Olympus OM-D E-M1X – Test Results
The results of our tests show the E-M1X to be a first-rate performer for an M4/3 camera and a better performer than rival professional-level sports cameras when it comes to stabilisation and autofocusing for still shooting.
The build quality and weatherproofing of the E-M1X are a match for the top pro cameras from Canon, Nikon and Sony and the dual battery system allows recording capacities to come close to those of DSLR cameras. This is no mean feat when the battery has to power the EVF as well as all the usual functions.
Since publishing a detailed First Look at the new OM-D E-M1X camera last week, we have been able to carry out some of our standard technical and user tests. This report has been prepared to complement the initial review, adding the results of our standard tests. Links have been provided to enable readers to jump between the two reports.
Angled view of the OM-D E-M1X camera. (Source: Olympus.)
As outlined in our initial report, the 20.3-megapixel 17.4 x 13.0 mm M4/3 Live MOS sensor in the E-M1X has been backed up by dual TruePic VIII processors, which roughly double the processing capabilities of the E-M1 Mark II (which introduced this processor chip). Since our initial report, more information on the technologies in the new camera has emerged so we’ll start by covering features we couldn’t cover fully in the initial report.
There are now eight CPU processing cores and this increase in processing power has enabled Olympus to build plenty of new technology features into the new camera and increase the capabilities of existing features. However, running additional processors produces more heat so Olympus has introduced a heat-dissipating panel that draws heat from the processors to the camera’s chassis, where it is able to radiate out to the environment.
Interestingly, the basic AF system is the same as in the E-M1 II, with on-chip phase detection using 121 all cross-type points for accurate detection of subjects. However, more processing power enables AI (artificial intelligence) processing to be used for the updated autofocusing algorithms, which are now capable of ‘deep learning’ and fast enough processing to apply real-time subject recognition.
‘Deep learning’ is an interesting concept because it relies upon a stored database of sample images that can be scanned through in a fraction of a second. As mentioned, the Tracking Subject setting in the Customising menu (page A3) can recognise three scenarios: motor sports, airplanes and trains. In each case, with the AF system is ‘trained’ to pick up the critical focus point from an outlined frame superimposed on the subject.
The system will continue to track the subject, keeping that area in focus, which involves a substantial amount of data processing. The detection area and type of subject will vary from one scenario to the next, which is where the scenarios stored in the camera come into play.
Future subject type additions are sure to be possible through firmware updates as popular scenarios like birds in flight and various sports are added to the database. We’ve displayed some of the results we obtained with the motor sports and airplanes settings in our previous review, showing the camera was easily able to identify and track subjects with each of the scenarios we presented it with, even with the M.Zuiko Digital ED 300mm f/4 IS PRO lens, which has a 4.1 degree angle of view.
Dual processors also means the camera can deliver UHS-II data to both card slots, instead of only one as in the E-M1 II. This will mean faster buffer clearing and, probably, greater buffer capacity. They also support two-stage noise reduction for improved image quality at higher ISO settings.
Improvements to the image stabilisation (IS) system have partly resulted from a re-engineered gyro sensor that raises the shake correction from a maximum of 6.5EV in the E-M1 II by one stop to 7.5EV. This has enables Olympus to add a handheld mode to the E-M1 II’s tripod-only High Res Shot function, which delivered 50-megapixel resolution for JPEGs (80 megapixels for ORF.RAW files).
Using High Res Shot requires the camera to control motion to a sub-pixel level for the duration of the exposure, after which the processor has to align 16 separate frames and integrate them into a single exposure. (Some very slight cropping can take place as part of this process.) This function requires the electronic shutter and is available in the P/A/S/M modes. It can to produce JPEGs at a resolution of 8160 x 6120 pixels and raw files at 10,368 x 7776 pixels.
Also relying on multiple exposures is the Live ND mode, which simulates the effects of a neutral density filter by applying a digital filter to reduce exposures and combining a series of frames in a way that keeps static elements in the scene stationary while moving elements are blurred. The results depend upon whether the camera remains steady. If you move the camera or zoom the lens during the exposure, some interesting blurring effects will result.
To use the Live ND mode, the camera must be in the M shooting mode because it is designed for use at slow shutter speeds. The camera will apply its own shutter speed limits, depending on the ND setting, with a maximum overall speed of half a second applied for the ND32 setting.
