My usual deployment: handheld video, with HLD-8 battery grip, Zeiss ZM 1.4/35 Distagon rin an adaptor, and a Zoom H5 audio recorder. I am working on fixing the hard/sharp/uncomfortable edges of the battery grip with a silicone putty compound called Sugru, and will post the results in a future post.
Better late than never (or, I finally get around to trying out the second coming): the Olympus’ E-M5 Mark II. Many of the long-suffering readers of this site will know that I had a period of enthusiasm for M4/3 gear (and specifically the original E-M5) before that abruptly came to a halt in early 2014. The reasons were simple: firstly, camera technology has moved on; what was an impressive size/quality ratio in 2012 is not in 2015. Secondly, my output requirements have changed; the cameras have never had sufficient resolution to make a meaningfully-sized Ultraprint. Thirdly, there was no real solution to the shutter shock problem of the E-M1, which produced unusable images under basically every shooting condition – from 1/90s to 1/350s*. We were amongst the first to use the original E-M5 for video because of its stabiliser, and continued to use the E-M1s for video (including all of the workshop videos after The Fundamentals), Olympus and I then parted ways, and it appears they found new champions less demanding of their equipment. But, why the change of heart for me?
*I demand critically sharp pixels and can achieve them with the same camera under other conditions. Different users may have different thresholds of acceptability and different levels of shot discipline and not see any problems. On top of that, I tested >80 E-M1 bodies including >70 at Olympus Malaysia HQ, all of which exhibited the problem. The initial review unit did not, because it was a preproduction unit with a shutter module from a different batch. A firmware update was subsequently released with EFC, but it only works in single shot mode.
Firstly, I must apologize for the delay in this review, but it’s difficult to write a meaningful summary of something you have not shot with (though it never ceases to amaze me how often this happens both online and in print). And before anybody asks about the black tape, I prefer not to retouch out reflections and do not believe in free advertising – undoubtedly just as Olympus Malaysia does not believe in free cameras. And despite what armchair experts on forums think, I’ve never been paid for a review*, never worked for them (you’re confusing me with somebody else) or gotten any preferential treatment other than loaner units, which is no different to other sites and publications. On top of that, I’ve had enough product images stolen by various unscrupulous entities – including major brands – that I’m going to make these as specific and unusable for marketing purposes as possible. Let me make it clear before I get accused of being an Olympus fanboy again that this camera is a normal unit bought at retail with my own money. I run a photography business. Any piece of equipment must therefore be able to justify itself by opening up the shooting envelope.
*If I did, good business sense would dictate I’d be reviewing stuff for a living instead of working as a photographer.
Those familiar with the original camera will find that simultaneously a lot and not much has changed – and this is going to be a recurring theme throughout the review. Initial appearances suggest that the whole thing has gotten just a little bit meaner and more serious: the paint finish is now rough, the shells are a bit thicker, the weather sealing a bit more complete, and the rubber a bit grippier. It reminds me of an E-M5 that mated with an E-M1 and an OM-4 for good measure. It’s nice in the hand, there are lots of customizable buttons, dials are positive and nothing is too amiss. But it’s still obviously an E-M5. I liked the original because it moved the price/quality/size envelope forward dramatically forward in 2012, even though it had its share of compromises. We are now in 2015, however, and technology has moved on.
So far, so good. Better still are the list of meaningful changes: headlining are a new IBIS system that is not only supposedly video-optimized (and includes an electronic mode that matches frames and crops a little) and delivers better performance for stills, but is also capable of moving in half pixel increments to combine eight images for a higher resolution image – in RAW and JPEG (more on this later). We gain the E-M1’s 2.36m dot EVF, along with a rotating touch LCD on the back. On top of that, it seems the shutter shock issue has been solved once and for all with not just electronic front curtain, but a full electronic shutter with a 1/16,000s limit and continuous drive modes with both EFC and full electronic shutters. Focus peaking is finally smooth and doesn’t visibly affect the frame rate. It is unclear if this is a new sensor or a new processor or a some combination of the two. It also has the upside of delivering an improvement in color, too.
In fact, the camera appears to have been largely optimized for video: aside from the stabilizer, you can now have 1080p in whatever frame rate you choose up to 60fps, with a 77mbps ALL-I codec, proper audio level monitors, audio out monitoring, phantom power, and a clean HDMI out feed. We’d already been shooting video – both the workshop videos and B-roll on commercial productions – with the E-M5s and E-M1s for quite some time. I still remember being laughed at by the first crew until they saw the stability of the handheld footage, and various ‘experts’ on the internet. Ah well. The IBIS system is really the crown jewel of Olympus’ cameras; even though Sony’s system in the A7II derives from the same technology, the implementation just doesn’t seem as effective.
However, it’s not all a bed of roses.
The high resolution mode needs a perfectly still subject, a rock-solid tripod, and even then frequently produces gridlike artifacts where it appears the camera cannot figure out if something is detail or false signal. And though you get a 64MP raw file, there’s at most ~36MP of detail in there. Olympus claims it can fight the big boys: realistically, forget about comparing it to the 50+MP medium format cameras; it’s not even close. It’s really a very soft file at actual pixels and 64MP. It appears to resolve at a comparable level to the D810: the latter has better pixel level acuity even after the E-M5II file is downsampled to match, but the E-M5II has much fewer artefacts, no false color, no moiré etc. In fact, the native color out of camera is quite a bit better – but you will have to shoot in the high resolution mode to get it. Still, perhaps there are creative uses for it when applied to non-static subjects.
Then we move on to the video side of things: even though the data rate is three times that of the E-M5 and 2.5 times the E-M1, there doesn’t appear to be three times the information. The camera is a significant improvement on the other two, but still produces soft-looking and undetailed images compared to the D810 or GH4, and dynamic range is distinctly lacking. Even if you use the most neutral profile with lowest contrast and saturation, clipping occurs very quickly; it really needs a ‘flat’ profile like the Nikons for easy postproduction grading. Fortunately, there is clean HDMI out though you cannot record both internally and using the HDMI device simultaneously. Nor can you adjust exposure parameters using the normal dial controls while recording; you’ve got to poke through the on-screen menus. Again, this seems like a significant oversight given that the cameras’ main video strength is in handheld shooting – and you’re certainly not going to have a spare hand to operate the touch screen. At least rolling shutter artefacts appears to be very well controlled.
When your precarious still life imbalances partway through the HR mode capture.
There’s more give and take here, too: even though we have L-R audio monitoring and phantom power, the camera’s preamps have a very high noise floor and there is pronounced hiss from any sound recorded with the camera’s line in* – much worse than any of the other cameras I’ve used, surprisingly including the E-M5 and E-M1. Perhaps it’s interference between the IBIS system and some audio components. You’re still going to need a separate audio recorder. It gets even better: you only get audio monitoring if you buy the (expensive, and poor quality with loose rubber ports and insecure stowage for the separate port covers) optional battery grip – the port is located on the side of that. To add insult to injury, if you intend to use any of the ports on the left side of the camera – say mic in and HDMI out, which would be sensible – the swivel LCD is no longer usable because it cannot swing out. At best you can view it in the same (vertical) plane as if it were in the stowed position. That would appear to be quite a serious design oversight since it entirely defeats the point of the swivel LCD to begin with.
