skip to Main Content

E-M5 RAW has better Signal-to Noise ratio and Dynamic than the X PRO 1??? (says Techradar)


The surprising RAW Signal to Noise ration chart from Techradar. The Olympus E-M5 is the best!

If you are from America or Europe you are very likely going to receive your preordered E-M5 next week! Meantime I have a new roundup of reviews for you!

Amy from Techradar (Click here) posted the full E-M5 review. She likes the “Excellent viewfinder, Tilting touchscreen, Customisable dials and buttons, High image quality at low ISOs and the New art filters“. What she doens’t like is the “Plastic unresponsive buttons, No in-built flash, Odd (sound) emitted“. But the real surprise is their RAW noise to ratio test (posted here on top of this post). The chart shows how the RAW files are clearly better than the Panasonic GX1 suggesting that they may not be using exactly the same sensor. The real surprise is that it even beats the Fuji X PRO 1! As usual take those charts with a grain of salt. As I learned from my physics studies no test is perfect and always depends from the testers method :)

And below you can find the Dynamic range test results (again on TIFF converted RAW files):

But overall that E-M5 looks like a tremendous camera and I can’t wait to get my hands on it! Thansk Amy for sharing this great test!!!

Black E-M5 body at Amazon (Click here), Adorama (Click here), Bhphoto (Click here), J&R (Click here), Henrys (Click here).
Silver E-M5 body at Amazon (Click here), Adorama (Click here), Bhphoto (Click here), J&R (Click here), Henrys (Click here).
Black E-M5 with 12-50mm lens at Amazon (Click here), Adorama (Click here), Bhphoto (Click here), J&R (Click here), Henrys (Click here).
Silver E-M5 with 12-50mm lens at Amazon (Click here), Adorama (Click here), Bhphoto (Click here), J&R (Click here), Henrys (Click here).
Black E-M5 with 14-45mm lens at Amazon (Click here), Adorama (Click here), Bhphoto (Click here), J&R (Click here), Henrys (Click here).

Black E-M5 body at Amazon DE (Click here), Amazon UK (Click here), Jessops (Click here) and Warehouseexpress (Click here).
Silver E-M5 body at Amazon DE (Click here), Amazon UK (Click here), Jessops (Click here) and Warehouseexpress (Click here).
Black E-M5 with 12-50mm lens at Amazon DE (Click here), Amazon UK (Click here), Jessops (Click here), eBay (Click here) and Warehouseexpress (Click here).
Silver E-M5 with 12-50mm lens at Amazon DE (Click here), Amazon UK (Click here), Jessops (Click here), eBay (Click here) and Warehouseexpress (Click here).

Black E-M5 body at Digitalrev (Click here).
Silver E-M5 body at Amazon JP (Click here).
Black E-M5 with 12-50mm lens at Digitalrev (Click here) and Amazon JP (Click here).
Silver E-M5 with 12-50mm lens at Amazon JP (Click here).
Black E-M5 with 14-45mm lens at Digitalrev (Click here) and Amazon JP (Click here).

  • E-1

    I’m not sure how often this needs to be repeated: This is not the G3 sensor. There is no indication whatsoever that this is the G3 sensor. None. Nada. No quote. Nothing. I’m fine with the sensor in my E-PL1 for the photos I make, so I’m sure whatever sensor it is it’s fine for me, but this rumor from frightened (why?) G3 and GH2 users needs to stop now.

    • Anonymous

      If M5 has diferent sensor than G3, I’m sure Olympus should say this very clear. And he is not.

      • Well first need Olympus say who so produce E-M5 sensor. ;-)

        • Why do they need to say? Can’t you make a decision yourself if it’s any good?

          The truth is, these cameras are all capable of great results and the differences between them are small in practice. But why anybody needs to be told who made the sensor is beyond me. Look at the photographs and decide if you think they are any good. Use your eyes.

          I have no idea who made the motor in my vacuum cleaner, but I know it’s really good at picking up dirt.

          • Rob-L

            @Thevoiceoverman – Thank you. Someone with some common sense. I stopped reading the forums at DPR because I couldn’t take all of the hand wringing over the smallest, inconsequential crap. I wonder how many people their actually take photos vs. just talking about cameras?

      • James70094

        Olympus did say clearly that this was a brand new sensor. Look at the number of photo-sites on the sensors, they are different. The G3 sensor has 16.6million pixels with 15.8million effective pixels. The OM-D has 16.9million pixels with 16.1million effective pixels. Saying these are the same sensor is like saying a 6 cylinder engine is the same as a 4 cylinder engine. They may perform similar, they are not the same. If they are the same, it would be the same pixel count since that is a physical difference.

    • Franky S

      These results are massively flawed. There’s a discussion on DPReview about it.

      I’m sorry to disappoint anyone, but the E-M5 does NOT have better DR than the D4/D800.

      “Just remember that 12-bit analog/digital converters have less than 12 EV Dynamic Range.”

      Therefore a result of 13 EV is extraordinarily improbable (impossible?).

      • Martin

        A result of 13 EV from a 12 bit AD isn’t impossible, given that the AD converters need not be completely linear. For obtaining measurements of larger DR than AD range, it suffices that the zero point of the AD is inexact, providing an offset rendering the first quantum smaller. Is this of any practical use? Hardly… In particular as the offset will certainly vary between samples of the same camera and even depending on operational conditions.

        • avds

          This does not increase physical dynamic range because the total amount of luminance grades does not change, although it could possibly improve perceived dynamic range somewhat if non-linearity of luminance mapping in the converter matches that of the human eye.

          • Charlie

            Nikonazis getting butthurt.

      • Dave

        I always smile when I see people regurgitate information they have read without understanding any of it.

    • Cal

      I spoke to someone in the camera industry (anonymous) and they told me that someone had opened the EM5 and the sensor is in fact a Panasonic sensor.

