My friend Tyson Robicheaud (Click here) compared the [shoplink 29074]Olympus E-M5[/shoplink] with 16 Megapixel Sony sensor with the [shoplink 22077]Panasonic G3[/shoplink] and 16 Megapixel Panasonic sensor: “I really wanted to see how these two sensors compared to one another as I have been very impressed by the G3.” Check out his website to learn more about the test!
Meantime I remind you that on July 18th we will see the new G5 which has the same G3 sensor but a newly improved processor. Let’s hope the new GH3 (again 16 MPX?) sensor brings a major improvement in RAW quality too!
Olympus E-M5 price and in Stock status check at [shopcountry 29074].
Panasonic G3 price and in Stock status check at [shopcountry 22077].
42 Responses
Sorry for being a little off topic, but any news on the new PENs, Admin?
Also, on topic, I agree that the G3 has always been a good deal compared to other m43 cameras.
Yep, i have some news coming…
Since Sony makes sensors for other camera companies and Panasonic only makes them for themselves and Oly, or at least they did, I suspect Sony’s R&D is going to move even farther ahead of Pany, now that Panasonic lost their only customer in Olympus. It could get a little bleak in the future, as far as sensors go, with a shortage of R&D monies. I hope I’m wrong, because I think Panasonic makes a better product than Olympus does. I guess this fall will tell the story.
So, where is the evidence that Sony made the sensor for the E-M5?
This topic http://forums.dpreview.com/forums/read.asp?forum=1041&message=41952049 seems to think otherwise…
I know, isn’t it incredible that the president of Olympus (Mr. Bamboo) went on record making bald-faced lies that Sony makes the EM5 sensor!?! He even got his Oly underlings to back up his statement when it was subsequently fact-checked by the Japanese press! Isn’t it pathetic Sony and Panasonic refuse to publicly denounce his lies and have allowed him to “save face!”
Lucky for us, we have very-nearly-bilingual forum pundits ready to reveal this latest Olympus conspiracy.
Actually, its a Fuji.
> Lucky for us, we have very-nearly-bilingual forum pundits ready to reveal this latest Olympus conspiracy.
Or idiots who think automatic machine translations work between two very different languages.
“new products bamboo biological microscope president”
As bad language as Finnish of international spammers.
You should read the whole thread before calling someone an idiot.
What is this latest conspiracy nonsense. Because oly had a few bad
seeds in its employ now the company is conspiring because it’s tired of the same irrelevant sensor question. If you’re gonna criticize the whole company for unscrupulous activities of a few employees, well better condemn every corporation. They are all the same.
+1, haha.
well said and I have been actually confused.
Exactly!
Where is the evidence?
Can someone compare EM5 (Panny) sensor with a known Sony 16MP sensor and show us any similarities that EM5 (Sony) sensor shows when compared to G3 sensor.
I bet they will find that EM5 is way way more similar to G3 that any camera with Sony sensor claimed to be in EM5. Why would that be if EM5 sensor is Sony? Huh?
What irritates me is that there is one group of tests attesting the EM-5 a phenomenal DR, while others can’t find any advantage regarding DR compared to the G3. Like this particular test, which even finds a slight disadvantage for the EM-5 in the DR chapter: “while the whole of the histogram for the G3 is shifted left, it has clipped less of the shadows in the blue channel and better preserved the highlights while handling the dynamic range within these particular shots according to the histograms”. A slight advantage for the EM-5 concerning resolution, paired with higher susceptability to alias effects, is generally conceeded, but for DR, the picture stays diverse. What could be the reasons? Is the variance in the G3 and EM-5 samples tested that high? Or is it badly controlled experimental conditions?
DR is difficult to measure and there are a lot of incompetent testers. You can go for a “mathematical” measurement of the sensor like DxO do and that might be the best solution, but it doesn’t say much about how the noise “looks” and e.g. the GH2 sensor does much better on real images than the DxO score indicate.
Most tests test on finished images, either JPGs straight from the camera or an image converted from RAW. Either way the data has been manipulated by the camer or the RAW converter in som unkown way and the results can’t really be trusted.
For one thing that landscape picture has clouds moving fast so unless you put both cameras to tripod and trigger them simultaneously there even wouldn’t be any change for having pictures with similar content.
Like spam said there has been some talks in relation to delays associated in getting support for new cameras that big commercial RAW converters might include some camera maker based automatic presets. (so hobbyer made RAW converters would be more neutral)
And like article itself explained JPEG is seriously bad format for this because data needs serious cropping/compressing to fit it into 8 bits.
16 bit TIFF would be right format for comparing histograms.
But if you want to compare processed picture’s here’s some testing with purposely misexposed images for seeing how well you can correct them in RAW conversion which shows dynamic range of sensor in highlights/shadows before data gets useless. (clipped colour channels or too much noise)
http://admiringlight.com/blog/olympus-e-m5-vs-panasonic-gh2-dynamic-range-battle/
Ahha.
So the sensor manufacturer being Sony is already considered as a fact. Nice. Any evidences, please? How the hell can Sony keep in secret that it manufactures 4/3 sensors from now – something it havent’t ever did before?