*I am using a Zoom H5 with the SSH-6 stereo shotgun and the XY-5 bidirectional mic. Recording directly, the recorder produces incredibly clean sound, and line out is clean on other recording devices.
Though the EVF is better than the original, and matches the E-M1, it doesn’t seem to have the same eye relief. It’s as though they plonked in a new and larger panel but never bothered to change the optics; I could see the edges of the original finder just fine with glasses, but now struggle. This was not a problem with the E-M1. Other ergonomic niggles abound, too: the strap lugs are still in a very uncomfortable position (unless you like pinched hands) – it seems this is an Olympus tradition. Even though we have no less than six programmable buttons, there appears to be no way to do three important things: assign one to AF-ON; have a consistent set between stills and video, and finally, save your preferences. I’ve had more than one instance where changing button assignments in video or stills mode arbitrarily affects the other mode without prior warning#. Finally, beware the high resolution mode – it can change your file format from RAW+JPEG to JPEG only without warning unless you manually change it while set to high resolution, and then manually change it again after switching back to normal. The setting then ‘sticks’, but who knows why.
#In all fairness, the Nikons also require resetting between video/stills modes – but at least they’re consistent. That said, the D750 manages this duality best – you can optimize U1/2 for video/stills and thus only require a turn of the dial.
Perhaps it is something Olympus can solve with firmware; after all, it wouldn’t be the first time they’ve had to patch something critical after release. They should probably look into a few other glitches while they’re at it – the camera locks up when the card is writing and you’re trying to change anything in the SCP; it still locks up under certain combinations of playback and zoom operation; it writes an additional .ORI file which is only readable if you rename the extension to .ORF**. I’ve also had a couple of black startups where you see shooting settings but no image – and yes, I did remove the lens cap. It might have something to do with the sensors in the collapsible Panasonic 12-32 and 35-100 zooms I’m playing with at the moment though (you do not get a picture unless the lens is extended and locked). I get the overall impression that the electronic portion of development was rushed and not fully tested, or at best compromised to focus on the high resolution mode.
**Updated.
I am honestly still on the fence about this feature. On one hand, it’s basically the same thing that Hasselblad has been doing for some time with its Multishot cameras; they really do deliver a bit more resolution so long as you’re working under fully controlled conditions. Unquestionably, there is more information in one of those files than a single 16MP frame. And unquestionably, color accuracy is both higher because of oversampling and higher because there are no false interpolation artefacts. But the shooting envelop for this feature is so small I cannot see myself using it; if I’m going to lug that size tripod around, I’d rather use the D810 – and if I stitch two of those images, the E-M5II is left in the dust. On top of that, the pixel quality is very, very low. There isn’t any more dynamic range, and acuity is lacking even with the best lenses. I’ve found the best results to be had by sharpening aggressively, downsampling to approximately 50% (32MP), and then sharpening again. This delivers pixel-crisp results with acuity comparable to the original 16MP single capture. Practically, I doubt most M4/3 users will be inclined to add the extra weight of a suitably sturdy tripod and head.
Single capture image quality for M4/3 appears to have stagnated for the last couple of years – everybody is still using the same Sony 16MP sensor, which is still good, but starting to show its age in several ways. There’s visible noise even at ISO 200; dynamic range is lagging behind the 24MP APS-C cameras with similar pixel pitches, and easily three stops down on the benchmark D810 and 645Z. Note: the only reason we’re comparing it with these cameras is because of Olympus’ own claim that M4/3 can now deliver competing image quality – sure, but if you’re going to take eight shots and stitch them, or overlap, or whatever additional processing that means extra capture time and the inability to use it on anything other than static subjects, that takes the 645Z to 300+MP, or a $300 D3200 body to 160MP even allowing for overlap. Part of that could probably be solved by going to a 14 bit pipeline – clearly the data handling ability is there, given the camera can shoot 11fps with a full global electronic shutter and 1080P60 at 77mbps. It is unclear why Olympus still remains at 12 bit for raw files since the data pipe appears to be more than large enough. On top of that, it appears that some minor shutter shock still exists when using the all-mechanical shutter option (necessary for flash, but you won’t usually see shutter shock because of the flash duration); full electronic (heart) isn’t always ideal because you will land up with some color shifts/banding under AC light sources and shutter speeds faster than the native power frequency*. I suppose electronic front curtain (diamond) is the best compromise. Lots of choice, but lots of confusion and you must remember to change modes or risk obviously compromised (double images, color banding) and unrecoverable results.
*This is not a fault of the camera but an artefact of the way our electrical system works – lights flicker, but just faster than our eyes can see – persistence of vision makes it look continuous.
I initially wanted to use the title ‘love and hate in small format land’, but perhaps that was a bit too ambiguous – even though it sums up my position at the moment. There are things about the E-M5II I think are absolutely brilliant – like the performance of the stabilizer with any and all lenses (the Voigtlander 180/4 APO becomes a light, incredibly sharp and easy to focus 360mm f4; the Zeiss ZM 1.4/35 Distagon delivers an absolutely beautiful video rendering) which makes both handheld stills and video very easy. That stabilizer also makes up in a big way for the sensor’s age – whilst I might need 1/250 to be comfortably stable^ with the D810 and Otus 85, I can shoot the E-M5II and 1.4/35 at 1/20s or less. I can also focus it consistently – I don’t have to guess with the optical finder, or use live view and a LCD magnifier and suffer all of the operational delays and glitches. The E-M5II is just that much more usable than the Nikon with manual focus lenses. Even discounting haptics, there’s a 3.5-4 stop advantage; it means ISO200 with the Olympus and ISO3200 with the Nikon. Whatever resolution, dynamic range and color advantage the Nikon/Otus combination had is largely gone by that point. Surprisingly, this means that practically and in low light, it’s probably a tie, and in good light, stick to the big guns – the high resolution mode is more of an emergency thing rather than something to rely on.
^Critically sharp at the pixel level.
I honestly think I may have found a creative use for the pixel-shift-stitch thing. It’s rather painterly.
I can work around the audio issue since I had to buy an external recorder anyway; I had to buy the grip anyway for additional power (though the power consumption of this camera is quite amazing – especially considering it is running live view and IBIS all the time, easily outlasting the D810 in a similar LV/VR configuration by a factor of three); and even though the D810 delivers a bit more video resolution and dynamic range, I lose some of that when I stabilize in post. No stabilization required with the E-M5II. The handling/firmware issues are annoying, but if you only use the camera for video or stills, not a big issue. Set it up for one or the other and leave it that way. If you use the LCD, eye relief is academic.
Perhaps I am being overly critical. But then again, is it too much to expect a $1,100 camera body that’s advertised to do all of these things to actually be able to do them? This is not a cheap camera by any stretch of the imagination, especially given what similar money buys you in 2015. People who buy one of these are going to want it to dance and sing, not dance or sing. I suppose the most frustrating thing is that in so many ways, the E-M5II is nearly ‘there’ – but is prevented from being so by a lack of attention and firmware that lacks common sense.
