Since I have a couple of nice Olympus lenses for the BMPCC (and GM1) I’ve been interested in the new Olympus OM-D EM5 Mark II. I love a large, light-sopping sensor, so find if difficult to trust MFT reviewers who believe full-frame sensors are over-rated. Kirk Tuck, at The Visual Science Lab shoots with the Nikon D800, so if he loves the EM5 what might I be missing?
One of the features that really intrigued me is the “Hi Res” mode, where it take 8 frames and interpolates them into one 40mp high res image. With a 16mp chip, I wasn’t interested in the “resolution” issue as much as the full-color information available to each pixel (light ray). Could the EM5 shoot super color-clean images like the Sigma? I figured I’d test (more on that later).
Unfortunately, I’m having trouble getting past my initial qualms about the MFT format. Certainly, outdoors, the EM5 takes very nice photos. It’s small and quiet. Built solidly. Moving indoors, however, the low resolution and light sensitivity hobble it. I found some old Nikon lenses at Goodwill a few days ago, so gave them a spin. I put the 105/2.5 on a speedbooster and took this.
In PS, I could not recover enough detail in the blown-out cheek. She was standing next to a window, in even, not especially bright light. If I had been shooting with an Olympus 45 I’m certain the photo would have been tack-sharp. If I had been shooting with a Nikon I could have saved the blown-out parts of the image. Okay, the camera isn’t a studio or portrait shooter, that isn’t news to anyone.
If I didn’t have, and love, the Ricoh GR I’d want this camera as my take-anywhere camera. However, I wouldn’t be able to put the EM5 in my side jacket pocket (unless I put a pancake on it). Once I have to put a camera in a bag it must compete with every other non-pocket camera. Yes, the EM5 is very small, but not significantly smaller than my current favorite, the Sony A6000.
There’s nothing like using a camera to see if what should work in theory, works in practice. I expected the in-camera stabilization to be as cool and useful as many people have reported. I found the opposite. If I take the camera outside and photograph a building at 1/8th of a second, I can get a nice sharp image, as good at an APS-C sized sensor at 1/60th, say. In real life, people are in my shots and PEOPLE MOVE! They will be less blurry in the APS-C size camera, shooting at 1/60th) than the sensor-stabilized camera, shooting at 1/8th. If I am going to take photos of buildings, I’m going to put the camera on a tripod or I won’t mind carrying a bigger camera. So why would I want to shoot with a 16mp, small sensor camera, when I could get a 24mp, larger sensor camera for less money? Or I’d use the Sigma DP1M.
Then there is video. Factoring out ISO and DOF issues, the Panasonic GM1 shoots beautiful video. Perhaps because of the odd video aspect-ratio issues, according to Gordon Laing at CameraLabs (which makes sense to me), the video from the EM5 is, as Andrew Reid at EOSHD points out, not very good (I paraphrase). The camera stabilization is good if you’re shooting with manual glass and standing fairly still. It doesn’t give the nice POV look of a camera on a stabilizer, like the Sony A6000 on the Nebula 4000.
If I had to have one camera with two prime lenses (say the 17 and 45), and I wanted the setup as small as possible (and rugged) the EM5 is THE camera to have. For me, having other cameras, the EM5 trades off too much photo dynamic range and resolution (using the MFT sensor) in comparison. If I take it with me, and I take a photo in sub-par light, I’ll kick myself for not bringing a bigger camera.
Again, in no way am I trying to talk anyone out of getting this camera. If one’s interest in manual-mode shooting, portability and composition than this is beautiful camera. If I had to take one camera on a European vacation this camera would be on my short-list. I can see how Kirk Tuck loves this camera. After lugging around big cameras in his day job it must be nice to grab the EM5 and hit the sunny streets of Austin. It’s the camera equivalent of riding free on a motor-bike.
Because the EM5 takes 1 second to create a Hi Res image, I had already given up on the idea of comparing it to the Sigma DP1. I forced myself for you few readers of Maxotics 😉 I picked a scene with a fence because I find bright lines (steel wires) are perfect to show aliasing/color issues. They should be neutral.
First, the two shots. I cropped the DP1M to match the EM5
As you can see, though shot at 1/400th of second, Pepper is blurry from moving during the 8-shot sequence. Now the Sigma DP1M.
I l’ll leave you to judge the two images. It’s no contest for me.
Next I looked closeup at the pixels through PS.
Next the Sigma
Even though the Olympus EM5 removed a lot of color aberrations common to bayer sensors, it still succumbs to it. I believe the problem is that it is trying to create a high-resolution 40mp photo instead of a color-pure 16mp image (which is what I want). My 2 cents is that Olympus should work on getting 16 million perfect color pixels and faster exposure times. If can do that, THEN is might compete with the Sigma.
The Sigma image, as usual, doesn’t exhibit color distortions. Of course, the Sigma struggles with red, and high-ISO shots, but in good light it approached medium format clarity. With the Sigma, neutral colors stay neutral. The sky in the Olympus is a bit over-saturated, like most bayer cameras. Although the Sigma has limited dynamic range between 6 and 8 stops according to Bill Claff (not much compared to Nikon’s 12 stops), when the exposure is within that range the color is perfect–as good, I would argue–as the tank-sized D810. Sigma cameras are proof to me that dynamic range is not a be-all, end-all.
I want to give another plug for FastRawViewer. I’m no longer shooting RAW+JPG in the Ricoh GR or Sony. The viewer is fast enough that I can scan through the images and open them in PS easily when I want to bring them to fruition.
Could you detail out the specs of the speedbooster that you used on the EM5?
It was very long ago. I used one of the first Chinese speedboosters that came on the market, I believe, for $100. For what it was, I had no complaints! My only issue was that it didn’t fit the BMPCC I had at the time, though a little dremel work fixed that problem. I’d imagine any speedbooster you bought now would be as good or better.