Olympus E-M5 gets tested by germans. Beats Samsung APS-C sensors!
At least according to B&H (Click here) we are exactly one month away from the first Olympus E-M5 shipment. But seeing that the E-M5 still remains within the top list at Amazon (pre-)order rankings (Click here to see it) I doubt you will get the camera on that day 
Anyway today we have a very interesting new review:
The German testers from Digitalkamera DIWA lab (Click here for the translation) tested the Olympus E-M5 with the new 12-50mm lens.
They first tested the lens and they were satisfied with the results with the only major issue being the high Chromatic Aberration at the wide end.
And now let’s talk about their E-M5 Labortest. First they discovered that the real sensitivity starts with ISO 120 (when the camera is set at ISO 200) and ends with ISO 14.200 (with the Olympus setting being at 25.600). This means that the E-M5 has a 2/3 stops difference from the official Olympus ISO value. The signal to noise ratio corresponds to a decent 40db at low 200 ISO (120 real ISO). It goes down to 35db at 1600 ISO. At high ISO the luminance noise is strong while Olympus handles the chrome noise much better. Digitalkamera says that up to 1600 the performance is good and from that point upwards the image quality visible decreases. Same thing happens for the dynamic range which up to ISO 1.600 reaches 11 stops (very good result). At ISO 3.200 it goes down to 10 stops, at ISO 12.800 it has 9 stops and at ISO 25.600 8 stops only. The sever German reviewer is positively impressed by the results that are better than the ones reached by the Samsung NX200 which uses a larger APS-C result.
To see the full graphs and test you have to pay a couple of Euros
More E-M5 bits: Olympus E-M5 video at Quesabesde (Click here). A finish hands-on at Sulanto (Click here). Another full German review at Chip (Click here).
Olympus 12-50mm test: Photographyblog (Click here) posted the full Olympus lens review: “Photographers who require maximum sharpness and resolution will of course be better served by one – or more – of the excellent, fast, sharp Micro Four Thirds primes; but the Olympus 12-50mm f3.5-6.3 is hard to beat on versatility and price-performance ratio.” Lenstip (Click here) also posted a full review of the lens.
—
E-M5 preorders (Click on the names of the stores to see the product page):
Black E-M5 body at Amazon, Adorama, BHphoto, J&R, Warehouse UK, Digitalrev and Amazon Japan.
Silver E-M5 body at Amazon, Adorama, BHphoto, J&R, Warehouse UK, and Amazon Japan.
Black E-M5 body with 14-42mm lens at Amazon, Adorama, BHphoto , Digitalrev and J&R.
Black E-M5 body with 12-50mm lens at Amazon, Adorama, BHphoto, J&R, Jessops, Amazon UK, Amazon Deutschland , Digitalrev and Amazon Japan.
Silver E-M5 body with 12-50mm lens at Amazon, Adorama, BHphoto, J&R, Jessops, Amazon UK, Amazon Deutschland and Amazon Japan.
gl
2 years ago |Here’s a fun little conspiracy theory:
Olympus to Panny: “guys, we can no longer afford to be behind your sensor tech by one generation”.
Panny: “you know we can’t give you our newest sensor, how would that look?”.
Olympus: “then we have to go elsewhere”.
Panny: “Wait – how about we give you a new generation sensor (not with the GH3 features but with competitive sensitivity & DR) – but you can NEVER reveal it’s from us. And you have to buy the office sandwiches for a week. Deal?”.
Olympus smiles politely.
: )
Chris
2 years ago |So according to the German reviewers’ methodology the EM5 gets 11 stops DR up to ISO 1600. Does anyone know what, according to their methodology, the 12 MP m43 sensor got?
TheEye
2 years ago |That’s a shitty review. More like a lame editorial on the E-m5 with crappy 12-50.
What happened to the proper camera tests German magazines used to perform, not to mention well-written reviews? They even tested shutter accuracy and similar things. They did field tests. Ah yes, those were done by ColorFoto and Fotomagazin. There was no DIWA then.
Someone please give me ring when the E-M5 is tested by the Swiss (Because they are neutral).
Bob B.