Addendum: There’s been a bit of traffic on whether the E-M5, E-M5II or E-M1 is the camera to go for – frankly, they all make identical image quality under ideal situations. If you work with stills and MF or native M4/3 lenses only, save money and buy the E-M5 – it has no shutter shock issues that I experienced in shooting with three bodies and north of 50,000 frames. If you might do video, or work under controlled lights in a studio, then the E-M5II is the best of the three by some margin. If shoot with legacy 4/3 lenses, then the E-M1 is the only one that offers PDAF. But otherwise I would avoid the E-M1 simply because you can only shoot single frame stills without shutter shock. Note that ergonomics might also play a part if you have larger hands: you must buy the expensive two-part grip for the E-M5; the E-M5II sells a one-part or two-part grip at proportionately increasing cost; the E-M1 doesn’t really need a grip (but has a single vertical component).
I suppose the final question is whether I would recommend this camera or not; and frankly, this does not have a clear cut answer. I think there are more compelling choices for M4/3 if you want compact and small (the Panasonic GM5, for instance) and better choices for mirrorless and pure video (GH4, G7, Blackmagic, Sony A7S, etc.). There’s no really compelling reason to upgrade from the E-M5 or E-M1 unless you need one specific feature of the E-M5II. But at the same time, there really isn’t anything else that has the same overall feature set and can fill the mission of a) handheld, stable video; b) acceptable stills; c) overall compactness. None of those other video options are stabilised without gimbals or steadicams (thus not quickly deployable or usable one-handed); and the smaller options are ergonomically compromised. This means it’s a good all-round choice of entry into M4/3, and despite of all its flaws, the E-M5II in fact delivers rather well on my original usage intention: handheld video for family and casual use, with an occasional still grab. If you take away nothing from this review, it should be that despite my previous experiences, I still bought one because it does what I need it to do. And that, I think is telling. MT
The E-M5 Mark II is available here from B&H and Amazon, along with the HLD-8 grip (B&H, Amazon)
The Zeiss ZM 1.4/35 Distagon is available here from B&H and Amazon.
The Zoom H5 audio recorder is likewise available here from B&H and Amazon.
Images in this post were post processed using my Photoshop Workflow II, available here.
__________________
Turn the mood and style up to 11: the Hanoi Cinematic Masterclass in association with Zeiss (21-26 July and 28 July-2 August inclusive) is now open for registration – click here to book and for more information.
__________________
Prints from this series are available on request here
__________________
Visit the Teaching Store to up your photographic game – including workshop and Photoshop Workflow videos and the customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!
Don’t forget to like us on Facebook and join the reader Flickr group!
Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved
Ming Thein | Photographer 
Hi Ming.
Would this camera with the pro lenses still match non full frame camera’s for IQ under iso 1000 or so? There’s still not a whole lot of attractive alternatives to it in its price range. At least not with weather sealing and good lens catalogue.
Probably not; the 24MP APSC stuff is really quite good.
Hi Ming, I got a good deal on one of these with the 12-40 lens. However I can’t seem to get the same punchy jpegs that were default in my original em5 and old pen too for that matter. Even the vivid and I enhance modes aren’t the same IMO..my question is what should I try in the settings to try and recreate that trademark olympus color I’m after? contrast perhaps? ..Cheers
At a guess, contrast and sharpening, but use the natural setting…
Would the em1 mark 2s colours be better or the same as the em5.2 Ming? I mostly shoot jpeg and can’t get the same look out of camera from the em5.2 as my original em5 no matter what settings I adjust..wondering if I should give the EM1.2 a shot or leave m4/3
I find they’re all about the same in raw…
Hey Ming Thein, I read your reviews a couple of times. Simply because I initially had major issues trying to decide whether to get the now 1 1/2 year old EM5 Mk2 or the upcoming new kid on the block EM1 Mk2. But re-reading your EM5 Mk2 and having posted questions on forums, I think the EM5 Mk2 will suffice for me. While I’d love to shoot some sports, I often shoot in the daytime in broad daylight so all I need is a good lens like the 75mm f1.8 (which I’d likely be able to afford if I bought the EM5 Mk2 rather than the EM1 Mk1!). Most of the time I’d shoot indoor family shots. Plus I’m moving up from a Panny GM1, so going to an EM1 will be a big shock to my system in terms of size! So thanks for your review, it helped.
The 5.2 is still plenty capable…I’d pair that with something fast and perhaps the 12-100…
Hi
I am having a hard time deciding between the E-M1 and E-M5 II. Where I live the E-M1 is slightly cheaper than the 5 II. Apart from the high res mode, is there any other thing where the 5 II is better than the 1? The shutter issue has been totally fix with the firmware 4 isn’t it? Even for the burst mode?
I will mostly use it with the 12-40 and 7-14 zooms, but might add the 45 and 17 as well.
Many thanks
Aside from the HR mode, the EM5II’s main gains are additional options for video frame rates (50/60p) and the fully articulated screen. If you don’t need these, then the EM1 saves some money – as you pointed out. Other than that, down to ergonomic preferences.
Many thanks for your prompt answer. Video will never be used. So I’m gonna head for the 1 🙂
Glad to be of help!
It seems a little bit strange to me, performing a review of the Olympus E-M5 Mark II without presenting an image or mentioning any experience using Olympus MFT – lenses for that review.
Please read my previous posts and note posting date. I have EXTENSIVELY covered M4/3 since 2012.
Great review, thank you! Olympus has since released a firmware upgrade that adds a flat profile and more. I am currently thinking about buying this camera and was wondering what you think about the new version. Does it make things better?
Can’t honestly answer that, I sold mine quite some time ago.
Hi Ming, quick questions on the OM-D E-M5 Mark II and camera servicing.
I recently had the good fortune of purchasing heavily discounted refurbished models of an OM-D E-M5 Mark II and a 75mm F1.8. Upon receipt, both look literally brand new and never used…but…the 75mm F1.8 lens was DOA as my existing m4/3 bodies failed to electronically acknowledge the presence of the lens. And despite the E-M5 Mark II having only 195 shutter actuations, it had a big greasy fingerprint on the sensor.
After speaking with Olympus, it sounds like they were completely out of stock of similar model refurbs and thus they were unable to do a direct exchange. So they offered to repair both for free in their service center. Given your professional experience with camera and lens defects and wear and tear, does it make sense for me to get them serviced under warranty, or should I just return them for a refund and wait for a future sale?
Generally, I’d be delighted to use these exact copies if restored to full working order and with a fully cleaned sensor. However, I’d prefer to not start from a bad point with both items if the servicing is likely to only lead to partial fixes and/or increase the risk of future issues. Is it possible for professional service centers to clean a fingerprint from a sensor without any negative repercussions to long term image quality?
Thanks in advance for your thoughts and happy holidays to the entire Thein household!
Uh oh. If somebody touched the sensor, there could be possible damage to the suspension mechanism for the IBIS. I’d be less worried about the lens (electronics are no big deal, but of course raise questions of longevity). Sensor cleaning is no big deal either, but the fingerprint suggests possible impact damage. I’d return it.
Thanks Ming! Sound advice. I’ll definitely return the camera body. As for the lens, I may ask one more time if I’ve got any other options. I sold a Sigma 60mm F2.8 and a Rokinon 85mm T1.5 Cine lenses to make room for the Olympus 75mm F1.8. So in the absence of that Olympus lens, I’ll have a hole in my telephoto portrait prime options 😦
Contax Yashica Zeiss 2.8/85 MMG. 😉
Thanks Ming, that would be a sweet lens on my m4/3 gear!
may I add…
Given the fact the supplier is lax enough to ship defective goods I would not rely on the repaired items being brought up to fully working standard. Who knows their could be other issues that you could not be aware of right now (so cannot inform them) that only become apparent when you get them back repaired. Refund.