Sasa from Olympus said it. I don’t think you can have more evidence than this 🙂
Well, in this case it seems that I missed the rumor. FT5?
also , several tech papers here in Japan confirmed it, so it must be a Sony sensor.
but I hoped it to be a Toshiba or Samsung or anything non-Sony chip.
Jake, could you, being from Japan, shed a bit of light on the adequate translation of the interview? The automtic one says “Sony made it” – not developed, designed, or anything synomyc. Is this a weakness of the automatic translation or has Sana carefully avoided any term pointing to design responsibility taken by Sony?
Here is the info: http://translate.google.com/translate?sl=ja&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fdc.watch.impress.co.jp%2Fdocs%2Fnews%2F20120703_544400.html
“How the hell can Sony keep in secret that it manufactures 4/3 sensors”
For some reason, also Olympus seem to keep that secret. Or maybe they just think that is irrelevant or why should people care…
because of the business deal between Oly and Sony, they kept it secret.
1. Buy an E-M5 and a G3,
2. Take out both sensors,
3. Compare the design rules of single pixel under microscope,
Better yet…
1. get an e-m5 and/or g3 if don’t have one yet.
2. find a buddy who has or can use the other
3. go out and take some pics
See results and decide. If nothing else, you had some fun taking some pics with a friend.
Now that’s a feasible statement, Full of possibilities, but there again, what about an argument on the cameras. Without using them of course like lots of others do? Or discuss the need for a 1000mm 0,95 lens at $250.00 or 200 MP sensor?
Come now be reasonable here.
I remember those days when the Admin used to test himself some of the MFT gear (for example, SLR Magic 12mm f/1.6 lens), and didn’t try to call Olympus camera as made by Sony (or by Panasonic) just because of the sensor factory 😉
BTW, I am typing this on my Sony Vaio laptop with build-in “Ricoh Co., Ltd Visual Communication Camera”, Intel MB and processor, Fujitsu HDD,…
The comparison is OK, the photos from both of the cameras are fine. The most surprising part for me is at least twice difference in price between G3 and E-M5. However as we know, the price drops for a camera body quite fast…
It is an interesting Test. A fotographic view on RAW quality. Thanks to Tyson.
I tested the E-M5 thís weekend in an olympus workshop. It is a nice camera, but it has its price.
Dynamic Range and the IBIS were the only criteria for me to purchase one. And it looks as if there will remain only one criteria, the IBIS for non-stabilized lenses.
The E-M5 is a very customizable camera. But energy consumption was annoying, a tribute to the IBIS I suppose. Every E-M5 was equiped with the additional batterygrip. Nevertheless a lot of participants ran out of battery during 2,5 hours of intensiv testing and were frustrated.
Also the noise was disturbing to me if you make fotos in portrait-orientation with your left eye. (and in that way the left ear oriented to the body:-))
So, nothing I would pay the price for (E-M5 + Batterygrip) at the moment.
Let’s be honest, the difference is night and day. The Sony sensor is worlds apart.
Reading the above article I come to a somewhat different conclusion. 😉
I will skip the EM-5 for now and get the new 12-35 instead.
Skip the 12-35 and buy the E-M5 with the 12 mm f/2.0 – a phenomenal combo.
+1
So true!
You didn’t read it.
Or at least failed to understand when it itself goes to explain that getting 8 bit JPEG out of 12 bit RAW requires whole bunch of (undocumented) manipulation of it.
Here are DxOAnalyzer results with comparison to two known Panasonic sensors (old 12MP and its 16MP derivative) and two known Sony sensors:
http://www.techradar.com/reviews/cameras-and-camcorders/cameras/digital-slrs-hybrids/olympus-om-d-1075717/review/page:5
Competing Imatest results with major increase especially in high image quality threshold DR.
http://www.imaging-resource.com/PRODS/GX1/Z00160_ACRauto_Step_2.png
http://www.imaging-resource.com/PRODS/omd-em5/Z00200_ACRman_Step_2.png
And non benchmark test for how much DR there’s in highlights and shadows in comparison to known Panasonic sensor of more advanced lineage:
http://admiringlight.com/blog/olympus-e-m5-vs-panasonic-gh2-dynamic-range-battle/
Either it’s Sony sensor or Panasonic finally managed to make modern sensor.
Just a question regarding the Techradar results:
– JPEG DR is higher at ISO 1600 than at base for GX1 and NEX-7? Maybe I should start shooting landscapes at ISO 1600 then?
– Also, EM-5 having more DR than X-Pro 1 … is that a confirmed result?
Those oddities are caused by JPEG being useless for doing trustworthy comparison of sensor’s performance.
On top of all cropping, folding, stretching and twisting of RAW data JPEGs are processed with varied amount of noise removal which further messes results because dynamic range consists from brightness range between darkest shadow whose signal is certain threshold above random noise and brightest highlight before pixel is saturated.
What that JPEG curve might mean are bigger changes in noise removal processing settings.
Of course it might also have something to do with quirks of Panasonic’s in camera JPEG processing which isn’t considered as good as in competing cameras.