2 years ago |I am curious….may be enough of a change for me to sell my G3 and GX1. Looks significant.
lnqo-M
2 years ago |When i see on E-M5 spec. 16.9mpx total, and not 16,7mpx so on G3 and GX1, on amazon.jp. I think this a new sensor, but now can we also look that is better.
Nawaf
2 years ago |Yes, at 25,600 it’s better than the nex! Awesome can’t wait to get mine. It’s a tiny sensor but does an excelent job.
As for the jpeg, well they didn’t state which setting the NR was left at. At off or low it would be fine, otherwise the results would be bad.
lnqo-M
2 years ago |Thanks, can we say this is a G5 sensor.
BLI
2 years ago |Interesting! Question: was it a mistake to include ISO 25 000? I don’t know, but check how it compares to Nikon D4… The D4 at 200 000 looks much worse than the E-M5 at 25 000. To me, the D4 seems to be 1-2 stops better???
Agent00soul
2 years ago |E-M5 samples up at DPreview now! Looks at least as good as the NEX-5. Maybe even better…
And the G3 is WAY behind. No, I don’t think it’s the same sensor…
Nawaf
2 years ago |It uses flash so I can’t check it out on the ipad
Bob B.
2 years ago |Hey..DPReview just posted their studio samples of the OM. The raw files look significantly better than the G3 and GX1. Maybe this is a 3rd-party sensor. First true evidence I have seen that may prove it.
amalric
2 years ago |You can now check the comparison samples at DPR. Let me remind that the only way to compare Dynamic Range is in Jpeg.
Resolution to can be checked through the comparometer there. Seems to me that it is on an equal footing to the NEX-5, so that the gap with APS is bridged.
Good news!
I am not going to get it for now, so it’s not a selfish observation. But I am happy that m4/3 has stopped being a Cinderella in IQ.
Best IQ+Lenses coming all the time = good investment.
Agent00soul
2 years ago |My supplier in Sweden writes “deliveries will be begun in april”
Anna_T
2 years ago |When will you get an E-M5 in your country ?
I see that B&H announces it for the 16th of April, but here they are not expected before the end of May !!?
I looked among all the online dealers. Most said delivery date unkown. Among the most reputable dealers they announce it for the end of May. One single dealer announces the kit with 12-50mm for mid-may and the body only for the end of May.
Ouch.. how comes other countries are getting it earlier ? and with free accessories ?
bli
2 years ago |Drooling for the GH3?
Bluebox10
2 years ago |Admin, the translated review is up on lenstip. And new samples at dpreview.
Nawaf
2 years ago |Finally, someone who can explain it to us instead of just stating a fact that means nothing. If something was fishy then wouldn’t this have shown in the tests with the GX1.
Thank you for spending your time to explain it to us.
Miroslav
2 years ago |plus eins
Ulli
2 years ago |in some way i agree; tests can be a guide for the people who need it. The real test is the moment when you have bought the camera and start using it, its actually a kind of gambling, the chances that it wont suit you is bigger then that it will fit your needs.
Anonymous
2 years ago |Sooo…. being a bit shocked you’ve been lied to makes you a troll…?
You’re right, we should all just be be fan boys
This does perhaps explain why my E-P3 is so badly exposed when using a light meter…
Guess I’ll try adding 2/3 stop from now on…
bli
2 years ago |90mm? Why?? The crop factor for the NEX 7 sensor is 1.5, isn’t it? 50*1.5 = 75mm.
Gabi
2 years ago |Thanks for the comprehensible explanation!
Fraenzken
2 years ago |I think many people don’t know what the DXO (and digitalkamera.de) ISO definition means. I have an Olympus E-620. DXO says that ISO200 on the E-620 is not really ISO200, but ISO125. ISO100, on the other hand, isn’t ISO100 – it’s also ISO125. So there’s no difference in sensitivity between ISO100 and ISO200.
What happens when I change ISO from 100 to 200 on my E-620 and keep the aperture constant? The exposure time halves. Why that, if ISO100 and 200 are (according to DXO) identical?
