is no misunderstanding at all.

  • Pixnat

    For all concerned by this pseudo-ISO issue, here’s an excellent explanation by Richard Butler of Dpreview, enjoy :

  • These numbers don’t excite me. It’s like looking at 24MP sony jpeg images, number great but real image output not so. And look at sigma DP merrill. Numbers not great but those images (at base iso) ultra stunning!

    What excites me is looking at images of DPreview coverage of photokina with E-PL5 and RX100, I think both cameras are superbly good. RX100 is the one a shocks me over and over and for E-PL5, i would say I’ve not seen any m43 images that can be so clean, detailed and sharp at those iso settings.

  • JimD

    7 Hours and waiting moderation. That means all the other posts were made before the subject was put up.
    What is going on.

    • Miroslav

      “7 Hours and waiting moderation. What is going on.”

      Admin has to sleep sometimes ;).

      • admin


        The server guys improved the safety of the 43rumros website and this seem to affect the comment system. It puts most comments into moderation. I will try to solve this next week. This week I am full at work and having a big exposition.

  • Lavinso

    I have been shooting with Nikon gear for a long time. Since the D700 came out it has been my main camera. I didn’t get E-M5 expecting better results than the D700, but I prefer the E-M5 image quality over the the D700 in many situations. Of course it can’t beat the low ISO performance or the shallow DOF of the D700, but it’s still a wonderful camera and I use it for shooting everything now.

    I am not surprised by the numbers and don’t even care that much about the DxOmark results since I can see the quality from the thousands of images shot under different conditions I have with me.

    • Dano

      I’d have to agree here. I also have a D700 but I prefer the EM-5 hands down over the D700 all day long. While all of you drool over DxO numbers, I’ll be out shooting some photos. It’s a pretty sad state in the camera world these days when people feel the need to argue over tech specs…

  • Miroslav

    Excellent. Surpasses Canon APS-C, comes close to APS-C cameras with Sony sensors. Olympus usually extracts a bit more from a sensor with every new camera, so new PENs may have even better results.

  • JF

    Finally !

  • JF

    Now, everybody should admit that it is a sony sensor:
    The cuves are pretty close and I think the slightly different high iso peromances comes from smaller pixel pitch of E-M5

  • JF

    I will say it again for those who don’t understand: FORGET DXO SCORES it’s useless and LOOK THE CURVES !! And whatever some can say, I think Dxo measurments are accurate…Come on, E-M5 is almsot as good as Nex-5N which has a wider pixel pitch what did you expect more ??

  • Narkissos

    So regarding to this ISOmania, can we all under expose an image using our camera and push up the EV in Lightroom to get better DR?

  • Mike

    In the light of these excellent scores of the OM-D sensor, it becomes clear to me why Canon and Nikon are suddenly pushing these “cheaper” FF cameras: the end of FF is coming.

    If the OM-D (and upcoming PEN) sensors can score this high (very close to APS-C), it is pretty obvious that the upcoming round of APS-C sensors will be very close to FF. I’m looking at the D7000 and D300 replacements.

    It is clear that in the coming years all of these sensors (1″ to FF) will become more than good enough, and the decision between these systems will be based on ergonomics and lens selection, rather than sensor size.

    I believe APS-C DSLRs occupy a sweet spot for ergonomics vs. acceptable size, and I believe m4/3 mirrorless occupies a sweet spot for compact size vs. acceptable ergonomics.

    • JP

      The D300 and D7000 are not as good as the EM-5. They are cheap compared to an EM-5. Have you used the good MFT lenses? Obviously not, they crush any DX garbage. Learn.

      • Matt

        What rubbish are you spouting? The latest and greatest m43 is only just catching 2 year old APS-C sensors and not even matching them. These results are predictably in line with expectations, they produce similar results to APS-C cameras that have SLT mirrors blocking 30% of total light to the sensor made by the same sensor manufacturer, they have similar results to a 1″ sensor made by the same manufacturer excepting that it has 1 stop better ISO performance which is expected given the larger surface area.

        ‘Crush’ is a very strong term, now you can discuss the relative merits of a particular brands lenses all you like, but this is an article about one organizations technical finding on a particular cameras sensor technology, which is completely line ball with expectation given the size of the sensor and technological maturity.

        It’s a good camera, why not go out and enjoy it and its remarkable technical merits.

        • JP

          Long comments are rubbish, your mother is rubbish, get to your point. Oh wait, you don’t have one. The EM-5 plus good glass defuckingstroys the D7000, not even a contest. You rubbish bama.

      • Raist3d

        You are delusional if you think the D7000 – sensor at least – performs worse than the EM5’s.

        • JP

          You are clearly delusional, read my comment. I did refer to the lenses, crush.

  • Luckily, I do not buy my gear based on DxO ratings. I am not slave to their ratings.

    I own both the E-M5 and the GF1 and guess what, the GF1 is still the same fun to operate and with a good lens, it takes terrific images which can measure up to the images I take with the E-M5 in many conditions.

    • Raist3d

      And that’s what Dxo confirmed you could do. Unfortunately apparently for some if Dxo didn’t put the EM-5 – Olympus latest darling child – at better than anything made on Earth, then Dxo s*cks, even though DXo ranked the Em5 phenomenally well.

  • IHUR

    Bah,, EM-5 is equal on both color depth and DR with Sony RX100

    And an almost 1 year older Sony NEX-5N Camera is way superior!

    But then what are those cameras compare to weathershielded EM-5.

    • JP

      The Nex-5 has a great sensor, yes, but it is built like a piece of crap, looks like a piece of crap and works like a piece of crap. Oh and the lenses are utter garbage. E-M5 plus lenses cannot be fucked with, learn.

Back To Top

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

What are Cookies?
A cookie is a small file of letters and numbers that is stored in a temporary location on your computer to allow our website to distinguish you from other users of the website. If you don't want to accept cookies, you'll still be able to browse the site and use it for research purposes. Most web browsers have cookies enabled, but at the bottom of this page you can see how to disable cookies. Please note that cookies can't harm your computer. We don't store personally identifiable information in the cookies, but we do use encrypted information gathered from them to help provide you with a good experience when you browse our website and also allow us to improve our site. You can watch a simple video from Google to find more information about cookies.

Cookies used by our Website
The 43rumors website,, uses the following cookies for the collection of website usage statistics and to ensure that we can . These are anonymous and temporary. By using our website, you agree that we may place these types of cookies on your device.
Read how Google uses data when you use our partners' sites or apps:
Google Analytics Cookie Usage on Websites: Addthis cookies:
Disqus cookies:
Vimeo cookies:
Youtube cookies:

Disabling/Enabling Cookies
You have the ability to accept or decline cookies by modifying the settings in your browser. Please note however that by deleting our cookies or disabling future cookies you may not be able to access certain areas or features of our site. For information about how to disable cookies in your browser please visit the About Cookies website.