Good point.
Thanks Michael, good point indeed! Thanks to both you and Ming, I’ve decided to return my E-M5 II for a refund. Fortunately, Olympus is willing to exchange my 75mm F1.8 for another copy.
The lens was the more important item for me as the E-M5 II was really only to replace my existing silver model with an all black model. I find the silver to be sexier and it’s a fun talking piece as people often think I’m shooting with a film camera. However, I was hoping to shift over to the black body to be more discreet (and to complement my black GX8 and GM5 bodies).
Thanks again!
I agree the silver/black bodies are very retro and have conned me into thinking film camera when spotted. I use E-M1 with blacked out logos, all my Oly kit is black. Contrary to what I read on-line, I find many uses for the 75 1.8, Stick with it, it is superb. I normally head out with E-M1 75 1.8 and E-PL5 17 1.8.
Thank you for your candor and objectivity. Rare to find a reviewer not pandering to the manufacturers. Oly does some things very well (IBIS, the whole 4/3 concept) and others you scratch your head and wonder, “what were they thinking”. Other mfgs have been able to squeeze a flash into their compact cameras for years and Oly just figured it out with the E-M10 II!!! How is it possible Oly could ignore and sweep under the rug the shutter shock issue for so long??? According to an Oly rep at B&H there won’t be an E-M1 update for at least a year while Panny is already using the latest 20MP sensor! Some of the ergonomics can only be described as bizarre as with the on/off switch on the E-M5, the accessory flash that requires keeping track of 3 separate covers, etc. Still on the whole I’m happy with the E-M5 as a great walking around camera for spontaneous shots. I look forward to your posts.
Well, Oly was in denial and then decided they didn’t want to collaborate with the person who found the problem for them. I say good luck to them…
Ming: Great review as always. You mentioned that the E-M5MKII hi-res mode would be part of your resolution comparisons section in the forthcoming Canon 5DSR part II. Can I ask how your processing the 99MP .orf hi-res raw files? Do you use the Oly plug-in or straight ACR? I’m finding there is a significant amount of detail lost with the ACR method. The plug-in seems to be the best way to start processing the file?
I just use ACR; the workflow for the Oly plugin is messy.
Ah yes of course, I understand messy. That said, to be fair in your comparisons, you might consider going the messy route so that you achieve the best possible result. The difference, when viewed at 100% in PS is really quite pronounced. Looking forward to part II! 🙂
The difference is that the plugin removes overlapping pixels but ACR does not. The first 4 shots are captured at 1px increments, but the following 4 are shot at 0.5px increments. Because of this, there is some overlap in the captured data, which is the reason that the 64 megapixel RAW does not look sharp. Both the plugin and camera processor compensate for the overlap, and this is why the resulting JPEG is 40 megapixels instead of 64 megapixels. I haven’t tried it myself, so I don’t know how the plugin works in practice.
I sent you a comparison.
Wouldn’t that mean the plugin is then up sampling instead? I’m not sure I see why we would need eight images if the overlaps are discarded anyway?
Not all of the data is overlapping. Here’s how DPReview explains the process. They also have a neat animation showing how it works, and at it leads me to think that each 0.5px shift overlaps with about 1/2 of each pixel that was captured by the 1px shifts. Like I said, I have no first hand experience with it, but to me it looks like the pure RAW data contains information that needs to be omitted for best sharpness. Downsampling to 40Mpx may or may not yield similar results, I’m not exactly sure.
http://www.dpreview.com/reviews/olympus-om-d-e-m5-ii/4
[quote]
Unlike the Hasselblad system, which uses four or six shots, the Olympus method uses eight. This is because it’s essentially doing two things: the first four movements shift the sensor in whole-pixel jumps in a square pattern. This is designed to overcome the limitations of the Bayer color filter pattern: moving each red and blue pixel into all the gaps that would normally exist between them and their nearest same-color neighbor, meaning you sample every color at every pixel. The sensor then moves up and left by 1/2 a pixel’s width and repeats its square pattern – theoretically doubling the vertical and horizontal resolution.
However, because there’s some overlap between the original set of four pixels and the set of offset pixels, the system doesn’t quite yield four times the original resolution (64MP). Instead, the camera compensates for this overlap and outputs a 40MP file. These files are captured both as 40MP JPEGs and as vast, 100MB Raw files.
[/quote]
Thanks. So it seems like the full overlap is for color information, which presumably is useful?
That’s what it looks like, yes. Judging by the DPReview image comparison with the D810, it seems to produce much better color because no interpolation is needed. The comparison is under the same link I posted above. If Oly can make this work at reasonable speed, say 1/40sec, they might have something here. Until then it is of limited use, mostly studio still life. Maybe it could work for watches as the larger DoF reduces the need for focus stacking? The 60mm macro should easily have the resolving power.
It doesn’t make sense for studio work because the net resolution is about the same (except now we have even more pixels to retouch), and I can’t honestly say I’ve had any issues with the D800E’s color under controlled light. Ditto DOF – you can always use the PCE macro.
Well there is still an advantage to be had: no moire. So it depends on your subject for sure. Also: I agree with the notion that if you are satisfied with whatever you have at hand there is obviously no need for anything better. My own thoughts on this are that EM1.2 with a good lens produces single shot files that are mostly hampered by my own shortcomings (bad composition mostly, sometimes setting a too slow minimum shutterspeed in fast sports).
For landscapes the Em1.2 Hires mode is very usefull though. I shot Noctilucent Clouds and here the difference between the Hires file and the Single Shot file is so big we are talking about and excellent pic with superb colours and a very clean yet detailed shot and a mediocre one. But that is when we pushing a m43 sensor to its limits. Also: Em1.2 as opposed to Em5.2 is much better at dealing with movement. It is not perfect for sue, but a lot better!
Hello Ming. The default resolution in Olympus cameras is 350dpi and I believe a very good print is at 300dpi isn’t it? Photoshop even ofer to open RAWs at 240dpi only. So what does that involve that Olympus as 350dpi on the 4000 something pixels wide image? That you can actually make bigger prints at 300dpi? Thank you!
The DPI count is arbitrary and only changes output dimension. 4000px wide at 400PPI will be 10″ – it’s nothing more complicated than that. Whether your output medium can actually resolve 400PPI is another matter entirely…
If the EM5 2 has anti-shock in continuous shooting modes, why can’t this solution be immediately implemented via a firmware update on the EM1?
Ask Olympus…your guess is as good as mine.
I’ve asked and they said they’ll get back to me. I won’t hold my breath ! Have you considered the Sony A6000 as fulfilling the family video/stills brief? Better image quality than m43rds, best AF in mirrorless world, better video than EM5 2. No IBIS but no problem if use stabilised lenses. No mic input or headphone out a problem?
Hello Ming,
which cable do you use, to connect the zoom with the camera?
Thanks, Regards, Stefan
A regular short 3.5mm interconnect I found on eBay.
Interested to see the results of your Sugru use.