      • I’m sure it is, if it wasnt Olympus PR would be all-over this by claiming “totaly new extremly redesigned totally different then previous sensor”

        But because they keep quite – i’m sure its a G3 sensor,
        although i’m supporting oly on this issue – as i’m sure they improved the surrounding electronics quite a bit (better a/d converters or whatever) so the final result is likely to be a bit better then g3\GX1 even thou their using same sensors

        • Olympus said something like: “It’s true that in the past we’ve used sensors from Panasonic, but in the case of this one (in the E-M5) we cannot disclose who made it. We can only say it’s a completely new one that hasn’t been used in any other camera”.

          • tmrgrs

            This statement clearly indicates that it’s not the G3 sensor. Why are so many here in denial about this?

            • Raist3d

              Because that statement doesn’t indicate what you think. Moreover Terada himself in another interview said “We can’t disclose who made it. It could be or it could not be Panasonic” (paraphrased).

              YES he said so himself that *it could be* Panasonic- or not, but they can’t tell.

              So if it wasn’t Panasonic, why just don’t deny it and move on?

        • Anonymous

          I had lunch with a Sony R&D today & told me that they discovered a POST-IT-NOTE inside the dissected E-M5.

          The message said “Fuck You. I win. You Lose.”

          Sony’s entertainment division are now trying to use the encrypted message to further secure the hacked PS VITA.

      • Daemonius

        Nikon uses Sony sensors for ages (either own design or Sony desing), but every time their use of Sony sensor is better in DR/SNR than when Sony useses it. Ofc theres price, price is paid by different (less accurate) CFA and stronger AA filter. Thats why D3x cant match resolution of A900, cause AA filter is stronger. But on other hand they managed to improve DR (altough in real life both camera are very similar to use, if you dont mind very different handling ofc).

        Same goes to Panasonic/Olympus. It depends how they use that sensor. If they dont need to look at price too much, they can do wonders even with average Panasonic sensor. Its just about electronics and firmware (and CFA/AA :D).

      • Rob-L

        @Cal – thanks for that indisputable evidence.

        • Neonart


          I know a guy, who knows a dude, who’s cousin is a Janitor for Panasonic, who found an internal email from R&D that said their new Personal Bikini Shaver would use a resistor that may be used in the EM5, which confirms without a shadow of a doubt that it uses the same sensor as the GX1.

          • Arnold

            hahaha right on !

    • I may be wrong, but as far as we know, isn’t the sensor a LiveMOS one? Isn’t LiveMOS Panasonic proprietary? Unless Panasonic has given Oly the next generation of sensors (which seems highly unplausible if we think about the past), its highly probable that it is some flavor of the GX1 sensor…

      • LiveMOS is an Olympus trademark, not any particular technology.

    • I wrote a comment some time ago saying the same thing, but not only because there is no indication, but because the differences are just to big to be done by processor or engine alone.The difference between the same 12MP used in both Panasonic and Olympus m43 are not that big, but in every comparison between the G3/GX1 vs E-M5, there are several differences, in ISO, DN, etc. so I hardly believe it’s the same sensor but if it’s, Oly is making miracles with the image processor.

    • Raist3d

      What you are saying is not true at all. We have a couple (at least) reviews suggesting it is indeed a tweaked or saying it is a tweaked GX1 sensor, and when you examine the raw files it does look like a GX1 performance.

      Now look at the results from Tech Radar and you honestly are going to believe this camera has over 13 EV’s at ISO 200 and only a loss of 0.6x EV’s DR to 12.5 EV’s at ISO 1600? That should strike you immediately as a red flag for whatever the hell they are doing. No camera has that kind of fall off by 3 stops of ISO.

      Moreover, Terada-san himself said in a later interview, that while they can’t disclose who made it, it could be or could not be a Panasonic sensor, and they are not telling.

      So, if it wasn’t Panasonic, for sure, why just not say “it isn’t” and move on?

  • Now watch all the people who don’t want these results to be true find reasons to fault the test. I get my E-M5 this week admin. Silver. Can’t wait.

    • admin

      I also chosed Silver ;)

      • GreyOwl

        I choose Black, but Silver looks good; I wonder if I made the right choice… To late to change now, so I’ll have to live with it.

        • admin

          Black looks amazing! I would have gone for Blak if I wouldn’t have silver lenses!

          • Robin A

            I got mine in black and the ISO 3200 is so usable

          • i would get silver only to look less like DSLR

        • It’s like a tasty cake. It doesn’t really matter what colour the icing is.

          • spam

            And you can always change it later:

            • GreyOwl


            • I got as far as seeing M8 stamped on the pink top cover, the video says M9 and then decided it was a stupid way to paint a camera. On top of that, he has not got a clue on painting or masking.

  • Camaman

    Black will come next week, but silver beginning of May, here in Croatia. :-(
    Oh well more time to sell my old gear… I’m broke as it is…

  • It’s a bit weird how the E-M5 can have better DR on ISO6400(!!!) than the GX1 on ISO200 – but nobody knows…

    • Good point.

    • The test results indeed look strange.

    • Mike1

      Better wait for the DxO test results.

    • Esa Tuunanen

      Completely overhyped G3/GX1 sensor doesn’t have really any better DR at base/low ISOs than that old 12MP sensor.

    • Raist3d

      It’s not just weird, it’s flat out wrong. Anybody who believes these results is delusional.

  • Fan

    As much as I would like the results, they are too good to be true and they are not compatible with other comparisons that we have seen.

    • MJr

      Indeed. Pretty much sums it up for now.

    • There have been no other similar tests, that I have seen, and certainly none posted here. So they cannot be ‘not compatible’
      But they do show what Oly has been claiming.

  • Paul

    Chose the black one to go with my 4Ti, coming late next week, hopefully. Reading the review it seems that reviewers are just starting to get the message about CSC’s, though often through gritted teeth, they still have some quibbles but on the relatively unimportant things.

  • JF

    :-o This is so good that I’m in doubt…

  • The results are flawed somehow. They go against everything we know about EM5 image quality compared to its APS-C peers.
    Remember too that there is the issue that the real EM5 ISO values much lower than indicated.

    • Neonart

      Almost all manufacturers do this. The EM5 is not the first or last camera to have ISO values discrepancies.

      People keep bringing this up as if it where unique to the EM5.

