CameraLabs posted the Olympus E-M5 vs Panasonic GX1 RAW noise comparison. While doing the test CamerLabs discovered something interesting: “But when framing and later analysing images from both cameras, it became obvious the E-M5 was capturing a fractionally smaller field of view to the GX1, suggesting a fractionally smaller sensor (…) The actual difference in field of view is sufficiently small not to have any significant impact on quality or coverage, but it does reinforce Olympus’ claim that the E-M5 uses a new sensor. Or at least uses it in a different way to preceding models from Panasonic.”
From a pure photographic point of view that question ins’t important. It would just have been interesting to know why Olympus is making a hype around the sensor by simply not disclosing the manufatcorer. Anyway, speaking of good photographic articles. As suggested by our reader “Dummy” (Thanks!) you have to read this: “Show Me Photographs, Not Gear: How to Improve Photography Without Upgrading Gear” by Robin Wong.
—
E-M5 shop search links at [shopcountry 29074].
GX1 shop search links at [shopcountry 26836].
85 Responses
Why are they doing it? People are still talking about the sensor! It was a good move.
MY sensor can beat up YOUR sensor! 😮
Calm down or I’ll come back with my FF brother 😉
As long as you don’t bring along the FF-equivalent ganstas! 😉
hehe…I call FILM!!!!!!! *film ”FILM SMASH!!!”
It could be because of IBIS?
Who cares unless you are in the industry.
The TRUTH:
One of the Olympus engineers steals a small piece of each sensor during the manufacturing process, takes them home… and is making cameras in his garage to augment his income. Its so obvious.
(I think he calls his side business Casio Cameras?).
It’s the 5-axis IS I bet, uses less of the senor as a result.
Definitely… Crop might be defined by amount of stabilisation used.
It’s not digital pixel shift fake stabilization.
Still, when your sensor moves (say, to the right) some border (in our case, right border) pixels get less light. To compensate that, they could crop a the picture a bit.
But cropping not only reduces the FoV, it also reduces the amount of MP. I assume that they mean that both images have the same amount of pixels, but different field of view, which does suggest that the size of the sensor is different.
The sensor was made over size to compensate for IS movement, I can’t find it at the moment but when the e-m5 was released this 43 rumours blog gave the sensor measurements, they were larger than the panny. The same thing came up about a month ago, the size had not changed in that time! That was again during a panny comparison.
There may be an electronic reason for the pixels missing, it may indeed be due to the IS, Bob B may also have found the Truth!
But it is insignificant it is less than the difference between lenses of the same focal length of many manufacturers. Leica and Zeiss (for example) have lenses of a particular focal length may have a small difference in viewing angle, therefore FOV changes. It really is irrelevant if your happy with what you see in the viewfinder, that is what you get it come off the sensor, If not change the lens.
Only by half of a step… I’ll stick to my G3.
Some people think DR is more important than High ISO. Where the difference is much larger to G3/GX1/GH2 cameras.
It is comprehensible. The 5 axis IS needs room to compensate the sensor move moment. The cropped angle should be the safety area. Panasonic GH2 used to be 18MP sensor technically but 16MP for effective area. think it is really not point to aim on discussion. It is good, no doubt. But it may explain why the when I am using my Pentax 25mm CCTV lens it get less vignetting.
Interesting theory but if the sensor moves, it should be an oversized sensor (as with GH multi-aspect ratio sensors), and the part that captures light should be the same as any m43 camera.
Also, the PEN m43 cameras have IBIS (just not 5 axis), so your theory should apply to them too.
The big question, is the Olympus or Panasonic so is more close crop.2.
The big question, is the Olympus or Panasonic so have crop.2.
Anyway this so is important, sensor in E-M5 have more DR.
Another paid shill of olympus. Too stupid to open up the camera to look at the sensor for the make?
Good point, why not just take off the lens and look at the sensor. It has Sony stamped in big letters right there on the front of the sensor.