Would using it as a permanent logo masking/blackout be something you’d consider, or not a consideration as you’re likley to dash the body once it’s outlived it usefulness?
Probably not. I don’t mind adding grip enhancements though since it’s unlikely anybody can argue with those…
I must say, your review is rich with negativity. So much so, I am not sure that you should have reviewed the E-M5II. Something else comes to mind, perhaps you should stick to taking pictures.
Well, one thing comes to my mind: you missed the point completely. Objectivity isn’t the same as negativity, and there were enough things I about it that worked that I bought one. If you were expecting perfect glowing praise, there is both no such thing as a perfect product, and there are plenty of other sites who are paid to say nothing but praise. Besides, if you think it’s the perfect camera for you, there’s no need to read what I think for affirmation. Just go make pictures and be happy. 🙂
Going from 12 to 14 bit RAW wont make any difference in real world. See here: https://photographylife.com/14-bit-vs-12-bit-raw and here: http://www.diyphotography.net/12bit-vs-14bit-raw-and-compressed-vs-uncompressed-does-it-matter/
Sorry, my eyes say differently. There definitely IS a difference between 12 and 14 bit, compressed and uncompressed, and it becomes especially obvious once you start doing any sort of tonal manipulation in low contrast transition zones (extreme highlights and shadows). Both color and smoothness are affected.
Agreed. Always a headache when it comes to post-processing the m43 RAW files. Tonal manipulation in low contrast areas(i.e. the blue sky) should be very careful or it would easily cause pic banding/posterization.
BTW, there seemed to be no official explanation from m43 alliance why there are no 14-bit raw to date, even the newly released EM1 Mark2 or GH5 are said to be stick to 12bit. If the claim is true that the new 20.4Mpix sensor has 1 more stop DR than its predecessor, 13+EV of latitude quantized to 12bit RAW could possibly make things even worse, right?
Maybe Sony is to blame as the sensor supplier?
I think it’s likely to be a sensor limitation – the EM1.2 definitely appears to have more than enough processing bandwidth for 14bit…
From what I understand the noise floor in u43 means that 14-bit output would deliver no improvement at all – the extra it would be lost in the noise. Some calculations based on the claims of improved noise on the E-M1ii seem to show that this is still the case, although it’s getting close to the margin now.
I’m not sure about this. M4/3 pixels are roughly 4u in pitch; this is similar to the pixels in the Sony A7RII, and the 100MP MF cameras. The A7RII showed a significant improvement between 12 and 14 bit, and the 16 bit MF cameras show even more dynamic and tonal range still, plus lower noise.
Nothing to really contribute to this discussion as I’ve never really shot much with m4/3, but would be interested to see how your video work progresses (not sure how you get samples on the net with copyright issues etc)?
And well written review as always – it’s been a while actually since you’ve done one (a good thing 😉 )
Thanks Junaid. Re. Video: some of it is up, some of it is with clients, some of it is embargoed, some of it is for their internal use, and some of it is my own family stuff which I wish to remain private. A lot of the pickups and b-roll in the workshop videos (especially HTS) is also shot by me. The other challenge is it’s never 100% your production unlike stills, which means it’s difficult to be fully happy with it – so I never show it. I guess it’ll happen one day 🙂
Hehe yes I can see how you can never be totally satisfied. So much possible tweaking.
I’ll have to convince you of a private showing whenever we next meet 🙂
Seeya in Chicago 😉
Very honest review and analysis. Very rare these days. Thanks.
Rare because it doesn’t sell cameras, bring you advertising or referral dollars or endear you to manufacturers…one of those cycles 🙂
Ming touches on an extremely important metric, one that just about every other internet reviewer ignores: price/performance.
This is a key design and engineering ratio, and is very high in the minds of consumers. Reviewers who are somehow comped by camera companies, or perhaps just want more click-throughs ignore it, other than say token throwaways when discussing something like a Leica. Ming deserves praise for going places few dare to, and calling out the manufacturers where they need calling out on price.
Looking at that video rig at the top of this article, I cannot help but think that this little guy is being asked to play in the “BIGS”. I think the reason I’m so impressed with the Mark II is that my expectations were in check when I opened the box. Cute little palm size digicam looks like a vintage recreation with pana20. I consider it nothing more than a disposable consumer imaging device, not at all in the same league with the known heavy hitters. It’s handsome to look at, it performs well beyond my expectations and it has a very nice compliment of lenses both native and 3rd party. All in all, an excellent value for what you get. I keep my cameras for 24 months before upgrading and I’m pretty certain I’ll recover half the cost at sale so it costs me a little over 20 bucks a month to own it. Add back in the fun and enjoyment dollar factor and the damn thing’s near free. Well done Olympus!
Thank you for this review. As a oldish cinematographer (did a lot of work with the Sony Digibeta) and still holding on to a NikonF2 film body with a crop of AiS lenses (incl the 50mm 1.2, 105 micro etc), its time for me to do the digitalvideo cum still combo. I have been seriously looking at the Nikon D750 with the 24-120 kit to do both documentaries and stills. This way, I could pack my bag with the F2 body loaded with B&W film and the new dSLR. My travels are to remote locations and I do want to cover both still and video. Does Nikon D750 today represent the best dual option. These project are self funded and later offered in the market for sale.
Depends if you have an external recorder for sound and video and proper stabilization…
I am looking at Tascam D70 and Zoom H6 for audio for my pro audio mics from Sennehiser. No plan to get a Atmos external recorder..I now use light tripods as the wrist grip of dSLRs is not to my liking. Would stabilization be a crucial factor.
Why not shoot with and without a tripod and see for yourself?
I live in a place where we cannot go out and check gear. The primary reason for posting on your site was the quality and content of what you like to share. Now that you have both the Nikon D750 and The Olympus e-m5-ii , for sheer image quality in both still and video, what are the considerations that you tick for using one over the other. Your comments will help as this decision on the camera will be with me for the next 5yrs. Thank you.
We can’t easily do it here, either – unless you are on the approved media list or chummy with the brand principals, the only time you get to try stuff is when you pull out your wallet. Those folks in the US have it better than they know…
I see what you mean about the product reviews getting far more comments than your (much more interesting) other posts…..
Perhaps it’s because it’s a lot easier to write off-the-cuff responses to a review? Internet readers are lazy too 🙂
Personally I’ve looked a few times at getting a used E-M5 v 1, it seems like for someone with small hands who takes only still photos with native MFT lenses, there’s no compelling reason to look at either the v2 or the E-M1. The high resolution mode is (to me) a cool feature I might use once a month, if I’m lucky…
Why didn’t you look at Panasonic for video? Lack of in-body stabilization only? Or other reasons too?
You’re right – the E-M5 v1 is much more compelling value than either of these options. Panasonic – yes, lack of in-body stabilisation. If I’m going to use a gimbal rig or steadicam then I might as well use the D810 (a gimbal is cheaper than a new system).
As for product reviews and comments…well, I’m starting to suspect it’s because you actually have to make photos to have an opinion on some of these posts… 🙂
Haven’t read all the comments here so I don’t know whether it’s mentioned but is the reasonably implemented image share feature another reason to look at this series of cameras,if you are into street photography. This article is really worth a read but there are potential ethical questions with its use.
https://photographylife.com/the-choice-of-presence-in-street-photography
As ever, a thought provoking review. Leaving aside the nonsense of dealing with a Manufacturer, I’m impressed with Ming’s ability to re-examine a previous position as near objectively as anyone could be expected to do.