Any way that kind zig zag curve isn’t natural and would be same like your car’s RPM meter suddenly jumping from 2000 to 5000 and then returning back to 3000 to continue rise while you kept smoothly pushing pedal deeper.
I agree, 12-35 f/2.8 is a good choice!
Interesting article for me because I’ve had the G3 for over a year and have been contemplating buying the EM5 (OR the 12-35/2.8-can’t do both now). I’m a raw shooter using Lightroom 4.1 and, like a lot of others out there, have been seduced by the love felt for the handling, esp. with the grips, IBIS, good high ISO, better DR, etc. of the EM5. I had resolved to hold out until Photokina to see what Panasonic does (I’ve been Panny since the G1 and GF1) but I’m a stills shooter so looking for both the best IQ but also the “experience” the EM5 folks tout LOL. Hard to be rational sometimes…….just read the comments above
This test is a good example of why amateurs shouldn’t publish test results. It’s just so flawed. He comments about differences in color without apparently realizing that every raw processor applies a standard profile FOR EACH CAMERA. The color differences he sees are as likely to be a result of the raw processor as the sensor. (And he doesn’t seem to disclose the raw processor, which makes it even harder for us to judge his test results).
And he sort of acknowledges that the OM-D underexposes relative to the G3, but doesn’t account for that in his tests.
On balance, the weight of the evidence from multiple tests is that the OM-D has at most a 1 stop advantage in noise and DR. Which is a good thing, but hardly revolutionary.
Thanks to Tyson for this interesting test comparing the EM-5 and G3 sensors. As stated the processing has as much to do with the IQ output as the sensor does and that is where Oly has excelled. I am not surprised to see the 1 year newer EM-5 output with a new sensor besting the G3 and many prefer the brighter more vivid colors put out by the Oly. However test reviews generally have noted the Pany colors while muted are more natural and if you select vivid or other option are much like Olys brighter colors. The new GF5 shows these new colors so very likely to see them in the new G5.
I noted the OM-5 sensor output in this review reportedly displays more artifacts and although sharpened further presumedly by weaker filter looks more pixelated when cropped?
I am surprised to see the repeated rumor that the new GH3 will have a 16mp sensor as the existing multi-aspect one is 18mp. Some have even rumored that the GH3 will be using the same Sony sensor as the EM-5 or is it a new sensor which will be shared with the G5?
Uhm, very unique review. I completely diagree that from my experience (of my G3, GF1, GX1 AND E-M5).
These comparisons are very funny.
Actually, E-M5 sometime surprise me with funny WB. Its actually slow in AF (I don’t believe Olympus ad: fastest AF.
)and non accurate AF (in any condition, Panasonic accurately focuses, but E-M5 often lose in some specific conditions),
BUT
E-M5 is about 2 stop low noise, 1 stop latitude wider (GX1 is very severe in EV comparison. if you set 1/3 more EV, white details begin loosing, but in E-M5 at least there is 2/3EV tolerant, that means more darker details can be captured.)
I never believe Panasonic semiconductor division can catch up with Sony’s sensor in 3 years (because of semiconductor development nature that I experienced.) This difference has been know by Panasonic ever (yes, that’s also my experience. A company always test the competitors’ product!). So current satiation is that Pana cannot do anything against Sony. In 3 year later, Pana’s sensor may becomes “current” Sony’s, but Sony’s run further ahead. Still, the sensor can be evaluated. Sony’s competitor may be Fuji. Olympus can choose Fuji, too.
I have no idea how Panasonic should do (just improve its sensor? ha? if it could, Lumix G series would be far better, before Olympus makes E-M5 “shock”.
My current conclusion is Pana failed in sensor strategy, and m4/3 product strategy. I don’t know if it affects to “m4/3” or not. For such failure, I have no any idea than just saying “Pana failed”.
But Olympus also is not a perfect company, especially it has finical issue. Pana’s sensor strategic failure is very serious.
Thanks to Tyson for a good article. He acknowledges it is not a scientific, but rather a pragmatic comparison. I have been using the Panasonic cameras from October 2008, and I would prefer to stick with this manufacturer.
I’m completely happy with the GH2’s performance for HD video (and now almost exclusively use the Voigtlander 17.5 lens for video). I’m less happy with stills, but I do shoot everything RAW and process in Lightroom 4.1. Others say the images are very good, but I still find the DR response lacking, and the colour depth weak.
I’m holding fire on new camera purchases until after Photokina. I don’t want to change the platform, and would prefer cameras that are equally competent in stills and video. The EM5 certainly “ups the ante” on stills, as did the GH2 on video. Canon’s MILC entry could also be a game changer.
@Admin – where has the Login gone? (I’m using Google Chrome)
> My friend Tyson Robicheaud compared the Olympus E-M5 with 16 Megapixel Sony sensor with the Panasonic G3 and 16 Megapixel Panasonic sensor
To me that just reiterated a well known fact that the G3 sensor is precisely at the “good enough” mark.
Better sensor is better, but for the 90%+ real world uses, the difference would be sufficiently negligible. (OK, occasionally the remaining 10% can drive user nuts. But even so, the sensor is only one of the many things which drive users crazy.)