DxOmark (Click here) just published the Panasonic G5 sensor score. The full test report should be posted soon. As you see from the image on top DxO still has to post the E-M5 results which they once unveiled to be “surprising”. Can’t wait to see this!
Panasonic G5 search links at [shopcountry 32274].
82 Responses
Ok – it’s not a Sony Sensor – the differences are not that high. But a little improvement is still there.
The Sony NEX 5 startet with 69 Points.
your comment likes SYSmark 2012 and AMD was resigned for BAPCo.
if not intel inside score so bad.
The E-M5 probably tested too good, and Nikon threatened to withdraw their payments.
I’m always asking myself whether people like you actually manage to believe their own lies or whether they just lie to get the satisfaction that insulting others brings them.
DxO isn’t a human being, but on the other hand, you just went on a full-scale rampage to insult me and call me a liar. Pot, kettle, psychological projection, all that stuff. Is that you, Amalric?
You started the defamation. You claimed that DxO’s results are payed for. If you start insulting others, be prepared to get paid back in kind. Why should we try to reason with you if you clearly are not interested in or even reachable by reasons?
You really don’t understand the difference between personally insulting a person and questioning the motives of a large company? (The innuendo in this post is meant to personally insult you, just for clarification.)
Of course insulting a corporation is less personal but you’d still insult the persons at the helm there. And a lie is a lie, whether it is about person or a corporation.
And how do you point out that somebody is lying without ‘attacking’ him? And I am really curious as why these lies are being put forward. Because I am really puzzled by that behaviour. And I think figuring out why somebody is lying is much more worthwhile than just saying: ‘This is a lie.’ Because to stop something (ie, the lies), trying to get to root of it is often the better approach.
El Aura take a chill pill man. It’s totally fair to question the motives of DxOMark since they haven’t even so much as given an explanation of why it’s taking they so long to publish the sensor results. If they want to retain their credibility they need to stay in communication with their readers instead of keeping everyone in the dark.
DoX may be accurate, the final result is every bit like false precision. To be taken with a huge grain of salt and there is no reasons to make this personal.
El Aura, the internet has spoken. Shut up.
If something looks like shit and stinks like shit, then it’s very probably shit.
TechRadar which uses DxO’s method and software got very good results for E-M5 both in noise and DR. Imaging Resource which uses competing Imatest beside visual inspection noted same from camera’s performance.
Dpreview’s RAW comparison shows E-M5 being fully competitive to APS-C sensors and both DR and noise performance have been noted as good in many other places.
And considering size of pixel (almost same as in Sony’s 24MP APS-C sensor) and Sony’s excellent technology that’s precisely what results can be expected.
All the while DxO keeps refusing to show results pleading to some mysterious problem no one else sees.
So I sure fail to see auras around shit!
You fail to see that 100% QE somehow is not a plausible result?
If you use THEIR equipment, and you get a different result from DxO, like TachRadar does, and if you use your EYES, and compare like ImageResource and DPreview, and you get the same result like TechRadar. And it differs, why defend DxO?
Please, look at the rawfiles (downloadable from IR)from Sony A900 and Nikon D600, and tell me how many stops of noise they are apart? 3 stops? 4?
Then go to DxO and try to believe what they say when they claim that they are only 1 (one!) stop apart at iso 6400!
Someone is doing a mistake. Either:
1) All other reviewsites, including those who use DxO exuipment, and your eyes
Or
2) DxO
What is your answer?
Techraders results are a complete joke and this is obvious to anyone with half a brain
He said ‘probably’, which is supposition – which means he thought his statement was a likely possibility. it therefore was not a ‘lie’, because he was not asserting that something was true.
That’s just the technical aspect. The main point, though, is that what he said was what you would call tongue-in-cheek-humour. Something said for the sake of humour, which the speaker themself does not really believe, or expect others to.
I have a feeling it is the opposite the reality of those pesky RAW files with no devious NR showed the E-M5 to be a minor step forward. Not the mighty leap forward the Olyboys claimed. DXO being decent chaps are afraid of a mass suicide and therefore are reluctant to post the results which could tip this unstable group over the edge , perhaps literally 🙂 .
It’s not in camera NR. Check Dpreview’s RAW comparison.
Completely expectable behaviour of noise and its increase when increasing sensitivity for tech what’s available.
But Fuji X-Pro 1 is definitely doing NR and lots of it even to RAW unless Fuji made Adobe to apply always heavy NR in ACR. It’s known to be Sony sensor and slapping different pattern colour filter array in front of sensor changes sensor’s noise performance as much as painting car with different colour changes its top speed.