As noted in our ‘First Look’ article, we were able to capture 1290 frames in the three shooting sessions where we tried the camera out and there was still more than 20% of the charge remaining in the first battery. For most of this time, the camera was in silent mode (which uses less battery power) and we shot using the EVF. However, but we did a lot more chimping than usual to check recorded images and scroll through burst sequences and that used the LCD monitor.
Subsequent investigation of the dual battery system and USB-C power delivery have revealed that photographers can’t run the camera from mains power and charge a battery simultaneously. If it’s operating with mains power it will do so as long as it remains connected and the batteries will retain their existing power levels. To charge the batteries from a USB-C source, the camera must be switched off.
This will work just fine for photographers who shoot time-lapse sequences and those who use the camera for recording long sequences of movie clips. The ability to recharge the batteries from a USB-C source will also be convenient during periods when the camera is not being used, for example while driving to a location. Charging speeds will depend on the voltage of the power supply and the type of cable used for charging.
Because no lens was supplied with the camera, we have reviewed it with our own M.Zuiko Digital ED 12-100mm f/4.0 IS PRO lens, which was reviewed in November 2016. We set out to run Imatest tests on both straight-out-of-the-camera JPEGs and also ORF.RAW files that were converted into 16-bit TIFF format with Olympus Workspace, a new application that replaces Olympus Viewer 3. Raw files from this camera are not yet supported in Adobe Camera Raw (ACR), our preferred raw file converter.
Our Imatest assessments showed the review camera was able to meet expectations for the sensor’s 20-megapixel resolution with JPEGs and delivered exceeding expectations by a comfortable margin with ORF.RAW files. With the normal 5184 x 3888 pixel JPEGs, the highest values were obtained between ISO 64 (‘Low’) and ISO 800, after which there was a gradual decline as sensitivity was increased. TIFF files from the equivalent raw files followed the same pattern, which is to be expected, and met expectation for the sensor’s resolution. The graph below shows the results of our tests.
Colour reproduction in JPEGs was a little more restrained than the E-M1 Mark II’s but both contrast and saturation were nicely balanced to provide images that looked good on screens yet provided scope for editing. Colour reproduction figures from our Imatest tests were close to, although slightly higher than, those from the normal ORF.RAW files. File processing is probably a reason for this similarity.
Unfortunately, some reason, the Imatest software wouldn’t work with the 461MB files produced in the High-Res Shot mode’s tripod setting, which produces 10,368 x 7776 pixel files. For shots taken with the handheld setting, which produces 50-megapixel files (8160 x 6120 pixels), we were able to measure the resolution of JPEGs but not raw files. We’re not certain why this occurred but given the way in which the High-Res Shot files are produced (multiple images are captured in quick succession and combined into a single high-resolution file), the resulting images may not have been ‘standard’ enough for the software to utilise.
Long exposures at night were generally clean and noise. free at ISO settings up to ISO 6400, where the first signs of image noise became visible in shots taken without noise-reduction processing. By ISO 25600, noise was apparent and images were slightly softened and a few colour artefacts could be seen. However, shots would be usable at small output sizes.
As mentioned in our First Look article, autofocusing performance when shooting stills was outstanding and tracking AF was the best we’ve seen in any camera we have reviewed so far. The AF system also performed well in very low light levels, locking onto subjects quickly with no trace of hunting for a sharp edge.
Auto white balance performance was better than most cameras we’ve tested in the past 12 months and the E-M1X provides plenty of ways to ensure the correct colour reproduction, including a ‘keep warm tone’ option for the Auto setting, seven presets, Kelvin temperature selection, +/- seven steps of adjustments on the amber/blue and green/magenta axes and four ‘capture’ WBs for custom measurements.
With the normal auto WB setting, the review camera produced close-to-neutral colours under warm-toned LED light and fluorescent lighting and came close to removing the warm cast of incandescent lighting. Very similar results were obtained with the ‘keep warm tone’ setting with shots taken under incandescent lighting ending up marginally ‘warmer’.
There’s no preset for LED light and the presets for incandescent and fluorescent lighting tended to over-correct. Manual measurement produced neutral colours.
The video capabilities of the E-M1X are almost identical to those offered in the E-M1 II so it’s no surprise that the resulting clips were similar in quality. The addition of Log capture should make it more competitive, although Olympus has yet to release a look-up-table to convert footage for use in popular video editing applications.