I am a recent convert to mFT and in particular to Olympus having had the recommendation to consider it from a highly respected source. He suggested that the combined benefits of size, non-intrusiveness, speed of use, ability to react quickly to rapidly changing situations, relatively high image quality and a wide range of good to excellent lenses fitted my needs better than the kit I was trying to use at the time. He was spot-on. I invested in the system – and in a single day achieved a step change in my imagery. Simply put, the E-M5ii freed me from kit constraints to such an extent that I could just concentrate on the images and the story. I kid you not.
Now, I own most of the best photographic equipment money can buy and even so, I’m firmly of the view that you can use ‘anything’ to make a good picture. So, what’s important is that you use what best supports your objectives for that particular situation. In the Olympus E-M5ii, together with some excellent glass, I have found just that. It isn’t perfect by any means, as Ming says, but, as he also says, there’s no better alternative for what it does. You just have to be sure that what it does best is what you need from it.
Thanks for chipping in, Michael! Also glad to hear it’s still working out for you.
As someone who has had the OMD EM5 MKii for a couple of months, it’s refreshing to see someone call it like it is. This camera is a camera I really wanted to like. On paper it’s my camera. The reality of it is that it is a superb camera that can cover a wide range of situations (speaking for myself), but there’s still a couple of things missing I value, so I decided it’s time to part with it.
Ming, also I note your comment : ” It is unclear why Olympus still remains at 12 bit for raw files since the data pipe appears to be more than large enough”
I wanted to comment on this- you can’t just add a 14bit RAW and automatically have an image quality gain. The Sensor / DAC needs to be capable of extracting real data to work with- otherwise all you re doing is wasting more bits, more power, reduce battery life, creating more heat, and need a faster processor.
Even when the first APS-C’s/FF came out from Nikon/Canon capable of 14-bit RAW, a lot of them weren’t really using it.
Thanks. Re. 12-14 bit raw, you’re right, the implication is that the processing should follow all the way through. I’m just looking at the amount of data flowing out of the sensor in 11fps and the 1080p60 modes and wondering why they couldn’t allow a lower frame rate with 14bit and proper processing – this suggests to me there’s a hardware limitation somewhere between sensor and ADC which is a holdover from the first generation E-M5.
Nice Review. I picked up an em5 here in Japan for around $400 (US) last year and really love it. In fact recently you can pick up a used em5 ($300) and the grip ($150) for less than half the price of the mk2. If money is tight, I think this is a great option. I`ll stick with my em5 until we see an updated sensor.
I agree – and Japan is a buyer’s paradise for second hand gear…I wouldn’t be surprised if you could get a used E-M5II for not that much more. We were able to find used E-M1s a month after release at about 75% of the lowest new cost anywhere else.
Nice! Where do you find second hand gear though? I did not see any at the big shops (i.e. Yodobashi, BIC Camera) the last time I was there
There are tons of options. In Shinjuku, Map is the main place to go. There are quite a number in Ginza (Lemon, Sukiya, Katsumido etc.) and a few more out at Nakano Broadway.
Nicely done.
http://wp.me/60y6a
Thanks.
Hi Ming,
I found your review a breath of fresh air. It was a honest appraisal and (imho) thought provoking something which seems to be a rare occurrence nowadays.
I think that people forget that this equipment costs money and for many of us, it’s a significant investment and that’s why I feel that your voice is important in the sea of ‘this is the best thing since sliced bread’ extremely happy that I’m finally able to read a warts and all review.
Blessings.
Thanks Darren. In this case, being honest about equipment costs me money too – since I have to procure my own review samples!
I would have preferred a full-colour-sampling mode (ie, four captures shifted by one pixel width) to the high-res mode. Not that it couldn’t be implemented with a firmware update. I haven’t seen a good description of what the 64 MP raw file actually contains, three colour values for each ‘half/quarter-pixel’? If yes, a raw converter should be able to extract a full-colour-sampling 16 MP raw by simply ignoring every second ‘virtual’ pixel.
That would be similar to the Pentax model for the K3II. I would agree it’s a firmware update away: tell the sensor to take three shifted captures, each moved by one pixel, and then read/write RGB values separately. It should be faster, too – since there’s quite a bit less data to process.
The review of the EM5II on DPReview (and elsewhere) indicates that the EM5II is doing a series of full pixel shifts in a circuit of 4 shots followed by a diagonal half pixel shift and another circuit of 4 full pixel shifts. For some as yet unclear reason, Oly chose not to offer an option for outputting one or both of the 4-frame full-pixel shift sets as a full color file (ala the Hassy/Pentax approach). It’s also unfortunate that Oly chose not to offer an option of outputting the 8 frames separately (they just give you the one ORI frame). With PhotoAcute-type functionality added to Oly’s raw processing software that would allow for handheld HiRes with subject movement. I’m hoping that Oly expands the HiRes options in the EM1 successor because there’s a lot of possibilities with super-resolution when done right.
Theoretically, there should be a way to extract both circuits from the raw file since all of the information is captured – maybe we don’t even need a camera firmware update for this.
I’ve always wondered about the visible noise at base ISO (with the original E-M5). A recent Dpreview article about shot noise was quite enlightening in that regard ( http://www.dpreview.com/articles/8189925268/what-s-that-noise-shedding-some-light-on-the-sources-of-noise ). I suppose a small sensor is that much more susceptible, especially with base ISO of 200. ETTR helps, but the relatively low dynamic range limits its use. Funny, but you may be right about the OM-D’s being best at low-light photography (as long as you don’t have to stop motion).
Things could be improved noticeably with a ISO 100 base value, and read noise has also been lowered noticeably with the latest Sony sensors (at least in Nikon cameras, eg, the D810). So, I think meaningful progress is still possible with today’s technology.
There is an ISO 100 setting, but it’s a ‘pull’ setting which means some inevitable dynamic range compromises. However, for the target audience/ intended use of this camera, I think a stop more sensitivity makes sense – you’re unlikely to use this thing on a tripod anyway; there are cheaper and better options for ultimate image quality (any of the entry level 24MP APS-C DSLRs, for a start).
Bear in mind that all of the current Olympus cameras are still pretty much on the same sensor generation as the original E-M5, from 2012; it’s also why we haven’t seen 14-bit color or any meaningful improvement in image quality, either.
ETTR helps but as you say, there isn’t that much DR to work with to begin with. At least the noise isn’t unpleasant or distracting though.
I’ve used both E-M1 and E-M10 extensively, and I must say that ETTR is of limited use since these cameras tend to clip highlights easier than shadows. If you take a shot at ISO200 that is way underexposed, and push it in post by +3EV, the noise is almost exactly the same as if you would’ve selected ISO1600 in camera. It is so close I can’t usually tell the difference. However, with the ISO200 shot you have a little more dynamic range to work with, so it’s actually going to be better especially for situations where you need some latitude with the highlights.
http://www.mu-43.com/threads/76508/#post-772443
Your comparison point for the A7S is once again out of focus and of course going to show less resolved detail. If you are going to conduct comparisons and then insist your conclusions are right, at least do so in a meaningful and scientific manner!