+1
Oh no! The EM-5 sensor is smaller? A crop m43 sensor? Now we really won’t be able to get any shallow depth of field! Remember f/1.4 on an EM-5 is really f/2.8 on a GH2 is really f/11 on a full-frame camera. I’m switching to a Nikon D800!
You forgot the sarcasm smiley… 😉
Why am I not surprised? I never believed that the GX1 and E-M5 sensors are identical, and there is a lot of evidence available for this assumption on the internet.
If you can open the camera and see what made the sensor is, do tell.
Forget it, that’s a crop as well.
Wow, they must have a precision tripod mount. The lens are aligned so accurately that a few extra lines are noticeable. Did they take the Earth’s curvature into account?
maybe its the work of the IBIS
maybe it doesnt have to do with the sensor size, but the difference in the shape of the camera bodies (ie the position of the sensor in the camera body).
different body shapes may result in a marginally different field of view as this may effect the distance between sensor and subject (if ever so slightly)…
No, flange distance is part of the standard. You can’t mess with that if you want everything to work correctly.
Theyre still different body shapes though, im sure this would make it very difficult to get exactly the same image when mounted on a tripod (without adjusting the tripod in some way)
But no one has standardized position of tripod mount making it hard to get absolutely same framing/perspective with two different cameras even if lens is same.
They should have used some 4/3 tele lens with own tripod mounting and then attached body to lens for eliminating that variable.
Yes! I agree, my point exactly! Seems a bit silly to blame sensor size…
I’m going for this theory too.
All it takes is a few mms back on the body for the tripod screw.
Exactly!
The OM-D has the same sensor as the GX-1!!!
THEY MUST BE THE SAME CAMERA!!!
I’m not even sure that people realise that not all camera makers make their own CCD, just like they don’t make their own LCD and processors.
LCD and Plasma panels are generally only made by one or two big companies, while there are a heap of TV brands out there.
Yes, that’s why I’m still happy with my GX-1. No need to move to E-M5 at all.
Well, my comment about them being the same camera above was sarcastic. They are quite different cameras. The CCD is about a fifth, or less, to do with overall performance.
It seems that some people think that it’s 80% or higher to do with results, but it’s not. There’s no point in trying to say they are the same camera, even if they share the same sensor. Even though they don’t.
Panasonic MAY have manufactured it, but who gives a shit?
CCD?? The Nikon D80 used a CCD, I think. Since then, most cameras have used CMOS, with Oly/Pana using their own version.
Yeah, well, sensor then! 😛
And SONY made the OM-D sensor. 🙂
Whoosh… I think the point just flew over your head.
As Bbking said, if sensor was most important thing I’d be using Fuji X1 Pro.
Deleted
And yet the images do differ between brands. Theyre made by the same factory but different designs and specs, i assume. Otherwise we’d allbe watching palsonic etc and getting the same quality..
Seems pretty logical because Panny and Oly don’t do the lens corrections exactly the same. I remember back to the reviews of the E-P1 where Panasonic fully corrected distortions and Olympus left them slightly under corrected. This will make a difference in the final image.
Software correction of lens aberrations definitely has effect to framing and used lens has that m4/3 standard higher than good geometric distortion covered up in software.
Even Camera Raw might well use slightly different settings basing to preferences from Olympus and Panasonic. (didn’t m4/3 include camera saving wanted correction parameters to RAW?)
Also camera’s JPEG-engine is often set to crop image size to multiples of 8 or 16 pixels (depending on use of chroma subsampling) to allow lossless rotation while raw converter might give you all pixels.
And E-M5’s sensor officially has little more pixels and outputs slightly bigger images, 4608 x 3456 vs 4592 x 3448, so cropping fixed size image from that should result tiny bit narrower FOV.
There are simply too many variables which can all have compounding effect to say that sensors are different size.
best theory so far.
Maybe they are under a non-disclosure agreement? Maybe they can’t say who manufactured the sensor because it’s from a competitor that doesn’t want micro 43 cameras to take away from their own offerings?