    • darrask.
      You evidently do not look at any of the links posted here. If you had looked at them you would not have made your statement. ‘They go against everything we know about EM5 image quality compared to its APS-C peers’.
      Before you go off into a screaming lather, go back and look at the links. Back for the past month, there are lots. So it will take you time a lot of time and yes, I have looked at them all.
      3 for you start.
      As I said there are lots so I don’t expect you back for days.

  • Alvaro

    I can’t understand this…
    Yesterday we saw how some german guys tested the X-pro 1 and said (in terms of DR) it was one of the best cameras ever… and now the E-m5 is better by a HUGE advantage?????

    Can’t trust in any review lately…

    I’m very concerning about low ISO noise and image quality and it’s being too difficult to decide which camera should I buy…

    • admin

      What matters is the camera real usability and your photography skills. Don’t hurt your head against the wall! :)

      • Alvaro

        I’m an ex-OM user so I felt in love with the E-m5 the first day I saw it. Even more I can use my 4/3-m4/3 equipment…

        For me the X-pro1 better ergonomics (sided viewfinder and manual controls) suits more my photo style… but I would spent a lot of money moving to a new born system…

        But you’re right admin what really matters is improving the skills not the camera… It’s not the Arch or the arrow…it’s the archer ;)

      • Fully agree. I got my hands on a demo unit today in Kuala Lumpur.

        Very simple to use coming from an E-PL1. To me, the most significant thing is that it didn’t feel new or strange, very straightforward. It is smaller than I expected, significantly smaller than my E-PL1 with the VF2 on. I love the looks (definitely Silver for me). The buttons do lack a definite click type feedback, but I assume this is to do with the weatherproofing. Easy to get used too.

        Even mounted the VF2 on it for a laugh, that looked bizarre! Focusing speed is great after my E-PL1, higher ISO very nice. But….I’m gonna try to resist for another 12 months. Not because I am waiting for something better, but because I don’t actually need this camera yet. I’m not a pro, I don’t sell any images, I can’t really claim to be a pixel peeper except out of curiosity. My friends aren’t going to scoff at the images I print because they are “noisy”, they’ll appreciate them for what they are. I feel I can wring another 12 months of good use from my old friend the E-PL1, by which time the E-M5 will be a bit cheaper, and the zoom lens issue will be clearer.

    • Look at the images and make your own mind up.

      • One of the best statements made. One with such wisdom is always welcome.

    • Martin

      I am afraid that all the reviews providing measurements have been systematically flawed by just testing a single sample of each camera. A reasonable test would have to evaluate a statistical sample and at least tell us about the variance, which now seems to show by completely incompatible results.
      One should also note that some of the results -in particular, siganl-to-noise ratio- may heavily depend on sensor temperature. And none of the testers even dares to tell us about ambient temperature during test and whether the cameras got a cold start or haven been working for a while before taking measurements.
      Hence, one should take all those results with a large grain of salt.

      • JF

        Hmmm, I doubt that performances variance is so important (at least I hope the manufacturing tolerances are tight for cameras at this price…) and I’m not sure the temperature variation has such an influence on SNR (yes there is an influence but +/- 5°C of variation will produce less than 1 dB of variation on SNR I think) To the contrary, I think that the protocol of measurement will have a big influence (lighting, processing, test samples, JPG/RAW/TIFF, RAW conversion parameters, etc…)

        • Martin

          I don’t know for sure about the temperature dependency of SNR on CMOS sensors, but for CCDs it is known to be in the range of -3 dB every 6 to 8°C. And especially in an EVIL, sensor temperatures can rise by multiples of those 6 to 8°C over extended periods of use. Which means that we may loose multiple f-stops SNR over time.

          • Perret cap Joe gear photo blogger

            One could put the em5 in the fridge before going out taking pics

          • So! We should ask the sensor if its cosy or cool before we go out?
            Should I buy the version for Finland if I go there on holiday as my Australian one may feel cold.

            Oh. They only make one version?

            How can that be? I read here that there will be a disaster because of temperature change.

            • Martin

              I didn’t say that the SNR would be insufficient for obtaining pleasing images in Australia. The sensor and the AD path has more than enough SNR margin for most (while not all, astrophotography being one of the notable exceptions) photographic tasks even when the sensor is hot. But if you care for these crazy measurements -and most here seem to take them for the holy grail of camera quality- then you should also know that they vary wildly depending on operational conditions. And that none of the testing institutes tells us very much about how they try to normalize measurement conditions, if they do at all.

      • bli

        ?? Do you think that testers have, like 5-10 cameras during tests just to handle possible variation in each camera??? I doubt anyone does that for any camera. Why should they start now??

        • Martin

          Because if they claim to do real measurements instead of subjective tests, this is the only way to go. Everything else just delivers random values.

      • Swejk

        It is also important what kind of weather at the time of the test, still day, moon phase, pollen flight and last but not least is whether the tester had good sex …

  • Anu Nyymi

    That is TIFF with who knows what processing, not RAW.

    Also their DR curves have the tendency of looking like heroin induces jokes, so no reason to trust anything they say compared to for example DxOMark.

    • Me

      > Also their DR curves have the tendency of looking like heroin induces jokes, so no reason to trust anything they say compared to for example DxOMark.

      Comparing those with the results of DxO, at least the DR at base ISO seems to be correct for all cameras in the graph. So I think it’s pretty much plausible that the E-M5 can deliver a bit over 13 EV of DR at base ISO. However, with only 12 bit A/D, that also means that tonal range will most likely suffer.

  • Sarek

    And not a single word about the video mode. Is it really that bad…?

    • GS

      > So I think it’s pretty much plausible that the E-M5 can deliver a bit over > 13 EV of DR at base ISO. However, with only 12 bit A/D, that also
      > means that tonal range will most likely suffer.

      but this values for DR should be normalized, so we do not know if the tonal range will suffer

  • E-1

    It also needs to be repeated here: 12bit A/D has nothing to do with 12bit DR. If you put “more” = larger range of DR into 12bit – to quote a DP comment – you will get poorer tone and shade distinction, “(posterized look to output).” but the DR (resolving very dark areas and very bright areas) will be ok.