Well, a sensor doesn’t have several sensitivities. It has one base ISO, that’s all. Let’s assume base ISO is 100. If you take a picture at ISO200, it only gets the half amount of light. The camera can compensate for this in two ways: either by multiplying the signal with 2 already on the sensor (than DXO calls ISO200 ISO200). Or by applying a different tone curve to the captured raw data when processing the jpg (than DXO calls ISO200 ISO100). The only difference is when the multiplying step is done!
Olympus goes the second way. The advantage: This doesn’t effect the amount of electrons stored in the electron wells on the sensor. Which leads to less highlight clipping – you gain one stopp in the highlights. But this comes to the expense of additional shadow noise: The multiplication on the sensor adds less noise than the application of a steeper tone curve when processing the jpgs.
Bottom line: The “ISO lie” isn’t a lie at all. If you take a picture with the E-620 at ISO200 and with a certain time-aperture-combination, and then take a picture with the Canon 60D at ISO200 with exactly the same parameters, you’ll get two pictures that are exposed exactly the same way. Importantly, by “faking the ISO” you don’t get less noise for a certain ISO, but more. At the same time, you gain some highlight headroom.
Fraenzken
2 years ago |I think many people don’t know what the DXO (and digitalkamera.de) ISO definition means. I have an Olympus E-620. DXO says that ISO200 on the E-620 is not really ISO200, but ISO125. ISO100, on the other hand, isn’t ISO100 – it’s also ISO125. So there’s no difference in sensitivity between ISO100 and ISO200.
What happens when I change ISO from 100 to 200 on my E-620 and keep the aperture constant? The exposure time halves. Why that, if ISO100 and 200 are (according to DXO) identical?
Well, a sensor doesn’t have several sensitivities. It has one base ISO, that’s all. Let’s assume base ISO is 100. If you take a picture at ISO200, it only gets the half amount of light. The camera can compensate for this in two ways: either by multiplying the signal with 2 already on the sensor (than DXO calls ISO200 ISO200). Or by applying a different tone curve to the captured raw data when processing the jpg (than DXO calls ISO200 ISO100). The only difference is when the multiplying step is done!
Olympus goes the second way. The advantage: This doesn’t effect the amount of electrons stored in the electron wells on the sensor. Which leads to less highlight clipping – you gain one stopp in the highlights. But this comes to the expense of additional shadow noise: The multiplication on the sensor adds less noise than the application of a steeper tone curve when processing the jpgs.
Bottom line: The “ISO lie” isn’t a lie at all. If you take a picture with the E-620 at ISO200 and with a certain time-aperture-combination, and then take a picture with the Canon 60D at ISO200 with exactly the same parameters, you’ll get two pictures that are exposed exactly the same way. Importantly, by “faking the ISO” you don’t get less noise for a certain ISO, but more. At the same time, you gain some highlight headroom.
evenflo
2 years ago |That’s exactly what I’m considering doing as well! Do you have any experience how, e.g., an ‘old’ Zuiko OM 1.8 50mm would behave on a NEX? One would of course need an adapter, and, as I understand it, there would be no autofocus. Depth of field check should be ok, since this is being done mechanically on the OM lens. And the (35mm film) 50mm lens would be something like a 90mm for the NEX sensor, correct?
Tom
2 years ago |About the real sensitivity with DXO. As long as the aperture and shutter speed combination is correct and the picture is good, its not an issue as I see it. Its the same thing with Canons.
http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/724%7C0/(brand)/Olympus/(appareil2)/483%7C0/(brand2)/Canon
Hecti
2 years ago |I feel like its been forever since we have heard anything about the GH3…
NiKo
2 years ago |> 14.200 (with the Olympus setting being at 25.600
SMART, so ISO 6400 is actually 3550 in real life. That explains things
Fraenzken
2 years ago |Right. In fact this strategy even worsens noise because applying a different tone-curve to compensate for underexposure comes to the expense of noise (in comparison to doing this directly on the sensor).
Bob B.
2 years ago |Has anyone done a meaningful test comparing the RAW files of the Olympus OM to either the Panasonic GX1 or G3?????? (hopefully using any lens but the 14-42mm X lens so that the images could actually be sharp?).