Mine IS NOT a resolution and sharpness race !!! those are casual difficult light comparison and put side by side a sony a7s (monster low light) VS a tiny sensor like OMD show that you can see a tiny difference only watching at 100% magnification(or printing an 1 meter print) ,and the difference is not far from 5-10%,
so WHY DON’T YOU DO THAT?take a super performance camera with low pixel count like the A7s in normale iso range -1600 3200 and compare with an OMD with those settings ONLY for the omd…..with a perfect focus plane and let discuss the result!
Sharpness 20
Radius 0.7
Mask 70
Noise reduction 10-12
detail 70
Add Grain 10-20-50
Then what are you comparing? A ‘casual’ test is meaningless because the results are not comparable. There is no way of knowing if the image appears to have less resolution/detail because of your lack of care with focus, camera shake, or for all I know, you underexposed one and pushed it by two stops in post. No test or experiment of any kind has any value if none of the variables are controlled!
Lightroom and Acr have no idea in How a Canon raw is made,a Nikon Raw is made,a Oly raw is made….so the generic standard profile for raw conversion is not adeguate because it’s a generic starting point!
Try open an omd raw with those settings,artifacts you see even at low iso will disappears and detail will rest totally intact,improving the usability of the file even pushing shadow a lot! ,the noise you see it’s a mix of noise and sharpness artifacts,
Detail 15
Radius 0.7
Mask 65
Noise reduction 8
detail 65
chrominance 15-20 as you prefere
Effects panel Add grain 12 , 25 ,50
Hi own a OM-D E-M5 Mark II since almost three months having purchased the camera in the first days of may.
It is my first Olympus camera, dealing with the menu has been very hard in the beginning.
I got cramps in my hands when holding the camera more than a few minutes , in my opinion the camera is too uncomfortable to hold without grip.
I use mostly to shoot architecture or landscape in hi-res mode and find it very useful and effective, minor quirks aside.
(I’ve got the two pro zooms and find image quality very, very good, when not phenomenal).
After almost three months of use, yesterday, suddenly the camera started to refuse to shut down when switched off, the only way to shut down is to remove the battery and when the battery is inserted the camera will turn on again, regardless of the switch position.
So the camera is now on his way to Olympus CH for repair. I am very disappointed, such a poor QC for such an expensive small tool.
Will keep you posted.
We had that problem intermittently with the E-M1s and E-M5, too. Lack of wide shift lenses makes it very difficult to use for architectural work.
I’m tempted by the Mark II… that or one of the Merrills, probably the DP2/3. But for what reason, and for how often? Before I found Ming’s blog, I supposed I intuitively knew about shooting envelopes without having consciously codified it. But when I sit down and attempt to justify a Mark II… I can’t. I can’t make that leap.
The truth is it only expands the envelope for a very few things. And if those things aren’t what you do, move on. Or take some medication for the GAS. 🙂
I disagree totally about Nikon FF gives 3 stop better Raw…you have to be careful derawing any omd raw because it need the right conversion,it’s nonsense open every raw files from any camera with the same settings in camera raw ,are sensor all the same?resolution is the same?? I can show you with those settings converting a omd raw improve minimum 2,5 better clean file with tons of detail
Sharpness 15
radius 0.7
mask 50
noise reduction luminance 10-20 from 1000 to 6400 iso
detail 70-75
noise reduction color 20
Effect panel
Add grain 10 – 50 – 20
Here you can see my test and download the files
http://www.mu-43.com/threads/76508/page-3
I do not open every raw file with the same settings, and I spend a lot of time finding the right profile for all parameters for each camera. Why would I spend so much effort chasing ultimate image quality and then compromising it? I think you’ll find that DXO also agrees: the D750 has a two stop noise advantage, to say nothing of dynamic range and bit depth.
However, your test is false because your Nikon files are clearly out of focus and thus not comparable. On top of that, you are using comparisons of 3D subjects which require sufficient depth of field to cover and fully resolve, yet you use the same aperture with both Nikon and Olympus – this gives the Olympus a 2-stop DOF advantage, which will of course make it look sharper. Increasing noise reduction – which will remove detail – and then adding back grain – which is not real information – is giving you the illusion of more information with no real signal. If you try doing that on a very high frequency subject, it’s clear the FF cameras have a significant advantage still.
Finally, you’re missing the point entirely: I also pointed out despite all of these limitations, the M4/3 camera still comes pretty close in practical deployment because of DOF, stabilisers, lenses etc.
More dof mean more area in focus not more sharpness! In my exaple there are a lot of 3d object for be sure there are some parts of the pics perfectly in focus…raw are different eatch other you know …and you know that it need to be built with the right conversion…if you comprare a d750 with a sharp lens even stopped to f4 VS a m4/3 omd camera with ANY lens you will find more apparent sharpness in omd files..why? Because the raw need an other approach….try yourself please! A nikon d750 + a nikkor 24 70 @f4 and the same composition with an OMD with equivalent 12 40….at moderate iso like 2000-3200 shoot a typical potrait …open orf file with my suggest settings and share the result….you can put my name on the post…you can find less than 10% in detail/noise compromize,not 1 stop,not 2 stop,and figure 3 stop ?? I hope you will try even to put contrast to -100% and look the shadows.Best regard.
Of course the noise difference is much less when you compare at the same DoF and same shutter speed, because that will force the D750 up 2 stops in ISO. The stabilizer levels the playing field even further when there are no moving subjects. So in practical use, the difference is not +3EV, and that’s exactly what Ming said.
However, that doesn’t remove the fact that the D750 is more flexible with DoF control and can exploit that +3EV in certain situations. In addition, it has the advantage of 14 bit RAW against 12 bit. That’s 4 times more information, and it really shows when you start pushing shadows and pulling highlights. With the OM-D, there is not much headroom in the highlights, it requires much more care with exposure.
If it’s not in focus, how can you judge resolving power or sharpness?
I’ve tried your suggestion several times and all it creates is noise over an underlying smear. And you can always do the same thing with the D750 file and we’re back to square one, except now both are smeared.
The .ORI files are readable by Bridge – ACR – Photoshop. All you need to do is change the extension .ORF — change xxxx.ORI to, say, xxxORI.ORF. Also, the reason why the 64MP high res files are not really 64MP is that the image pixels obtained from the half-photosite shift are not independent of the unshifted pixels. The pixels effectively overlap. http://philservice.typepad.com/f_optimum/2015/04/olympus-om-d-e-m5-ii-analysis-of-high-resolution-images-and-comparison-to-sigma-dp2-merrill.html
Note on ORI. Seems rather misleading to output and claim 64MP though…
Its a 64MP raw because its raw. The jpegs are 40MP for a reason.
While, as both an E-M5 and E-M5II user, I don’t agree with all that you say, it’s still a great and refreshing review.
I’m using the High Res mode with some nice results with only a light (but high quality) travel tripod.
You mention the the ORI Files are useless – rename them to ORF and they turn into regular 16MP RAW files (it’s the first of the 8 frames in the High Res photo). 😉
Well, I come from the perspective of somebody who was heavily invested and a big M4/3 fan until things started going sloppy, both technically with the E-P5 and E-M1 shutters, and support-wise from the principals. It makes no sense to compromise if you don’t have to – but here, we have a solution that is compromised but no other alternative. So there is no other conclusion than it is still the best tool for the job – and that is always what I am looking for.