    • reverse stream swimmer

      That makes sense.
      I can compare with doing postprocessing my TIFF files, which can be both 8 bit/ch or 16 bit/ch. But when converting to JPEG, they are only 8 bit/ch.

      Both the resulting pictures 48 bit TIF and 24 bit JPEG cover the same dynamic range in the picture, but the JPEG with lesser bit resolution are more prone to posterization.

      • Martin

        Well, the 8 bit/channel and the 16 bit/channel TIFF do not cover the same DR, unless you round everything up in the conversion. If you do a standard round to nearest in the reduction of bitwidth, some non-black pixels will be mapped to plain black, thereby tremendously reducing the theoretical DR. This example does, however, show how unimportant this single DR figure actually is: It demonstrates that the most accurate conversion (round to nearest) scores much lower with respect to the DR figure than pretty inaccurate conversions, like always rounding up. Measuring DR as the contrast between the darkest non-black pixel and the brightest vastly overemphasizes the importance of the DA’s threshhold for considering an input value non-zero. While there is some point in being interested in that treshhold, namely when trying to find texture in the darkest shadows, it fails miserably in actually characterizing sensor and AD quality. But as people prefer simple numbers, they tend to tacitly ignore the more meaningful, but also less comprehensible figures that could also be obtained.

        • Me

          There even was a time when people knew that due to its 8 bits per channel, a jpeg can not cover a DR of more than 7 EV…

          • E-1


      • Gianca43rds

        Thanks E-1: very nicely and succinctly said.

  • spam

    They are ufortunately doing something wrong in the RAW-processing. You’d expect a steady drop in DR/Signal to noise levels when Iso is increased. This happens with all current sensors.

    If you get curves that go up at higher Iso then you know there’s something wrong with the test. It will happen with many cameras if you test JPG-output too as stronger noise reduction will hide the noise from the testprogram (which usually dont take detail into account).

    So, it looks like they do RAW-conversion and use some default noise reduction or (even worse) applies noisereduction themselves when converting to TIF. The problem is easy to see on several of their graphs if you read the whole test. Btw, there are several sites who do similar stuff and obviously dont understand how interpret their results (or do it properly).

    • If there was an electronic change made to the sensor at certain ISO levels then a difference would be seen in the RAW output. So indeed a rise may be possible. Have a look at the early E-m5 ISO shots you will see a change in noise and clarity at about 6400

      • spam

        You can possibly see an improvement from a really low Iso like 50 which is lower than most sensor’s native (designed) sensitivity. At high Iso there is always a steady decline in Signal to noise levels and if you see an improvement in the measurement then you know that the RAW-data has been manipulated.

        That’s not bad in itself (Panasonic has done it, Sony too and probably most others), but the improvement is caused by noise reduction in the camera. And when you look at files processed thru a RAW-converter (as most of us have to do) then noise reduction might hva been done automatically by the converter, and different converters do different things. Actually, Adobe’s converter apply some default noise reduction that even varies from camera to camera.

        Anyway, noise reduction eats detail, reduce the “apparent” noise (of course) and “tricks” DR-measurements to some degree as they measure uniform (gray) diagrams.

  • Nawaf

    A grain isn’t enough! All I need to look at are the photographs. The Fuji is amazing if you were able take the photograph in the first place.

    The more interesting comparison would be the E-M5 vs Nex 7 vs Nex 5N. DR & sharpness please just to see how far these half frame sensors have come. No graphs and RAW files only. A person can only dream.

    Now when will my E-M5 be shipped :)

    • inorog

      indeed , My x100 AF is practicaly useless in dim light (indoors) all we se are good pictures from x1 pro no word about those photos that go straight into trash can …
      and that’s where my m4/3 is worlds ahead …..

  • Perret cap Joe gear photo blogger


    • Charlie

      Fuhgeddaboutit. ;)

  • BLI

    Common wisdom says that 12 bit A/D conversion limits the DR to 12 EV (e.g dpreview Wiki). ColorFoto (Germany) published a base ISO DR of 11.3 less than a week ago. My knowledge in these things is too poor to speculate. It is theoretically possible to get higher DR by combining several shots into one, at least if you can bracket the sensivity. Or by having a high pixel sensor and average groups of pixel values (Nokia sensor, X10 density) — that will, however mainly average out the noise, I guess. Ooops — there, I speculated.

    The results seem good (too good…). If it turns out to be “just” of GX1 quality with refinements, I’m happy with that. Out in the sun with my PEN I go!

  • adriaantie

    Tanks to Olympus for supporting and financing this very funny test.

    • Vivek


      All these “tests” and “review” sites are going go down in the trash (they already are).

      Good show! ;)

  • If this is confirmed by other tests, it is really amazing and very positive for m43 system! In other aspects the new Oly is great – just the sensor was thought to be the leat jump forward.
    NOTE: Oly themselves has said they have a DR improvement over EP-3 with 1/3 step – the graph above shows 3 FULL steps. If Oly had really a 3 step improvement, they would have told us long ago.
    I have my doubts….
    Another test showed X-pro1 had way better dynamic range than: Nikon D700, Leica M9 and Canon 5DMkII. at ISO200 two steps better than nr 2 (5DMkII). Being a potential buyer of X-pro1, I have my doubts also in this case.

    • Terada say we testing two NEW sensor, one 12mpx and one 16mpx but 16mpx have 1/3 better DR.

    • Oly said the E-M5 was 1/3 rd better than the E-P3. They never explained what 1/3 rd was. Other people suggested it meant 1/3 rd of a stop but it never sounded like that from them.

      • Neonart

        This is correct. They never said 1/3 stop.

        If they meant 30%, it’s in line with the 3 stop difference from the EP3s ~10.

        I have my doubts, but looking at DPReview Raw samples and tests like these it certainly shows this is a great sensor for it’s class, in a great camera.

        If someone doesn’t like the EM5, or owns another camera, they should still acknowledge it’s merits. But nowadays no one seems to be able to do this. It’s either your stuff is like my stuff, or it sucks. C’mon folks.

  • You can dismiss TechRadar tests as much as you want – too many vested interests here, but they use DxO test glass filters and uniform light. Then they process the shots with DxO analysers. There might be a slight difference due to the difference between nominal and real ISO in Olympus, but that is only a fraction of a stop.

