I just received the Olympus E-M5 Mark II in for review today, and I’ve only had a few hours to shoot so far. However, like anyone interested in this camera, the very first thing I had to test was the 40 megapixel Hi-Res mode. The E-M5 Mark II has a feature that shifts the sensor in 1/2 pixel increments for 8 shots, then assembles a final image that yields a true 40 Megapixel image.
For this rather informal test, I put the E-M5 II on a tripod and took back to back shots: the first at the native 16 megapixel resolution and the second in the 40 Megapixel Hi-Res Mode. IS was turned off for the 16MP shot, and triggered with a 2 second self timer to ensure the sharpest shot possible. Hi-Res mode was triggered with a 1 second delay before the sequence started.
The first test I performed was a simple shot near infinity of the city skyline, with the Scioto River in the foreground. The 40MP mode requires no movement in the scene (and no movement of the camera), so I was curious how it would fare with the moving water. The full shot, reduced from the 40MP original, is below:
I then took the 16MP image and enlarged it to the same pixel dimensions as the 40MP image. This will allow a direct comparison of the extra level of detail offered by the 40MP mode. I could show both at 100%, but then discerning the real detail is harder…viewing both at the same magnification (as one would if you were printing both to, say 36″ wide), makes a lot more sense, as I’ve done here. The 16MP image is on the left, while the 40MP image is on the right. Click to enlarge the image and see it full size. Verdict? If you take care and use a sturdy tripod on a static scene, you get a massive increase in true resolution and detail with the 40 Megapixel mode.
About that moving water: Indeed, the river shows some artifacts in the 40 megapixel image, as expected. However, at least in this scene, they are fairly minor. A 100% crop is seen below. It’s worth noting that taking the 40 megapixel image and downsampling to 16 megapixels will eliminate the artifacts seen in the crop below. It also results in an extremely sharp 16MP file that shows more detail than the standard 16MP file from the E-M5…so even if minor motion artifacts appear, a very good image can still be had throughout, though you will lose some of that resolution that you’d gain if the whole image were still.
For the next image, I did a closer image of something completely static. The processed image is below:
Again, I took 100% crops of the two images, with the upsampled 16MP image to the left, and the full 40MP image to the right. Note that the crops below are from the unprocessed image, straight out of the camera. Again, the 40 megapixel image shows a very stark advantage in detail. Again, click the image to view the crops at full size.
Early conclusions? If care is taken on a truly static scene, the 40 Megapixel high-resolution mode is no gimmick: it really does produce a massively high-resolution file with a very clear advantage in detail.
I’ll be exploring more about the 40MP mode as well as everything else about the new E-M5 Mark II in my upcoming review. Come back for that shortly!
13 thoughts on “Olympus OM-D E-M5 Mark II – A Quick Test of 40MP Mode”
Which lens was used for this tests?
Olympus 12-40mm f/2.8
Just for information two reports analyzing the HiRes mode in more detail:
There is definitely an improvement (about 10% for MTF50), more small detail is captured, too, albeit with much less contrast. The benefit in my opinion is the lack of moire and aliasing.
Well my experience is very different but bear in mind I only had the camera for about an hour. I found it incredibly hard to get an image in 40Mpixel mode which was sharp. I tried with the 12-40 f2.8 and the 40-150 f2.8 and finally with a Panasonic 7-14mm It was only with this lens and running an 8 sec exposure (it was getting pretty dark) that I got a sharp image at hires. The camera was on a very solid tripod! I should add that the images taken with the 40-150mm were with an exposure time of 1.6 and in every case the lowres image was razor sharp but the hires was awful. It would appear that camera movemnt in the order of microns is enough to wreck the hires mode.
Since you’re getting sharp long exposures and blurry short ones, it leads me to this: did you set a shutter delay before the sequence starts? You can set up to a 4 second delay in order to allow all vibration from the shutter press to settle before the sequence starts. I’ve gotten sharp images in high res even with the 75-300 at 300mm, so it’s certainly not impossible. It does need to be very steady though.
The HR mode really requires very, a very sturdy tripod. Just like if you want to get sharp results from Nikon D8xx or Sony A7r, you need very sturdy tripod, set shutter release / mirror up before release to get vibrations off etc. And more often you want to add some extra weight top of the camera to really dampen the vibrations.
And think what it will be with those 50Mpix cameras, as now with Olympus on 64Mpix photos you have difficulties.
A minor correction:
“The E-M5 Mark II has a feature that shifts the sensor in 1/2 pixel increments for 8 shots…”
The HiRes mode does not shift the sensor in 8 half-pixel movements, but first shoots 4 pics w/ a whole 1-pixel shift, yielding full RGB coverage in every pixel for increased color bit-depth, and then in direct succession another 4 pics series w/ a half-pixel shift to increase resolution. Still in all a total sequence of 8 pics though.
Indeed…I was simply trying to shortly explain what was going on. I’m aware of the square patterened whole pixel shifts then offset similar four shot square pattern. I will be covering the details in my review, but was summarizing the effect for this short article.
Basically that is like a half-pixel movement, but in one pixel and half-pixel movement 🙂
Meaning if you move first half pixel to right, then again half pixel right, you get half pixel AND full pixel.
It is same as moving full pixel and then moving half pixel, as you will end up to take 4 photos that gets one pixel to cover all RGB values.
I hope you could take few test photos from moving water in HR mode, where you use long exposure mode, meaning that you set camera to take 6-8 second shutter speed and this by using a ND filter. You should get much much better results by this way.
I have already done so. Check the first image in my portfolio (link at the top) for a sneak peek. It works ‘better’ but shows uniform zigzag artifacts even with long exposures. Solution? A 1px Gaussian blur on the water and sky (they were blurred by long exposure anyway). Great final result.
I find this amazing regarding the EM5 MKII and it’s hi red mode.
Do you think the Oly EM-5 II 40 MP shot with timer [ 3 sec. ] would give a good result
using a small hand size tripod [ manfrotto ] of good quality ?
I use one when I travel as it fits anywhere and works really well with my EM-1