Though that isn’t suprise considering Fuji started experimenting with these heavy in camera NRs already almost ten years ago. (and then claimed having revolutionary sensor just like now)
Changing the pattern of the colour filter array does indeed change the noise behavior wrt. any reasonable definition of noise. Just think of it: The patterns have different densities of red-, green-, and blue-sensitive cells. If this wouldn’t affect noise in the respective colour channels, then why not go to the extreme and just invest a single red pixel for the whole sensor? Because a higher number of respective sensor cells delivers a better signal-to-noise ratio. Well, the differences between the Bayer array and Fuji’s pattern are not that extreme, but you should still expect the Fuji pattern to deliver less noise in the green and more noise in the red and blue channel.
Ratios of different coloured pixels changed only by tiny amount: 20 vs. Bayer pattern’s 18 green pixels in 6×6 pixel area can’t increase luminosity sensitivity much.
I miss not only the OMD. Also Fuji, Samsung and Sigma are still absence.
According to Techradar (using DXO-methods) the OMD-Sensor is overwhelming.
Other tests say, that Fuji can easily compete with the MARK II or even Mark III.
And there are tests, that say that the Sigma-Foveon surpasses in Raw and low ISO situations all FF-Cams. Not bad for such a camera.
DXOmark tests only the booooooring mainstream – the mass-market. They don’t see the real innovations.
That’s a absolutely wrong business-strategy, although most earnings come today from the mass-market. But tomorrow?
If I look on TechRadar, the OM-D outperforms most cameras on RAW while it is worse than all cameras on JPG. This leads me to believe there is some RAW noise reduction going on… making it impossible to test the camera. The dynamic range is as impressive as all recent Sony sensors.
For Fuji I guess the problem is that there are no RAW converters. Either way I have serious reservations about the DxO conspiracy theories.
Dear Admin,
It would be better to compare to GH2 results as there are surprising similarities and the result is also quite close. So I wouldn’t be surprised if the sensor was the one from GH2.
(The just a bit higher score may prove that there is a little tweaking on the original GH2 sensor though).
It’s almost certainly not the GH2 sensor – it’s the wrong size and aspect ratio, for a start. 🙂
G5’s total and effective pixel counts are same as for GH2 if you check product pages.
And these results basically confirm it being based to GH2 sensor with some tweaking.
Limiting output to only 4:3 aspect ratio in firmware shouldn’t be any problem because camera’s operation is controlled by firmware.
+1, seems very likey it’s the GH2 sensor or a variation on that design.
> So I wouldn’t be surprised if the sensor was the one from GH2.
It IS the sensor from the GH2. More than obvious not only because of the same characteristics, but also because of the same total pixel count of 18 MP, which is of course due to the GH2 sensor being oversized for multi aspect and they didn’t change the size for use in the G5.
+1
It’s very probably GH2 sensor – same pixel count, similar results. A bit disappointing they’ve put 2010 sensor in a 2012 camera, limited it to 4:3 aspect ratio and improved it only slightly.
It even has worse high ISO performance than GH2 and G3. Better DR than both is a plus, though.
until the E-M5 Olympus had the same sensor crap 12mp sensor in everything from the E-620,E-30 and E5 on through the PEN range all of them. The GH2 sensor is a good step up from the 12mp models and at a good price point
I am still using this crap sensor, and I don’t mind it being so crappy. 😉
I too, 😉
Good for us!
7 months since announcement and no results for the EM-5 yet.
2 months for publishing the results from the G5.
10 days for the results for the Nikon D600.
(compared to announcement date from their site)
How not to conclude that DXOmark is biased…
+100 dxomark has become a joke to me, their inability or unwillingness to show the em-5 results after so many months without offering any kind of real explanation speaks of some serious flaw in their system or even worse economically inspired one-sidedness.. forget dxomark.
Olympus is not the only one: almost 10 months after the announcement of Fuji X-Pro1 and no results yet. There are now two more cameras with E-M5 sensor and one more with X-Pro1 sensor, so biased or not, DxO will have to publish the results or they’ll loose what credibility they have left.
What credability does it have, it has none no more, they have had their moment ofglory it is time they closed down and retired into Nikons back room, DxO has become a joke throughout the industry.