    The above explanation that some hidden noise reduction is at work, doesn’t explain at all the huge difference between two Oly cameras metering the same, the E-P3 and E-M5 – so Olympus haters are still getting their fair share of ridicule.

    There is now an enormous body of evidence showing that the sensor provides disruptively better performance, not only tests, but also photographers on the field reports. Even the comparometer at DPR shows that the sensor is better than any other mirrorless, not only at 200 ISO but also at 3200.

    Therefore repeating a lie won’t make it any truer. For a more in depth discussion:

    • Franky S

      Wrong on so many levels. The only thing being tested here is the software being used to convert the image from RAW. Which appears to have been provided by Olympus according to your link.

    • Anonymous

      they call it the OMD in the review far too many times. It speaks to their attention to detail. The camera name is the E-M5, not the OM-D.

      I want their tests to be accurate, but have serious doubts.

      • Anonymous

        The camera name is OM-D. The model number/name is E-M5.

      • Many parts of the world refer to it as the OM-D or OMD and not E-M5, the Asian countries and many European countries. It would appear that the followers of blogs such as this are the ones who call it the E-M5, the rest of the world less so.
        Just look at the sales sites.

    • Raist3d

      So why then are DXo numbers um… a bit different from them? Who knows best.. Tech Radar or Dxo which makes the software used and ships a professional RAW converter? hmmm..

  • Narretz

    Seems like Olympus still can’t do buttons. I remember on some EP-3s, the mode dial fell off, because it was just glued on. I need to get my hands on an EM-5 to see if it is really an issue.

    • Miroslav

      The buttons are like that because of weather sealing.

      • Neonart

        Yup. Same in E3 and E5. I don’t have any issue with them.

  • thisguythere

    Well yesterday on there was an ISO test performed on the X-Pro1 at some german site, the curve looked very different than this one, showing best result at 400ISO… so yeah, until I see these sensors tested on DXOmark I won’t believe the result. Also, pretty much every picture I saw from the X-Pro1 had me think the dynamic range is incredible whereas with the E-M5 I never had this reaction so something tells me either the E-M5 samples already posted everywhere suck or this test is flawed, I’ll go for the second.

    • aqasem

      I agree with you let’s see wait till the DXOMARK issue their results.

  • My OM-D will arrive this week for a review.

  • Don Pope

    “If you are from America or Europe you are very likely going to receive your preordered E-M5 next week! Meantime I have a new roundup of reviews for you!”

    Do you have any new information or are you still going by the date posted by B&H?

  • GS


    • Larry

      I received my black body with 12-50 on Friday (in Seattle).

  • It will be nice to see those who bought the X Pro 1 kit at twice the price of an E-M5 realise that the latter has the same IQ with twice the reactiveness, and almost 300 lenses to go with.

    NEX owners might be seriously annoyed too, poor things. They came in flocks to troll, and now they must go home to lick their wounds.

    I guess sensor wars have just begun.

    • Jason

      X-Pro 1 appeared to be a good deal for about a second, then the E-M5 proved to be a much better camera and now only a moron would buy the X-Pro1 over the E-M5!

      • Raist3d

        I think only a moron would believe that the EM5 has the same image quality level of the X-Pro 1 :-)

    • You make it sound like X Pro owners and NEX owners are snob members of select clubs that are fueling on jealousy.

      In real life a met a few NEX users and once a X100 user. You know what, they were all grown up adults. Heck, I even had some cool discussion with some one using Olympus ;)

      • jules, don’t let this out! some one will come and get you. take you away. You’ll be reprocessed and photoshopped………..

        • Photographers are so violent!
          They will frame me, shoot me and then hang me on a wall


    • Raist3d

      That won’t happen for several reasons, but the most basic one is that no, the EM5 doesn’t have the same image quality of the X-Pro 1.

      Anyone who is going by Tech Radar’s results is simply delusional.

      • Steve

        EM5 has great image quality. I’ve been using mine for several weeks and am VERY satisfied.

        I am not at all interested in lab results, I am interested in real world results and for that, it delivers. It handles better than the X-Pro 1, it focuses MUCH faster than the X-Pro 1, it has a vastly larger library of lenses to choose from than the X-Pro 1. Not only that, it’s heaps cheaper.

        • Raist3d

          That’s fine Steve and I am not saying that the EM5 doesn’t have its pros, but image quality over the Fuji X Pro 1 is not one of them. What I am talking about is not a lab results either.

          There’s a difference between seeing the EM5 as a fantastic camera with its set of pros and cons and seeing the EM5 as the solution everything photographic which hardly anyone does better. The Fuji X Pro 1 does have its advantages.

          BTW, have you used an X Pro 1 to know if it handles better? I am not sure it does, though the AF sure seems faster on the EM5.

          • Steve

            Yes, I’ve used an X-Pro 1 in the shop a few times, enough to know that I didn’t particularly care for how it handled. I was hoping to like it, actually, as I always liked the Fuji MF cameras and how they handled.

  • Miroslav

    Olympus said E-M5 is optimized for DR.
    Let’s hope these results are true…

  • Let’s wait and see what sites like DPR has to say. I trust their analyse better than “whoever” tester using whatever instrument and methodology.
    Still we have the laws of physics and underlying technologies that sets limits and creates possibilities.

    I will not buy anything before Photokina anyway :-)

  • pdc

    These test results may be flawed, BUT it further raises the flag for m43 and that is good, as I’m in it for the long term. Why? m43 is a great compromise between compactness, system functionality and image quality. Just buy the cameras and be happy.

  • safaridon

    Admin -Why have you chosen to omit showing the JPEG signal to noise graph from this review which shows the GX1 having the highest signal from ISO 200 to 600 over both the E-M5 and the Fuji X1Pro and about equal at ISO 1600? Surely the JPEG signal ratio to noise is more relevant than TIFF signal which few use?

    • Charlie

      1) It’s JPEG
      2) Panny JPEG images can’t touch Olympus.