YouDidntDidYou
2 years ago |@Robbie
just had a quick flick through quite a few UK photography magazines and they are all gushing of the E-M5 in the previews and quick reviews
MP Burke
2 years ago |I am far more interested in seeing how well the EM-5 performs at the base ISO rather than finding out whether the ISO values are “correct” according to DxO. If you look at e.g. the DxO measurements for the Nikon 1 sensor you find a similar “low” ISO trend: such discrepancies are neither new nor particularly uncommon.
lnqo-M
2 years ago |It these trolls have very bad faculty of vision, so them need a test to tell them this.
YouDidntDidYou
2 years ago |@Kylberg
I think the Olympus E-M5 interface is much improved, I played with the E-M5 for about an hour and was able to quickly change all the settings I wanted to quickly and believe I change them often!!!
BornBad
2 years ago |if i get a darker image, read less light, then it’s not the same image or ISO.
lnqo-M
2 years ago |AF speed on 1/100000 sec will be more creative so a Lytro camera, but a ISO on 10000000 need be noiseless.
Zonkie
2 years ago |Exactly, good that someone explained this correctly.
No one is cheating. Each manufacturer tunes this to get the best performance, but the ISO speed in user terms is correct in all of them, no matter what DxO measures.
Zonkie
2 years ago |This is just a technical thing that is not relevant for end users. When you shoot at ISO 200, you’ll surely get a correct exposure for that ISO value (that is, same shutter speed at same aperture as you would get in a Panasonic -or any other- camera at the same nominal ISO – bearing metering differences which in case of happening would result in different “brightness” of the image).
So it’s not important and it’s not a trick by manufacturers to give better results in tests.
OlyFan
2 years ago |ooooo by 0.01 and 0.06 seconds!. Massive!!
bilgy_no1
2 years ago |NEX-7 is a great choice if you’re going to use mostly legacy lenses.
If I’d be looking at getting into a system with native lenses, I’d go for the E-M5.
Robbie
2 years ago |LOL look how the trolls are pouncing on the so called FAKE ISO again.
It just never cease to amaze that there is this systematic slandering of Olympus going on that the same things will happen simultaneously at different forums.
Nawaf
2 years ago |+1
agent00soul
2 years ago |Well, this is what ANY camera maker MUST do, to be able to have soft clipping of the highlights.
Remember, all sensors have a strictly linear response, so this is how you emulate the S-curve of film.
Jorge
2 years ago |Test with even better result a german site chip.de. 11 stop dynamic!
Test:
http://www.chip.de/artikel/Olympus-OM-D_E-M5-Systemkamera-ueber-1.000-Euro-Test_54946190.html
Highscores (SLR over 1000 Euro):
http://www.chip.de/bestenlisten/Bestenliste-DSLR-%C3%BCber-1.000-Euro–index/index/id/969/
Second best camera! IQ better than Sony A 77!
ProShooter
2 years ago |GX1 ? Kein IBIS, Kein Interesse !
Kylberg
2 years ago |“Fastest AF” often turns out to be under very special conditions only. It is more about marketing than an absolute truth (like with EP3). G3, GH2, EPx, G1X all have fast AF – and EM5 too, in some cases mabe even better.
How much AF speed do we need? 1/100000 sec? How much ISO do we need? 10000000? – Probably not.
DrWu
2 years ago |This issue has been discussed ad infinitum on DPReview forums. Nobody here is fooling anybody. Camera manufacturers ( Japanese at least) have another standard or method for measuring sensitivity values of imaging sensor and the Holiest of the Holies, DxO, has in their bottomless wisdom invented their very own method. Note: This is not my information… this is what the DPReview testers regularly give as explanation why their and camera manufacturers´measurements are different from DxO tests.
Kylberg
2 years ago |I still think the EM5 is a great camera. That ISO thing is not much of a problem. Using the camera you will get well exposed images. Camera X may be able to perform slightly better at highest ISO, but who cares. You can still get a shot of a black cat in a dark room with an EM5/G3…!
Unfortunately I have a few issues with Oly user interface, will wait to see what Pana will bring in September.
BTW: I Sweden EM5 and Nex-7 costs almost the same unlike rest of the world. It is the Nex that is cheaper.
Nawaf
2 years ago |Stating facts about the ISO doesn’t help us understand. Atleast comparing it to something would have helped. Where can I find a similar test done on a nex or whatever?