The cry of bias is the longstanding battle cry of the chip on the shoulder Olympus fanboys going back all the way to the film days, who needs AF, who needs anything but a small sensor,Why don’t we top the sales chart we are the best, I want my Mommy , no one needs more than 12mp, the board is honest , fastest ever ,you get the message girls lol
LOL!
I love Olympus, I have a collection of their film cameras and the OM-D, and I thought that comment was spot-on and hilarious.
I don’t know if they’re biased or not, but it does seem incredibly odd.
Then again, waiting for raw file format updates from Apple, Adobe, Phase One works much the same way. Canikon updates appear immediately along with medium format and Pentax and Olympus updates may be a year late.
> How not to conclude that DXOmark is biased…
DxO is not biased. The DxO is “flawed”.
DPR e.g. also uses the DxO suite, but they do not try to dumbify the results into few easy-to-compare numbers – they show images instead.
Images often taken at different physical exposures and non consistent lighting.
All taken out of the equation when looking at the graphs on the DXOmark website.
If I look on TechRadar, the OM-D outperforms most cameras on RAW while it is worse than all cameras on JPG. This leads me to believe there is some RAW noise reduction going on… making it impossible to test the camera. They are probably waiting for Olympus to release a firmware update that lets them turn off the noise reduction. The dynamic range is as impressive as all recent Sony sensors.
Hope G5 will get the multi aspect function with Hack firmware.
That’s a good camera for who want multi aspect.
i hope so.^^
The E-M5 score is either surprisingly good or surprisingly poor; so why will DxO not publish the results after all this time? What’s the problem, I wonder?
Dxo are pornography for technocratic wanker, not useful stuff for photographers.
+1 (but I still like to watch results)
AT LEAST SOMEONE HERE GETS IT. Going on these sites looking for specific information is a nightmare since you just end up seeing this creepy community built around technophiles who are so far removed from photography it’s insane.
Funny. I have many friends that are well known pro photographers and none of them give a crap about sensors makers, or even know what Dxo is. They don’t need IBIS but they all take amazing photos, many under dark and difficult conditions. One laughed when I told him Olympus makes only silver lenses saying they were too reflective for some situations and distracting for some candid work at night. I have to agree, most of this talk is just that, talk. What people call “pro” photographers today in many cases is anyone loaded up with technology and photoshop to make up for a lack of skill. In the end the pictures may be better form a technical standpoint but the content sucks or is mainstream junk. Of course there are some exceptions.
whatever the issue is with DXO and the EM5 – the end result is that the reputation of DXO is trashed, its unfortunate since I have found DXO a valuable tool and previously accurate.
Those scores for dummies were bad thing from the start but DxO’s dynamic range results have been good for comparing sensors with others testing cameras in uncontrolled basically random circumstances. Now also TechRadar uses same controlled method.
But for noise performance Dpreview’s RAW image comparison tool is lot more telling of real world performance because DxO’s synthetic method completely lacks way to take effects of noise to details into account.
But that should have been obvious since they gave Canon Powershot G11 only little better noise score than G10 when those 10MP pics were lot better than the latter’s pics of 15 Marketing Pullshit.
DPreview test protocols are a joke, they have different lenses,different apertures, different DOF, different lighting levels , different exposure { they do not even bother to compare actual exposure levels. For years their DR test was done with jpegs seriously BS
I have got a full frame Nikon D700 and an E-M5. I think that the IQ is better on the Olympus…
Could it be that dxo is waiting on Olympus to donate a camera?
I don’t think that’s big enough payola…
And how they have supposedly previewed it if they don’t have it?
DXO said a couple of months ago that the E-M5 was tested but they couldn’t understand the results (likely higher than expected). So they have decided not to post them. I wonder what they will do with the new PEN series or the GH3 that are using a similar sensor.
They cannot hold out forever.
This convinced me of upgrading from G1 to G5. Excellent. Many people on sites like this upgrade their camera everytime there is a new model, but often new cameras nowadays is just a tiny improvement over last years model. Going from G1 to G5 is a big enough leap in dynamic range to justify an upgrade.
So disappointed. Even a smaller sensor size, the sensor inside Sony RX100 is a lot better with 66 overall score, better color depth, and better dynamic range.
http://www.dxomark.com/index.php/Publications/DxOMark-Reviews/Sony-Cyber-shot-DSC-RX100/Sensor-Performance
The Sony RX100 does very well, but for anything above 200 ISO, m43s are better. The 66 score is based on low iso scores almost exclusively. With that lens, if you zoom a bit you need to go up to f3.5 or f4 or so. Pushing you into that territory in less than ideal light. But that is DxO, the only thing that really counts is low ISO performance.