    • Riley

      Since cameras no longer produce .tiff files the .tiff are generated from RAW files, so their relevance goes to RAW noise examination. The jpeg format will include NR that is ‘adjustable’ in the camera but sometimes cannot be turned off.

      If you shoot jpeg all the time perhaps jpeg is more relevant to you, but RAW processing has consequences for both RAW to image (jpeg or tiff) format transformations, hence is considered the better guide, without being flawless in itself.

      All these people claiming .tiff isn’t a RAW examination are clueless twats, and .tiff are used as a reliable means of RAW transformations including tests such as IMATEST, and .tiff is still considered the format of choice for printers (pro) as it is less corruptible with no compression.

  • MP Burke

    Some people here seem certain that the DR of the EM-5 can not exceed 12EV, but I can’t find any information that confirms that the E-M5 records 12 bits per pixel. If DxO says the NEX-7 records 12 bit yet its graphs show over 13 EV DR, does this mean DxO have got it wrong as well?
    I was one of those who was sure that the sensor in the EM-5 would be the same one that’s in the G3 and GX1. However, there is a discrepancy in the manufacturer’s specifications (total pixels 17.2M for EM-5 compared to 16.7M for GX-1) which suggests it may indeed by a different sensor.
    This in itself does not mean that the sensor is not made by Panasonic: they may have started making another 16MP sensor which they intend to put in the G5.

  • Perret cap Joe gear photo blogger

    I expect a hard landing for oly brand apostolists when the first real reviews are out

    • bli

      Your prediction abilities will come in for a hard landing: no-one expects the E-M5 to be as good as this test indicates; as long as it is slightly better than the GX1, buyers will be happy. But I doubt you will realize how flawed your expectations are. And I doubt anyone cares.

    • How is this not a “real” review? Because you don’t like the result?

  • If Techradar’s tests had shown that the X-Pro 1 beats the E-M 5, would Admin and all these posters be expressing skepticism and warning us to be wary of believing the results? Probably not.

    It may be that Techradar’s methodology is flawed, or it may be that it isn’t; we have no way to know for sure. All this disparagement is purely the result of confirmation bias.

    • MJr

      Well obviously, what else would you expect, when against all odds the little duckling is claimed to beat the wolf. It’s only logical to question the illogical.

    • Raist3d

      On first pass it seems that way but ironically the confirmation bias you talk about is for those who are cheering this test. You see, the issue here isn’t that some may be “against Olympus” and thus pick on Tech Radar- the problem is that the results they posted don’t make any sense.

      For example, Look at the DR drop from ISO 200 to ISO 1600 on the EM5 in their test, let alone the measured DR at ISO 1600. We see a drop of only around 0.6x EV’s stops for 3 stops of iso (!) then the DR at ISO 1600 is 12.5+ (??!). If the EM5 only drops 0.6 ev’s for 3 stops of ISO increase, it would be like the first digital camera ever designed in consumer space with that ability in all the history of digital camera designs.

      Doesn’t that strike you as a bit odd? Forget there’s competitor cameras. Forget everything else that people are comparing the EM5 to an X Pro 1 and the like. Just ask yourself- does it seem reasonable to you that the EM5 is the first camera ever to behave much better in DR drop per ISO increase over anything else ever designed?

      This isn’t confirmation bias. This is straight forward scientific common sense. Of course, if this is repeatable- particularly you know DXo who makes the software they used, then it will be very interesting.

      Do keep in mind- DXO’s numbers and Tech radars differ on cameras. Who you think knows best? Tech Radar a review website or DXo who makes the software they used and ships a professional raw converter? Once again, that’s not confirmation bias, but following evidence.

      I would say to you though- you talk about confirmation bias.. did it occur to you that the overwhelming cheering for Tech Radar’s review are those with it? Seems like proof positive of it in liew of what I just said.

  • Anonymous

    Did this “test” use calibrated ISO-values? Or the ISO-values shown by the camera (in which case you would have to move all E-M5 points to the left….

    • BLI

      It is possible the curves of all cameras would need to be shifted to the left: it is hardly only Olympus who uses a different ISO calibration.

      • Riley

        not if one wanted to comply with CIPA ISO standards. The standard allows that a camera will deviate 1/3 stop, and this happens due to less than desirable amplification across a range of ISO. A camera may begin at ‘0’, and progress to +/- 1/3 stop over a range of ISO, or visa versa.

        Unlike what people are seeing as ‘cheating ISO’, in effect compliance with CIPA ISO says that isnt happening.

        DxO on the other hand, DO NOT COMPLY with CIPA ISO standards, their measurements are taken differently, they mean different things. Hence their ISO figures are different, and this is true to every camera I have seen tested there particularly when looking over a range of ISO.

        CIPA are the agreed standard that all camera manufacturers must (and do) comply to. DxO hold no such agreement.

  • I checked TechRadar method: they use test glass from DxO and Dxo Analyzer. The only discrepancy might be in the RAW developer. But if there was built in noise reduction (more than in other brands, as trolls pretend) then the image samples would show it.

    Instead if you go to that page you will see beautifullly detaailed hair, and a counter Sun temple, which proves the dynamic range.

    Trolls can’t have it both ways. This is a DxO test, and Tiff is compared against Tiff, and the advantage is even better than in Jpeg. Now Tiff is the uncompessed visual equivalent of RAW, so there is little to object too.

    Even if you argue against the RAW developer used for Oly (why don’t you argue against the others?) one can see that the E-M5 has 2-3 stops advantage against the E-P3, *using the SAME developer*, so how to explain it, if not with a *much* better new sensor?

    But trolls will be trolls, especially those who come from the dSLR world and now see their big guns suddenly obsolete.

    • pdc

      Good one! And thanks for the clear summation on the validity of the results.

      • Charlie

        QFT, amalric!

    • Anonymous
    • Riley

      I think (outside of the rant) what Am says here is essentially right. Whatever they are doing to test DR (there are claims of noise abating software) theyre using the same method they use on everything else.

      For now in DPR we at last begin to hear how this camera might have RAW NR, but guess what, just like almost everything else. And I note if you go to TR’s review at around page 4 you can see resolution tests at high ISO, and just like 5DIII resolution begins to melt at high ISO. Yet EM5 remains within parity to 5DIII, so nothing ‘unusual or unfair’ is happening.

