This should be a fun comparison. I have the pleasure of having three of the best mirrorless cameras around in my possession right now: a newly acquired Fuji X-E1 with 35mm f/1.4, my trusty Olympus OM-D E-M5 and the Panasonic GH3, fresh into my hands for review. Expect full reviews of the Fuji X-E1 and Panasonic GH3 in the coming weeks. Edit 3/2/13: My full review of the Panasonic GH3 is now up. Edit 4/17/13: My full review of the Fujifilm X-E1 is now up.
Note: Since this review has been published, I have updated the Fuji X-E1 RAW crops after developing with Lightroom 4.4 Release Candidate, which has much improved demosaicing of the Fuji X-Trans sensor. Crop images show the updated processing now.
Anyway, I thought it would be fun to pop the Panasonic Leica 25mm f/1.4 on the two Micro 4/3 cameras and do a controlled studio shot against the Fuji X-E1 with its Fujinon 35mm f/1.4. Due to the different sized sensors, these setups result in an almost identical field of view, with the 4:3 aspect ratio of the m4/3 cameras allowing for a little wider field of view in the vertical direction. The 35mm on the X-E1 is slightly narrower than the 25mm on Micro 4/3, however (equivalent to the field of view of a 53mm lens on full frame vs the equivalent field of view of 50mm for the Leica).
As a result of this minor difference in aspect ratio and field of view, the crops you’re about to see will make the Fuji look like it is rendering things slightly larger. All images were taken on a tripod with 2 second timer, and all were taken from the same position.
This is a simple test scene with a few items to focus on, a color checker, and a crystal decanter to provide specular highlights in the background. Below is a few of the test setup:
As I mentioned, all cameras were mounted on a tripod in the same position. Focus was taken on the book, which is at the exact plane of focus of the right-most flower where the 100% crops were taken. All images were taken in RAW and processed with Lightroom 4.4 Release Candidate, which has much improved support for Fuji X-Trans processors. The new Adobe Camera RAW demosaicing results in a little more neutral file with the X-E1 shots, but color can be brought in line with the other cameras with a very minor vibrance adjustment.
I shot the Fuji in RAW+JPEG, and have included the Fuji JPEG as well for comparison. White balance was set using the lower left white square on the color checker. Sharpening was set to 50 at 0.5 pixels, with no luminance noise reduction and default color noise reduction.
Exposure: Those who have looked at DxO’s ratings on the OM-D may be mistaken on ISO measurements, since DxO measures RAW well satuaration for its ISO measurements, rather than the output image brightness, which is what matters for the end result. Please note that all three cameras were set to the exact same exposure value. The OM-D and GH3 were set to f/2.8, with ISOs ranging from 200-12,800, and shutter speeds increasing from 1/3 sec to 1/200 sec. The Fuji X-E1 was set to f/4 (to equalize depth of field between the two formats), with ISOs also ranging from 200-12,800, and shutter speeds increasing from 0.6 seconds to 1/100 sec. All files were shot in manual, and these settings are identical manual exposures. You will see that the Panasonic is rendering very slightly brighter than the other two cameras (which appear equal in brightness to my eye)…maybe 1/6 stop difference, and the Fuji is rendering very slightly darker. The Fuji is about 1/3 stop less sensitive than the GH3.
The crops below are 100% crops from the same area of each image. IMPORTANT: Click on the image below to open the image larger. This will darken the screen and make the image fill the screen. At the bottom of your browser, you will see some blue triangle buttons, a red circle X button, and a green arrow pointing down and to the right. CLICK ON THE GREEN ARROW to view the image full size to see the 100% crops. If your monitor is not of high enough resolution to view the full width, you will still be able to scroll around the images.
On the camera front, the OM-D and GH3 are extremely close. The OM-D may have a very slight edge here, but it’s minimal. As has been hinted at by many people, I think it very likely they use the same sensor, and Olympus usually gets just a little better performance out of the same sensor. However, the story of this test is the X-E1, which easily bests the two Micro 4/3 cams in terms of noise control by about a full stop, and perhaps a little more in the 200-1600 range. The Fuji is providing extremely clean files up to ISO 1600, and still controls noise well while maintaining strong detail at the higher ISOs. For those who know the X-E1 and know that ISO 12,800 is not available in RAW, I underexposed the 6400 shot by one stop, then pushed the file in RAW by one stop, which is what the camera does for the JPEG images anyway.
As part of this, the two outstanding normal lenses can be similarly viewed. Both are quite sharp in the central region, so I wanted to show a closer look of how these two lenses perform. Stopped down, they are both excellent, but how about wide open? Looking at the shots above, they are still both quite good in the center. Below are the overall scenes with both lenses wide open at f/1.4, focused in the same spot. Click to enlarge.