Gabi
2 years ago |All these tests really don’t matter at all. I prefer to look at real life samples and then decide if the camera is good for me and for what I want to shoot. I can’t remember of any official test site to report on the problems of the Panasonic X lenses (just to give an example).
molikmolik
2 years ago |test and super score by chip.de :
http://www.chip.de/artikel/Olympus-OM-D_E-M5-Systemkamera-ueber-1.000-Euro-Test_54946190.html
JF
2 years ago |40 dB is better than GH2 DxO results but I suspect it is not comparable measurements (JPeg with noise reduction vs true RAW) (I’m not defending GH2 I would like very much that EM-5 is better !)
Jb
2 years ago |i think NX200 not comparable with E-M5, is not good camera,
NX200 is comparable with E-P3 and low pen cameras
JF
2 years ago |Who buys the test and makes a summay ??
lnqo-M
2 years ago |Right!
lnqo-M
2 years ago |Well my car have 16″55-205 tyre on now, but orginal is 15″65-195.
Also back to the film days is this a normal by “develop tweakes” like this.
safaridon
2 years ago |Olympus claims that the EM5 has the fastest AF of any camera yet this test shows that the Pany GX1 is slightly faster!
khoa_sus2
2 years ago |This is nothing new. Look at any camera from Canon or Nikon made in the last decade
BLI
2 years ago |The German test site claims to have used DxO software in their test.
digitalkamera.de
2 years ago |Den Signal-Rauschabstand kann man nur als befriedigend bezeichnen
Allerdings gab es eine Kamera die in unserem Labor schneller war als die OM-D E-M5: Die Panasonic Lumix DMC-GX1
the hype seems to be close to over…
khoa_sus2
2 years ago |Newsflash: check CaNikon’s cameras and their reported vs actual ISO values. It’s the “industry norm” to under-report.
Ralph
2 years ago |Why not? Its the same proportions. Depends what you want it for; if you bought it speciffcaly for ISO 26,500 and it can’t do it – not even close; not much better than half in fact – then thats pretty poor.
Agent00soul
2 years ago |There is nothing cheating about it. It just means that Olympus reserves 2/3 stops for highlight headroom and adjusts the tone curve to compensate. This will give you better highlights at the cost of slightly more shadow noise.
JF
2 years ago |Yes, it makes me happy !
www.MilosJanata.com
2 years ago |At least they can’t trick Dxo mark .. yet
BLI
2 years ago |I don’t think your comparison is valid. Of course, it is highly unfortunate if the ISO numbers are not accurate. But in this case, the error will actually make some people happy
— I guess.
Ralph
2 years ago |How can olympus get away with it? Surely there is a point when it becomes false advertising? a few % either way is reasonable, but being 66% out is pretty poor.
If they sold a ’20mp camera’ that turned out to only have 12mp people wouldn’t put up with it…
Brod1er
2 years ago |Clearly an International Standard is not very standard internationally!!!
Do they do this just to make it look like it handles high ISO? Tut, tut!
amne
2 years ago |According to their test Digitalkamera DIWA lab use DXO for their measurements (see last paragraph), that would explain why they find a difference in ISO values – same thing has been found by DXO for all the recent PENs as you already mentioned.
Steve
2 years ago |+1. Something that many seem to conveniently forget or ignore.
deluxer
2 years ago |Samsungs aps-c sensor, is one of the worst on the market, nevertheless good news for olympus, i just can’t convince myself ordering one for some strange reason, being an OM nut, the magic is lacking to me, i will order a nex 7 to use it as digital back for my old om lenses.
the more i see output from the nex 7 there more it’s clear to me this camera is a game changer, the e-m5 is a very good update for those with many m43 lenses, and well deserved for the (much beaten over the last years…) oly crowd
lnqo-M
2 years ago |I hope DXO be soon finish by test them last camera so is come now.
Me
2 years ago |> This means that the E-M5 has a 2/3 stops difference from the official Olympus ISO value.
That’s true for all PENs, too. And also one reason why the PEN’s SNR always seemed better than on Panasonic cameras using the “same” ISO value. Can be seen at DXO also.