      But I cannot say strongly enough that Sony, and hence Pentax and Nikon have been doing these things for years and have managed to escape much criticism.

      If its ‘fair’ for Sony/Nikon/Pentax to do this and have their DR reported as a solid 14 (example) stop, then its going to be ‘fair’ for Olympus to do the same things.

      now go back to sucking eggs….

      • Thank you Riley. Sometimes one can feel very lonely trying to bring some rationality to the fold.

        The fact that Olys underexpose, and therefore there’s hidden noise reduction has been blown out of proportion, compared to what has seemingly been achieved.

        We are speaking of a 2 stops progress compared to a 1/3 stop ISO difference. And as you remark, other brands do the same thing.

        TR has some sample images which show great detail in a portrait with long red hair. That wouldn’t happen if noise reduction was so intense by default.

        Same goes for skies when shooting towards the sun. DR cannot be faked when you look at real life, single shot, images.

    • avds

      The Sun shot from the EM5 looks very promising indeed.

      That said, this is not a “DXO test” since, whatever techradar’s process, their graphs don’t match those from dxo sensor database for cameras that they run DR tests, with a typical difference of about 1 EV and up to 3 EV at some ISO stops (eg Nex-5n dynamic range graphs).

  • CRB

    If indeed true, i will be getting one. But i rather wait for IR, dpreview and DXO reviews to be sure of it….

  • Vivek

    The posters supporting the “tests”:

    Buy a few of the EM-5 at the current price!

    I will get one when the prices are fair. :)

    • Panda9

      > I will get one when the prices are fair. :)

      You must be outside the US. In the US, the E-M5 price is probably $200-300 less than its price position should be compared to other cameras. It’s only a couple hundred more than the E-P3’s list price, and several hundred under the Sony and Fuji competition.

      I’d say that the price of the E-M5 in the US is already more than “fair.” :)

    • ronnbot

      “I will get one when the prices are fair. ”

      For the bodies, an e-m5 is $200 cheaper than a nex-7 and $700 cheaper than a xpro 1, both of which do not have 5-axis is, full wheather sealing, touchscreen, accesory port, etc. We like stuff to be cheaper but The price is fair and may not go down as quickly as other m4/3. I will most likely buy one (silver body) soon when i see/try it locally.

      • Vivek

        Add the grip and it becomes more than the NEX-7. I can not understand it (the price) at all. It will have to come down considerably.

        • Digifan

          So the NEX7 has built in grip?
          Dream on, there’s no battery and vertical grip for the NEX7.

        • Neonart

          Add the weather-sealing and 5-axis image stabilization and how much is the NEX7? Add the extensive lens selection to the NEX7, and how much is it now? Oh, wait. The NEX7 has none of those things… Nevermind.

          The grip is an accessory for those who want it. It’s not required for use, operation, or image quality. At least there are 3 grip options on the EM5. How many for the NEX7?

          In the US at least, this camera is a good deal. Will it be less in the future? Of course. But the current release price is great. (Not so great in EU, UK, etc.)

          • The Real Stig

            The UK price isn’t so bad if you get it now with the free grips promotion.

            I have mine pre-ordered here in Ireland for €1350
            and will be getting the free grip deal also

            That is $1,766.25

            The same combo from B&H in the US is $1600

            The price difference is only $166.25 in total which could easily be accounted for by tax differentials.

            Try comparing Apple product price differentials between Europe and the US.

            I believe there are four grip options as there is also a hand strap available.

    • Agent00soul

      I’m getting one now, because the price is already fair.

      • Ya, F@#k the NEX system!!


    Yeah, next thing you know it’s going to out-resolve the 645D.


  • Khµfµ
    • He shows as much sense here as he does painting a Leica pink.
      He seems to think its a pany sensor, but what would any one who paints a leica pink know.

  • My dear colleagues,

    As Wikipedia says:

    “Dynamic range, abbreviated DR or DNR, is the ratio between the largest and smallest possible values of a changeable quantity, such as in signals like sound and light.”

    So DR has *N*O*T*H*I*N*G* to do with the RESOLUTION of representing these values, that is, the bit depth of the AD converter. Assuming the sensor I/O is not necessarily linear the ratio between the largest possible value S[2^n – 1] and the smallest possible one S[0] can be virtually any quantity you want (n is the bit depth of the AD converter).

    Hence, the E-M5 dynamic range – at least theoretically – could be >13, no matter if the AD has more than 12 bits or not. A completely different question whether the E-M5 really has >13 EV DR (I have serious concerns about that).

    • Riley

      dear clueless
      “So DR has *N*O*T*H*I*N*G* to do with the RESOLUTION of representing these values”

      All DR tests calculate DR from a given point where noise becomes unacceptable to the white clipping point. If in a process you can reduce noise, therefore you increase the range of DR.

      Likewise if a sensor, a file, or a process has NR (noise reduction) involved, then the base from which DR is read will shift deeper into the shadow side, hence additional DR from the furthest white-point clip, to the ‘acceptable’ noise in black.

      If NR is happening it will scrub out detail, so if you check the progress of resolution as ISO goes higher, perhaps compare it to another camera, you can see if resolution falls, indicating the presence of NR. Or if resolution is consistent, indicating a good correlation to existing products, no more or no less NR than is acceptable elsewhere.

      • Esa Tuunanen

        Cameras doing noise reduction at sensor/RAW level is nothing new, I think it was Canon who was first caught doing it something like half dozen years ago…
        Which also coincides with time when Canon’s sensors were considered better than others.

    • Esa Tuunanen

      If Sony can get that fair 13 EVs of DR out of only slightly larger pixel/photosite there are no laws of physics which would prevent nearly same performance for EM-5’s sensor. (or same considering sensor being newer)

  • napalm


    digitalrev hands on:

    • MJr

      That crap was probably the most uninformative ‘review’ ever, with completely useless ‘conclusion’. So it looks like the old OM, get over it already !

      And yes i know what DigitalRev is like, but usually there’s at least something useful inbetween the bullshit as well.