I also think the RX100 explains that an unbelievably high score will not prevent DxO from testing it To me, it also shows that if the OM-D uses the same tech it will of course score l a lot better. Probably somewhere near the 75 pnts (for what its worth).
If we compare the RX100 results on dpreview, we see that GH2 sensor performs visibly better. More detail in RAW and less noise in higher ISO’s. OM-D (and probably GH3) being yet another top better than GH2/G5.
I don’t think DxO will test EM-5 sensor after all….we don’t know the reason since there’s no official announcement from DxO. And I don’t think DxO still waiting for EM-5 to be tested, they already test it but they decide not to publish it for whatever reason.
If you scroll the ranking af all sensor tests you can see, it is a ridiculous test with no relationship to the real world. I don’t know what they to, but it is not interesting.
Based on my experience of G3 and Sony RX100 I belive the EM-5 will score very high: The RX100 and Oly EM-5 sensors are the same generation. The RX100 produces better quality images than my G3. Noise it a bit higher in the RX100 files at 100% on screen though. But printed the higher resolution in RX100 equalizes the result. RX100 has better colors and significantly better DR. This was my unscientific judgement. I was pleased to see tha DXO comparison RX100 vs G3. Their scientific conclusion was very close to what I percieved. From now I love DXO 🙂
Low score for little sensor. It seems fair.
The E-M5 will not be better.
LOL…
And you obviously have a little brain, so we will show more tolerance regarding your BS comments. Seems to be fair.
DxO didn’t publish E-M5 results because it would thwart their belief/religion that larger sensor has to be better.
They might post results once Canon and Nikon release new APS-C sensors which score better….
Seen Elvis lately?
Larger sensors of a similar generation are “always better” it is rather obvious. lets say Sony made the sensor in the E-M5 whoopee now lets say Sony made a FF sensor with the latest tech four times larger than the E-M5 what do you think will happen.
True, but that could still mean the OM-D has a sensor performance close to the Nikon D3200, which is pretty damn impressive and could put it pretty close to current generation Canon full frames…
i see there is progress in the ISO from g3 to g5. panasonic you gotta get pissed off and develop some faptastic sensors to compete with sony
Panasonic is electronics giant fully the size of Sony so might be that instead of R&D capabilities question is about Sony having had research farther and patented so much things which now prevents Panasonic (and others) using results of their own research.
If they use the apple methodology they will sue anyone who has a sensor that is flat, shiny, has electrical contacts and is sensitive to light.
Don’t worry, if they don’t test the OM-D, they will test the E-PL5 and the E-PM2 which have the same sensor. One day we will know, for sure 😉
Be satisfied if you shoot better images with your camera. Forget these guys, they don’ know anything about reality or maybe they don’t tell anything about reality because they are payed
Who cares about dxo, probably Olympus is not a good client:)
Even this went to moderation.
I’m glad that I made a decision to buy RX100. This test just confirmed my actual experiences. RX100 sensor is a lot better than G5.
I am sooo sick and tired of waiting DxO to make a test of E-M5! Perhaps, it’s true that the results were so stunning, that they could not post them!
Anyway, I got the camera myself and am really satisfied with the improvements over my E-P2 (which I still own). The camera suits my needs in every aspect!
“DXO said a couple of months ago that the E-M5 was tested but they couldn’t understand the results (likely higher than expected). So they have decided not to post them. I wonder what they will do with the new PEN series or the GH3 that are using a similar sensor.”
And yet strangley enough they understand them well enough to use them and support the E-M5 in their DxO raw converter! 😉
The G3 to G5 Comparison is great. You look at the DXO numbers and G5 gets 61 and the G3 gets 56. Then you actually look at the data. I don’t see any difference in the curves. Really None! Yes NONE at all. The DR one for the G3 is wonky and probably needs repeating.
So I think this is the best comparison to show that the DXO numbers are meaningless! The curves are good to look at, but the numbers are all meaningless.
I also have a feeling that the GH3 has the G5 sensor, which I think is different or modified from the GH2. As too much of the data lines up between the G5 and GH3 specs.
The DXO Marks are there, just look at this link with a score of 71, whatever that means: http://www.dxomark.com/index.php/Cameras/Camera-Sensor-Database/Olympus/OM-D-E-M5
DxOMark just posted Olympus Om-D E-M5 sensor results ….