I took a 100% crop from image on the decanter to evaluate wide-open bokeh and specular highlights, and then took a 100% crop from each image on the left edge of the color checker. See the 100% crops below (again, be sure to enlarge the image to full size):
Both lenses produce extremely pleasing bokeh to my eye. Of course, the 35mm f/1.4 has greater blur due to the longer focal length on the larger sensor, and overall, the Fuji is just a little creamier. Both also show a very slight green fringe on the specular highlights, but despite this, I think both look great here.
On the edges, the Leica 25mm f/1.4 is a leg up, being considerably sharper on the image edge wide open. However, it’s quite clear that these are two outstanding normal lenses. Both are among the best I’ve used for any system.
Well, the X-E1 is a camera with fantastic image quality, that much is certain. Not surprisingly, it produces cleaner images throughout the ISO range and retains great detail. Is the Fuji the best of these three cameras then? In pure image quality from the sensor? Yes. In other ways? Not so fast…. Wait for my full review of the X-E1 for more detailed discussion, but both the GH3 and OM-D are much more responsive machines when it comes to autofocus. Still, Fuji has a winner on their hands. It’s also great to see Panasonic put out a body with very high image quality to match the OM-D on the stills side.
37 thoughts on “Mirrorless Battle: OM-D vs GH3 vs X-E1”
cool. thanks for that. im still descovering stuff on the gh3, like raw stacking, fun on a bun.
regarding corner performance at F1.4 an APS sensor has a .75 stop advantage / disadvantage meaning that the 35mm f1.4 shooting at F2 would give the exact same DOF as a mFT at F1.4.Comparing lenses at different DOF is rather pointless. With the better high ISO of the XE-1 there is no downside to having to stop down to F2 assuming you want to compare like for like.Or conversely pop on an the F0.95 25mm for almost identical DOF
Nice comparison Jordan! Looking forward to the full reviews.
Interesting; thanks. Makes me want that 25mm even more… (it’s on my wishlist since a while)
I got the 25mm some time ago and it never leaves the camera. Somehow it is a perfect lens for M43. It gives nice colours, bokeh, and it’s plenty sharp. In my opinion, you can’t go wrong with the 25mm 🙂
good write up , thanks ,
the panleica 25mm 1.4 is outstanding , my favorite af lens for m 43
What is your favorite non af
Really nice comparison! I also own the amazing OM-D with the Pana 14/2.5, Pana/Leica 25/1.5 and Oly 45/1.8
Three great primes (28, 50, 90mm FX camera equiv. focal lengths), all I need for the type of photography (street) I really enjoy! Thanks Jordan
Great write-up and thanks.
Any comment on color? I’m struggling a bit with my EPL5 (same sensor as the OMD) with producing too much red/orange. It was the same issue I had with NEX (convinces me it’s a Sony sensor). In your pics above, any comment on color accuracy? Skin tones, especially, are a concern for me.
Set “[WB Auto] Keep Warm Color” to Off if you want less of an orange tint to your images.
Try Photo Ninja, I really like the color rendering that produce.
So the fractalizing of Fuji higher-ISOs is actually feature? I thought it was a bug, but it doesn’t seem to be going away. Sad times. :/
Twaddler, I was wondring this too, as I’m very interested in the new Fuji X100S as a “travel” camera!
Hi Jordan, in your comparison could you check the highlight recovery capability of the x trans sensor and omd sensor.
I find the latter much better in that regards.
Thank you for this detailed review of these systems. Personally, I have been researching and considering the OM-D, Fuji XE-1, and lately, the Fuji X100S for a “Travel” system to complement my heavier (though very high quality) Pentax K-5 system. However, what gives me geat pause on Fuji, is an apparent edge artifacting issue mentioned on several sites, notably here: http://diglloyd.com/blog/2012/20121206_1-FujiXE1-artifacts.html
Frankly, I don’t see this issue being addressed by Fuji, and it worries me that it could still appear in the new Fuji X100S, which to my mind, would be a phenomenal camera (albeit if one’s style can accommodate a fixed 35mm lens). Now, I’m a sharpness and detail fanatic (as you can see on my site’s images), which makes me wonder if I should just sell my K-5 and get the newer K-5IIs without the AA filter. I wonder if that system would provide even more detail in larger prints than would these mirrorless systems. Thanks again, sir!
All Fuji owners know that OOC Jpeg-s are stellar. Fuji has managed to create Jpeg without artifacts. If some artifacts are observed it is in case raw files was converted offline. Greatest problems has Adobe. Fuji should not be complained, but the used raw converter. According Fuji all information to convert raw files was provided to external partners. Fuji owners should not shoot RAW ONLY, but raw + jpeg. Use Jpeg if artifacts occur in raw. Fuji Jpegs are really excellent, in most cases raw converters are not able to provide better noise vs. detail quality.
I shot some jpeg with Fuji xpro1 and I can see artifacts….It’ s not only a raw converter problem.