      • tmrgrs

        Kai is an idiot and his site is a joke.

        • napalm

          he’s a nikon fanboy and seems he’s not interested in any other camera than the D800 and D4, just watch his D800 vs D4 video hehehe. and they have a 15-min review for the D800

          • MJr

            Well i wouldn’t go that far, guys. And he’s certainly not a Nikon fanboy. He loved the Olympus E-P3, especially with the 12/2, and he uses a Leica M9 himself i believe, while everything is filmed with the Canon 5DII. But yes i guess he prefers Nikon for SLRs himself.. and we’re all allowed a preference no?

            • napalm

              dont worry, he’s a self-confessed nikon fanboy. i think he said that in one of his reviews :)

              to be fair, some of their fans were complaining on the focus on mirrorless cameras so they might just be balancing their reviews

              • MJr

                Ugh, yes i guess, ironically. I just hate the word fanboy as it’s used in the wrong context 99 out of 100 times.

                Anyways, one can only review cameras that are there to be reviewed, and you guessed it, lots more mirrorless than SLRs.

        • hannes

          In my opinion Kai’s reviews are most entertaining!

        • Raist3d

          Did you think the same when he reviewed the newer fast Olympus primes?

      • napalm

        their “reviews” are for entertainment purposes only, to me at least. i never take their “technical” comments seriously :)

        • MJr

          true. and i liked it, when the ‘entertainment’ was a way to bring information, but he seems to neglect the useful part of it these days. and why would he bother, us mindless youtubers will watch anyway.

    • If you go to the latest hands on or reviews, and check the comparometer at DPR you’ll see that the E-M5 behaves pretty well.

      At all ISO, except perhaps the highest it has more detail than the NEX and the X100, which were believed to be the champions until now.

      These are Jpeg comparisons, but if you have a look at foliage and hair in the samples pages, it still behaves very well. So I think that Oly found a fair balance between resolution and noise.

      One can’t really as for more, especially if DR holds while ISO is increased.

  • Gianca43rds

    The real question is if the silver model takes better photos than the black one.
    Can’t wait to read tests about that!
    I preordered the black model: what if the silver one is better???
    The suspense is killing me.


    • Gianca43rds.
      Its easy to see. Just look at the pictures taken. Put all the good ones on the right and the rest on the left. Put a piece of paper on the right pile with ‘Silver’ written on it. Put another piece of paper on the left with ‘Black’ written on it.
      Now we will all know through this deep scientific reading of labels and checking the lable spelling and positioning, which is better!

  • Bob B.

    I don’t know what to think about this camera….the reports are just all over the map.
    At any rate, at this point in time the over scrutiny and variation is boring at best.

    • MJr

      “At any rate, at this point in time the over scrutiny and variation is boring at best.”

      lol, yes, over scrutiny and variation, how boring is that. we want only one single review that is perfectly average and exactly as expected. that’ll get the adrenaline pumpin.

      • Charlie

        Something sneaky happening ….. too much variance in testing around the interweb.

    • Bob B.

      yeah…there is just such a difference in the reports and speculation…it is all over the map?!?!?!

  • chris

    I don’t get all the hate directed at the EM5, Xpro1 or at every other camera. The reality is that the Nex7/Xpro1/Em5 and so on all do the same thing, you push the button and capture an image. The differences between cameras when processing RAW’s are pretty small these days. Great photos are created by the photographer, not the camera. Arguing about JPEG’s is a waste of time.

    All this arguing about fractions of DR numbers is laughable, buy what camera serves your needs and shoot the hell out of it. In the end you’re going to have to commit to a lens mount, pick one or two and run with it. Who gives a shit about the name stamped on the sensor? Or if camera X is “better” than camera Z. Digital sensors have reached a nice maturation point, the lenses you buy now make a bigger difference than the body because most have settled in around 16-18 MP for APS-c and M43 and that’s plenty of resolution – especially since most photos are just slapped on Facebook and Flickr.

    Why do so many people try to justify their purchase by bashing other cameras? I have a Nex7, a GH2, a 5d2 and I’m going to get a EM5. Each camera is awesome. If I want low light video I go 5d2, 1080/60p Nex7, more portable video I go Nex7 and Gh2, better stills 5d2 or Nex7 depending on how much I want to carry and so on.

    I now avoid DPreview forums because every thread is just an endless amount of bickering about mindless things like this. What’s the point?

  • Leo
    • Anonymous

      Are those hepatocytes with immunologic stain? BTW, nice captures, I really want to try like this to.

    • Fiatopichan

      Nice capture, I really want to try it on microscope too. BTW, is that an image of hepatocytes with some kind of immunologic staining?

      • Leo

        Sorry, I dunno what kinda cell they are.

  • sorry my inglish

    put em5 guts into a 620 and take my money

  • marilyn

    why i cant put my comments??

  • lorenzinonkatu

    Mmmmh, suspicious results.
    Good camera though.
    Who cares who made the sensor. It is the camera that appears to be nice as a whole. And now we have nice lenses to play with. This is what matters…

  • Vincent

    I don’t believe in techradar result. I will wait and see how E-M5 performs on dxomark labs test.

    • Richard

      And what happends if dxomark confirm this results? Whos test are you then going to wait for, becuse the you will probably will not belive them either. Like all other Oly haters that has posted above.

      • Raist3d

        Actually I would believe in a big part DXo’s if it confirms it, though the best is to check the raws for yourself. Dxo has a track record I have corroborated with different camera makes and models that is ballpark of what you can expect relative between cameras.

        The reason why believe Dxo and not Tech Radar, besides the pointed out red flags in Tech Radar’s results, is that Dxo makes the software Tech Radar allegedly used, and develops a professional RAW converter-i..e they know what they are doing.

    • Raist3d

      Yup. It’s amazing that it doesn’t ring a bell that the EM5 is doing over 13 stops at base ISO only to fall to 12.5+ EV’s at ISO 1600. That has spelled “super wrong” all over it. But people believe what they want to believe.

  • Hi,

    The most unclear problems in upgrading for a DSLR is knowing the main difference between cropped frame and full frame cameras.

    Andrew Childress

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.