Now we come to the fun part – how does the E-M1 perform compared to the D800? I looked at three aspects:
- Rendering of detail – resolution
- Low light capability
- How large can I print with the E-M1.
First, consider that any discussion of imaged detail is really about the system of lenses and camera. To make it comparable, I have used the top “Pro” lenses for the D800 and E-M1, the standard 24-70/80mm f/2.8, for comparison. On the E-M1 this means the new ED 12-40mm f/2.8 Pro (24-80mm 35 equivalent) and for the D800 the excellent Nikon AF-S 24-70mm f/2.8 ED. One of the comparison shots is from a local park. Both cameras were shot side by side at the same equivalent focal length, ISO, f-stop and exposure. A crop from each then allows an evaluation of detail. Each image was sharpened in Lightroom as I would normally, the E-M1 at 1, 40 and the D800 at 1, 60. The crops were resized to be the same size – 640x 930 pixels. This would equate to a 48″x64″ print, based on normal screen resolution of 72dpi – much larger than I would ever print. The D800 crop clearly shows the effect of the much higher resolution, as expected. (Click images to enlarge).
|Entire image E-M1 1/2000s, f/4, ISO 200||E-M1 Crop||D800 Crop|
So what does this really mean? There are a couple of obvious conclusions. First, it takes a pretty large print to see the differences between the images and second, the D800 images can be cropped much more heavily. This just means that using the E-M1 requires pretty good framing of the initial image if large prints are required. This is clearly a subjective comparison, just representative of several I did to reach that conclusion. On comparison of 17″x22′” prints, one has to look very closely to see the differences. At normal viewing distances, there is not much difference in detail perceptible.
So, how about printing? How big can one print from these images and still get good results? Imaging-Resource.com has done an extensive amount of work on print quality in their camera reviews. A summary of print quality versus sensor size and ISO performance from Imaging-Resource is contained in the accompanying chart. (More on E-M1 high ISO performance a bit further down).
As mentioned above, at base ISO, very large prints are possible from the 16MP E-M1. Since I never print larger than 17″x22″, this is good news. However, it clearly shows that emphasis on getting the shot right in the camera is very important for large prints. Again, for the stated objective of a light-weight travel camera-no problem. I do have to keep in mind that for years I was satisfied with my D100 6MP and D300 12MP results. The D800 is just so far above everything else the comparison requires some perspective.
There are over 30 lenses for the M4/3 system and many more that can be used in manual mode with adapters. The question is what range and which lenses to choose. As with most system cameras there is good, better and best. I am looking for a combination of excellent quality and coverage. Ideally, from 14mm to 300mm (35 equivalent) in a reasonable travel size. There are a couple of do-it-all wide to tele zooms, but they are not of the highest quality optically, so are not under consideration. After a lot of reading, testing and shooting, the following is my travel kit:
- Olympus OMD E-M1
- Olympus Pen E-P2 backup body
- Olympus lenses (M43)
- 25mm f/1.8
- 45mm f/45
- 12-40mm f/2.8
- 40-150mm f/4-5.6
- Lumix G 7-14mm f/4
- Lowepro Photo Traveler 150 backpack
The whole kit with batteries weighs about 7.5 lbs. The 40-150 does not really fit with the rest in image quality, but will do until the 40-150 f/2.8 Pro lens is available later in 2014.
I routinely test new lenses to ensure that my copy has no problems (I have found sub-par copies of specific lenses which were replaced). I use Imatest with a standard slant target to measure resolution at various focal lengths and some different apertures. Several results charts are included here for reference. Please note that the LW/PH numbers on these charts may differ from other reported results because of specific sharpening and test setup parameters. However, all of the presented tests were performed in the same manner and are valid relative to each other on the charts. Repeatability of results is good. Each lens is tested at several focal lengths and selected apertures. Each chart value is the from the best of several test shots from manual focus and auto focus, if applicable. All results are for the center of the image area only. Results for corners and edges are available in the many tests by dpreview.com, Photozone.de, slrgear.com, imaging-resource.com, and dxomark.com, among others. These excellent sites also check for distortion, vignette, and at all of the various f/stops. My tests are just to verify that I have a good copy.
For my purposes, I came to the following conclusions:
- The Olympus and Lumix m4/3 lenses are actually sharper than most Nikon FX lenses on the E-M1.
- Sensor size is dominant in determining resolution for these high quality lenses, at least to 24MP.
- The E-M1 with a quality lens is capable of very large prints.
- Most of the time I will use native M4/3 lenses on the E-M1.
High ISO comparison
These cameras all have setting up to ISO 25,600, but how good are they pushed that far? I have been very pleased with the D800’s high ISO performance and routinely use ISO 3200 for low light sports action. Does the E-M1 measure up? To find out I set the cameras up side-by-side and shot with equivalent focal lengths at the same exposure. Processing of the raw files consisted of sharpening identically in Lightroom – no noise reduction. The full image shot, showing the crop area is:
|E-M1 ISO 400||D800 ISO 400|
|EM-1 ISO 800||D800 ISO 800|
|E-M1 ISO 1600||D800 ISO 1600|
|E-M1 ISO 3200||D800 ISO 3200|
|E-M1 ISO 6400||D800 ISO 6400|
|E-M1 ISO 16,000||D800 12,500|
|E-M1 ISO 25, 600||D800 ISO 25,600|
- E-M1 – Excellent performance up to ISO 3200 – 6400
- Above 6400 E-M1 exhibits considerably more noise than D800 and loses detail.
- ISO 12,600 and 25,200 are still useable if the images are not enlarged too much (8×10 -5×7).
- The high ISO noise is primarily luminance – Chroma noise is well controlled. Looks like film grain rather than digital noise.
Go to Page 3 – Conclusions: http://afpphoto.com/writings/tests-2/olympus-omd-e-m1-conclusions/