I strongly suggest to use in comparisons Fuji OOC Jpeg with LOWEST noise reduction in-camera setting ( value -2 ), these files provides NO chroma noise, but best details. These files could be directly compared with converted raw from other cameras. In contrast to Fuji files for all other camera manufacturers converted raw files provides better detail vs. noise then OOC jpeg.
This artifact is discussed quite fair here:
The issue is not serious but minor (if you do not try to show worst possible artifacts).
Its like also Nikon D800 has ( or any other camera with AA filter) moire artifacts but less visible than D800E ( like without AA filter )
The issue is much related with raw converters and your procedures.
There is already new raw converter version of Adobes LR / ACR with considerably improved results.
Did the OM-D used in this test have the latest firmware upgrade (v. 1.6)? It is my understanding that it improves shadow/highlights detail among other things.
I just did a quick and dirty test comparing my Panasonic 20mm f/1.7 with my Voightlander Color-Scopar 25mm f/4 using my OM-D EM-5. I set both at f/4 which should give the Panny an advantage since the Color-Scopar was shot wide open. Both sets of images were shot in RAW and opened in CS5, magnified to 100%.
The test image was a red tile roof shot end-wise to capture the rough end grain as well as the sharp edges.
I found the manual focus Color-Scopar wide open to be somewhat sharper in fine details than the Panny. I made two shots with the Panny; one AF and the other MF with no discernible difference between AF and MF. It would appear that the Color-Scopar; an older LTM using the M mount adapter is the better lens notwithstanding the Panasonic lens being a newer aspheric formula. Of course, in low light situations, the Panasonic would be the preferred lens, but the difference in IQ considering sharpness and contrast of the Color-Scopar was still remarkable.
I wonder how the Panasonic Leica 24mm f/1.4 would stack up against the 25mm f/4 Color-Skopar.
Thanks for the nice comparison. It will be pleasure to see the X-E1 review. For AF test please use Fujinon 18-55/2.8-4.0 lens, which is excelent, has good AF and is best choice for most X-E1 buyers. In any case use last camera/ lens firmware, which improved AF considerably.
One thought about the blurry edge on the fuji v. the oly is that it is a depth of field issue. with both at F1.4 the fuji will have about one stop less depth of field (because DOF is determined by (1)focusing distance which is the same here and (2)focal length of the lens and the fuji lens is a 35 and therefore more magnified than the 25 and therefore less DOF)
When focusing on close objects (rather than say a landscape scene) the distance from lens to main subject is very much less than distance to the edge of the scene. It is kind of like an isosceles trangle with the hypotenuse being the distance to the corner. So, it is possible the the edge is out of the depth of field area, which was established by focusing on a central subject.
I wonder what would happen with these two lenses if you did a second experiment where you focused on the edge. If then the two lenses were equally sharp at the edges, one could conclude in the first test that softness with the fuji lens was really just a depth of field problem.
Do the three cameras have all the same lens mounts? Meaning if I have invested in glass for one the gh3 then I can use all the lenses on the fuji or olympus also?
The Panasonic GH3 and Olympus E-M5 can share (as well as other micro Four-Thirds camera bodies). FujiFilm’s X-Mount is unused anywhere else.
I’ve been reading your review and following u on FM, great reviews!
Currently I have xe1 with 35mm 1.4, vc 21mm and zm 50mm. The AF is too slow for my need. (I Have a 4 years old son) but I love the IQ from it. I’m not sure if I should switch to em5 before the promotion end on late march. I know that u love ur em5 and it seems like u love the IQ from fuji xe1, are you going to keep the xe1?
Which one would u recommend?
Sorry for the abbreviation, I’m typing on an iPad.
Thanks! I love both cameras. I am keeping the X-E1, and I’m rolling with a three lens Fuji kit when I bring it out, with the 18-55 f/2.8-4, 35mm f/1.4 and an adapted Contax G Zeiss 90mm f/2.8 Sonnar. I did sell off some of my lesser used Micro 4/3 gear to pay for the X-E1 stuff, but I’m running with two systems right now. The X-E1 has incredible image quality, with really rich files and beautiful high ISO work. However, the AF is definitely not suited to following around an active child. For that, I grab my OM-D or GX1, as the m4/3 cams just destroy the X-E1 in autofocus speed and low light acquisition. However, I’ve gotten accustomed to the X-E1 and am having fewer instances where it fails to lock now. That said, when I shot theater a few weeks ago, I brought both cameras and ended up using the OM-D with Panasonic 35-100/2.8 for about 95% of the shots, simply because it was so responsive. I didn’t miss a single shot due to AF, and it was really a brilliant combo to use.
So, I’m using the Fuji kit for my typical day to day stuff, and grabbing the OM-D when I need a wider selection of gear, macro, or will be shooting anything moving.
In your pics above, any comment on color accuracy?
Good write-up. I absolutely love this site. Keep it up!