They might have a NDA, not likely though. What admin seem to forget is that the manufacturers rarely tell who made the sensor. They simply don’t comment on that.
So when “Olympus is making a hype around the sensor by simply not disclosing the manufatcurer” then that’s actually what they almost always do (or don’t do).
Even smaller than 43 lol………….
Lol… but what are you doing in this forum ? as you clearly don’t like 43, you must have huge time to waste…go take some photos with your FF and share it if they are good !
I am guessing it is an Aptina.
There is absolutely no evidence that the E-M5 would have a physical sensor inside. I think it rather has just a simulated one (obviously by software).
I’ve looked, there is a sensor there, pretty strong evidence IMO, and if you read some reviews you’ll find that quite a few have checked.
Oh, no. The thing you’ve looked on is just the AA filter. You cannot be absolutely sure that there is actually a real sensor behind it. 😀
The AA filter is glass (of course) and is invisible (or pretty close) so what you actually see is the sensor. I’d like a virtual one though as it would solve a lot of problmes (like dust problems).
it’s pretty obvious by now that the whole camera industry is virtualised, but they give you physical representations of your purchase just because they know the market isn’t ready for that yet.
That test doesn’t say anything about the sensor size, that’s just the output. The bitmap could be just smaller because of the sensor shift, they probably have to crop it. Otherwise borders might be outside the golden center of the image circle.
Um… btw… stupid question.
I got my hands on an 12-50 Zuiko yesterday. And while I know the Zoom is always motorized… the manual one feels kind of scratchy. Is it normal to feel like there’s sand in the gears ?! Or is the lens defective ?
(I cannot really test the lens… as I am still waiting for my E-M5 to arrive…)
It is normal. It feels like the manual zoom has to move the motor of the motorised zoom against it´s will.
Ahh alright. Thx for the answer.
I’m in the store again, today, however, so I’ll compare it with another unit. Just to make sure.
Mine is like that too. It’s the missing chips off the sensor grinding in the gears.
Olympus might have reasons (PR) to not talk. But if it is the GX1 sensor, it would be in the best interest of Panasonic (camera division, prob. not sensor) to say so.
Why would it be in Panasonic’s interest? Saying so would only make sense if they considered the E-M5 a benchmark product they’d have to reach at. But their perspective probably is different: They do likely consider the E-M5 a competitor product trying to line up with the GH line. And obviously, you would only dare to say that you share features with a competitor if you have the feeling that it’s your product line which is lacking behind.
Panasonic is under NDA with Olympus which prevents them from telling who’s sensor it is. That’s actually the only thing that is known for sure.
One has to wonder though why would such an NDA exist in the first place if Panasonic had no relation to producing the damn thing.
Olympus has a totally new sensor, produced to new specification which takes into account the distance from lens nodal point to any point on the sensor. As we must accept that distance will be greater in the corner than it is in the center, there is therefore a slight change of focus, Olympus has taken this into account and has designed a bowl shaped sensor. However the technology has not stopped there as the sensor has a flexi base which will change it’s shape to allow for the different characteristics of each lens used, therefore if cameralabs performed this test with a different lens they may find a miniscule difference in sensor size from lens to lens.
Yeah, I like that one. 😀
Its a water bubble sensor to go with the water bubble lenses they are going to release later this month! Perfect match of lens and sensor every time.
Actually the IS does not move the sensor either, what one sees is the water being moved around in the bubble to compensate for movement as well as lens characteristics.
The main conclusion is the same: there are no major differences in RAW of GX1 and E-M5. I care less who made the sensor.
For me, the form factor of GX1, which is following the GF1, is preferable in comparison to the cameras with integrated EVF.
“The main conclusion is the same: there are no major differences in RAW of GX1 and E-M5”
Lol !!! I’m really wondering from where (sensor size ???) or how you make your conclusions because based on the tests I read so far, it seems that E-M5 RAW are signicantly better than GX1 and closer to Nex-5N…Dxo will give the finall word but I bet that that E-M5 will be far better than GX1 !
You’ve missed my point.