M4/3 cameras have an advantage over competing ‘full frame’ DSLRs because the smaller frames are easier to use at full frame width without pixel skipping or binning. Unfortunately, the AF system was slightly slower and less consistent when tracking moving subjects in movie mode, particularly with the high-speed recording settings (an example is shown below). Aside from these modes, we found picture quality to be generally good.
Soundtracks – in modes where they could be recorded – were clear and possessed a surprising amount of stereo ‘presence’, although this could only be heard when playing clips back on a computer with stereo speakers. The camera provides settings for wind noise reduction and adjusting the levels of both recording and microphone levels. We found the recording level adjustment to be quite sensitive.
We carried out our timing tests with two SanDisk Extreme Pro SDHC UHS-II U3 Class 10 cards, each with a capacity of 32GB. Both cards claim read speeds up to 300MB/second (2000X) and write speeds up to 260MB/second and are optimised for the latest cameras and 4K camcorders.
The review camera took just over a second to power-up for the first shot. Autofocus lag averaged approximately 0.1 second and it was eliminated with pre-focusing. Shot-to-shot times averaged 0.45 seconds. It took just under a second to process each JPEG file on average.
Having dual card slots enabled JPEGs to be recorded to one card and raw files to the other, so this time also applies to raw files and RAW+JPEG pairs. When we removed one card and measured the processing times for JPEGs and raw files separately, there was little difference between them.
The E-M1X offers six continuous shooting settings, with high and low frame rates for the regular sequential shooting mode, which uses the mechanical shutter; the silent sequential shooting mode, which uses the electronic shutter and the Pro Capture mode, which combines high-speed shooting with fast autofocusing and a 120-fps high-speed viewfinder refresh rate. We only tested the high-speed settings for each mode and recorded JPEGs to one card and ORF.RAW files to the other.
Using the mechanical shutter, the review camera captured 64 Large/Superfine JPEGs in 4.4 seconds before the first hesitation occurred, which works out at a frame rate of just over 14.5 fps. It took 11.8 seconds to process this burst. With the electronic shutter, continuous recording paused at 125 frames, which were captured in 15.5 seconds, which is a little over 8 fps. Processing took 10.2 seconds.
We used the Pro Capture H mode on the default settings with a frame rate of 60 fps, 14 frames recorded pre-shutter and a burst limit for 25 frames. By our assessments the frame rate was maintained. The camera allows users to select frame rates between 15 fps and 60 fps, up to 35 pre-shutter frames and a burst limit of up to 99 frames.
Please Login or Register to access the Conclusion.
Based on JPEG files taken with the M.Zuiko Digital ED 12-100mm f/4.0 IS PRO lens.
Based on ORF.RAW files converted into 16-bit TIFF files with Olympus Workspace software.
High-res shot mode JPEGs.
All shots taken with the M.Zuiko Digital ED 12-100mm f/4.0 IS PRO lens
Auto white balance with incandescent lighting.
Auto white balance with fluorescent lighting.
Auto white balance with LED lighting.
ISO 100; 60 second exposure at f/4, 24mm focal length, available light exposure at night.
ISO 400; 30 second exposure at f/5.6, 24mm focal length, available light exposure at night.
ISO 1600; 10 second exposure at f/6.3, 24mm focal length, available light exposure at night.
ISO 6400; 5 second exposure at f/8, 24mm focal length, available light exposure at night.
ISO 12800; 2.5 second exposure at f/9, 24mm focal length, available light exposure at night.
ISO 25600; 1 second exposure at f/8, 24mm focal length, available light exposure at night.
Live ND shot with ND 32 setting, 100mm focal length, ISO 64, 1/2 second at f/5.6; camera jiggled during exposure.
Live ND shot with ND 32 setting, ISO 64, 1/2 second at f/4.5; lens zoomed from 100mm to 24mm during exposure.
Still frame from Cinema 4K (4096 x 2160 pixel) movie clip recorded at 25p.
Still frame from UHD 4K (3840 x 2160 pixel) movie clip recorded at 25p.
Still frame from Full HD (1920 x 1080 pixel) movie clip recorded 50p.
Still frame from Full HD (1920 x 1080 pixel) movie clip recorded at 25p.
Still frame from Full HD (1920 x 1080 pixel) movie clip recorded at 24p.
Still frame from Full HD (1920 x 1080 pixel) high-speed movie at 120 fps.
- Build: 9.2
- Ease of use: 9.0
- Autofocusing: 9.5
- Image quality JPEG: 8.9
- Image quality RAW: 9.0
- Video quality: 8.8