I shoot in RAW, and based on the RAW comparisons the sensor is almost the same for me with the minor variation probably resulted from the thinner antialias filter. There have been numerous speculations on what kind of Analog-to-Digital Conversion have been applied on the similar looking sensors, and the way the Panasonic and Olympus marked their ISO zero-points.
I’ve tried myself both the E-M5 and GX1 in RAW-data mode, and my personal conclusion concurs with the review of the Cameralabs.
It’s all important technical details, but my point is that the difference is not large enough to base your choice solely on the sensor imaging qualities.
So sounds like you couldn’t care less. I was wondering what you care less about…
I thought it was a new sensor but after checking the review out , I am confused and there is no previously reported one full stop advantage with the Only over the Pana, and in fact, the way they renders blue and brown like red are exactly the same to my eyes and I think if the difference is this small , I must get the Pana to save 500USD.
I have to agree with jake and Igor. Seeing this real life test just put my E-M5 pre-order in check. Also I must take in account what VK wrote a couple of days ago: “I am working on next PTool that will have support for G3 and GX1. Hope it’ll be available in June.”
Ah, well.
According to Pekka P, pictures taken with the 75/1.8 can not be released until after June 11.
I don’t think this is because of the IBIS : Oly claims the EM5 has a 16,9 Mpixels sensor and the GX1 has 16,1 Mp. The 800 000 pixels should be because of the IBIS. So anyway, it is a new sensor, never used anywhere else. But it could just be a bigger GX1’sensor.
DxO, come on, we want your results !
The numbers concerning sensor sizes are actually wildly varying even for sensors known to be identical: According to DPReview, the sensor in the G3 features 16.6Mpixel total and 15.8 effective, while in the GX-1, that very same sensor is claimed to feature the same 16.6Mpixel total, yet 16 effective, despite both cameras offering exactly the same set of selectbale resolutions.It does, however, seem that Olympus delivers a marginally higher maximum image size with 16 more columns and 8 more rows at maximum resolution (accroding to DPReview). But as this is just 0.3% wider and 0.2% higher than the maximum G3 / GX-1 image, it certainly is within the tolerances of the M43 standard concerning frame size and by no means an indication of a sensor of different density.
I regret to repeat an over-stated sentiment, but the sensor inside the E-M5 (or any other camera) is interesting to me just like old shoes. Lenses, focusing mechanisms, processors, sensors, AA filters, viewfinders and strap lugs all have one purpose: to help the photographer transfer their vision into physical existence as seamlessly as possible. What does the identity of the manufacturer of one little piece in the jigsaw have to do with that?
+1 (followed by more zeros than I can be bothered to type). Let’s call it a nice round googleplex.
Most likely a combination for ibis and the distortion auto correction. I know that because picasa do not support the panasonic distortion correction, the jpg always has less fov compare to my raw file. Each camera autocorrection for each lens differently, so it is not about the sensor size but how the processor process the info.
The images at Camera Labs look similar but are obviously not identical. Could it be that the cameras actually use the same sensor but with different camera processing? And why is the brand of the sensor such a secret?
Do you care who made the lightbulbs in your car, or the hinges on your fridge door? Course you don’t. So why does anbody give a shit who made the sensor? It is not secret, just nobody cares. All that matters is whether the sensor does a good job or not.
You assume too much. Of course I care about the low- and high beam lightbulbs in my car. Not all bulbs perform the same, and I’m picky. I don’t care about the blinker bulbs, but I’m sure someone does!
On my E330, which is nominally a 7.5 mp sensor, if you use RT (Raw Therapee) to process the raw files, you can actually get 8 mp images. Sometimes the edges got sort of weird with color shifts and pieces missing but generally speaking with some judicious cropping you can get very close to 8 mp.
This is getting kind of dumb. Why make a public product review with funny speculations about what’s inside it when it only takes a screwdriver to get right to the truth? If you don’t have the balls or budget to screw (literally) the warranty, then just shut up at wait until others pray the thing open…