ORI files: aha! Thanks for the tip. No program could read/preview them before that…
About the black startups: I’ve had them very badly right after getting the camera – in fact could not make it work in the first hours with the camera and thought about returning it. Suddenly it started to work, and now I’ve had not got them in a long time.
But when these black startups were occurring, I could track it down to one cause: something in the camera mount. Removing the lens and mounting it again applying a little bit of force when turning the lens into place made the camera turn on ok (not FORCING the lens really – only putting it in locked position firmly, turning the lens until you sense that it is really in the position, not just mounting it fastly). But now even this was not necessary, no more black starts – I think that something might “conforms” in the mount’s connectors.
That was the same conclusion I came to, because I have not seen it with adapted (i.e. completely non-electronic) lenses or non-collapsing ones. I suspect it may even be the contacts inside the collapsing lens barrel not fully engaging or returning an erroneous signal.
The OM-D menu system is a huge mess and responsible for many of the usability quirks. Optimally, they should just re-build it from the ground for E-M1 mk II. The SCP should be totally customizable. Additionally, there should be a custom menu that you can call up from an Fn button. Possibility to quickly set live view boost on/off, I hate it when I have to dive into the menus when using flash. Possibility to completely turn off the LCD, now it is always back-lit even when nothing is displayed. Possibility to do something useful with the trashcan button in shooting mode, now it only works in playback. Possibility to manage files in camera, like creating folders and moving files between them. Some settings have default values that make no sense (shutter lag, EVF refresh rate, JPEG quality etc), these should be set for best possible performance as default. The list goes on.
Oddly, no camera manufacturer I can think of is innocent of setting meaningless defaults – none of the cameras give optimal quality out of camera. JPEG basic, super saturated consumer auto and crazy ISO limits? Ugh. But yes, the menu was okay with the very first generation because there weren’t that many options; it’s now gotten so unwieldy that it takes a long time to find what you need. And some important options are just not available (AF-ON, for instance).
I don’t understand, do you just mean that there isn’t a dedicated button marked AF-ON? Even on my little E-M10, I am able to use the Fn1 button for back-button focus, are you not able to set focus-acquire to anything other than the shutter button on the E-M5 MkII?
No, I’m smart enough to use a programmable function button.
The problem is none of the possible combinations give you AF-ON on one button, AE-AF-L on another, and shutter only for shutter. And on top of that, there are two places to set your buttons: one in the custom button menu to assign AE-AF-L, then another to set shutter behaviour which affects AE-AF-L. And to top it off, this behaviour isn’t consistent between still and movie modes – and there’s no way to make it so.
Ah, got it, thanks! You never know with new cameras, sometimes they take things away that you had even with older versions! 🙂
Yeah…but sometimes they never get it right, either – even though it’s only one little line of code in a firmware update away 😦
It may not be so easy.
As a software guy (in recovery), I’ve worked with chip designers, and they just don’t think like you and I. And from looking at the manual, I can sorta figure out what they’re doing, and it isn’t going to be a one-liner. It certainly looks like it can be done, at least to me, but I wouldn’t guarantee it without seeing the source code. The problem is that the abstraction of the button is tied to the AF mode, so that the function of the button is dependent upon the behavior of the shutter release. To me, as a software guy, this is nuts, but as I said, chip designers are a different breed, and it probably makes sense to them. Or maybe its one of those “we’ve always done it that way” things left over from 15 years ago.
Oof. I take back what I said then…but given that it’s so easy for them to reassign buttons in other modes, it makes no sense. All of the Olympus cameras I’ve owned and used have behaved this way – even the E-1. That said, consistency between shooting modes can’t be that much to ask for, surely?
As a long-time AF-ON BBF user (on Nikons and Canons) and a recent m43 convert with the E-M5 ii, I’m confused about your concern. I assigned “AEL/AFL” to Fn1 and am quite happy with the results (so far using S-AF, but not having experimented with C-AF and not missing it so far). Even ergonomically, Fn1 is pretty easy to find due to the location behind the obviously tactile lever.
I do appreciate separating AEL from AFL on the Nikon bodies (even on models without a dedicated AF-ON button, most recently relying on shutter half-press for AEL on my D600 with AFL for BBF using AF-C on the AEL button), but having a live histogram and EVF makes up for that as I can easily use EC if the exposure is non-optimal.
Also, regarding black screen startups, that seems to have disappeared with FW version 1.1 (or is it 1.2; sorry can’t remember). Did you take the update?
I’m on latest firmware, and it still blacks out occasionally.
There are situations in which you’d want to decouple AE and AFL (off centre subjects, tricky light). I find this need quite common for video work, though I agree with flashing highlights it’s less of an issue with stills. It’s just much faster to move the camera til the histogram looks right, hit the AEL button to lock exposure since it’s a toggle, and then set initial focus precisely with AF-ON. There’s just no way to do this.
On top of that, even if you set AF-ON to fn1 with the S3/C3/M3 modes, when you switch to video, your AF-On moves to the DOF preview button and fn1 defaults back to AE-L. This continual change more than anything is extremely confusing – and completely unnecessary…
Thanks for another great review and some more stunning images. I’ve been looking at this and others you mention. I’d grab a Ricoh GR but have a hard time without the VF…never get the framing (and can’t see much at all in bright light)
Thanks! Too bad Schneider never delivered their promised 14/2 for M4/3…I suppose there’s the Panaleica 15/1.7 though.
Doh! I hate to be baited by gear… but since you mentioned it…
The ZM 1.4/35 Distagon clearly must be able to justify itself! Did you ever think you’d own another M-mount? If you balance speed/size/build/performance it actually looks like remarkably good value.
Hard to say, actually. I think it depends on whether a) the price is right and b) it delivers sufficiency or beyond (which I suppose goes back to the whole relative value question). But that ZM 1.4/35 is really something – it’s remarkable that it does well on M4/3 despite the super-thick sensor filter packs; I’ve not seen this from any other M-mount wides. I see why Lloyd Chambers gushes about it now. 🙂
Lloyd certainly makes it sound like good value (fair price?) when placed in the same tier as Leica and perhaps even Otus lenses. With its contrast and pleasing bokeh, I imagine it aught to justify itself by occupying a unique cinematic portion of the shooting envelope (and no doubt others).
I suppose with M4/3rds only the central portion of the lens, where the rays are relatively ‘parallel’ (or perhaps more accurately, not oblique), is being used. This might make the thicker filter pack more of a non-issue than on larger (non-native) mounts where the ray angle in the corners may cause trouble.
Zeiss certainly appear to be motivated in pushing the limits. It is interesting observing the engineering from the sidelines. What an age to live in!
I wish you well with it. I have no doubt you will put it to good use!
(Also, your images make me realise I take ceilings for granted – I should look up more often)
I’m not sure it’s quite up to Otus level, but definitely better than the Leica options (which are all overpriced) but it is of similar price/quality ratio as their own 2/135 APO.
Even though M4/3 only uses the central portion of the lens, there are still serious issues with a lot of lenses because they are not even telecentric enough there – there’s a reason all of the M cameras have to have offset microlenses to compensate for this.
Though, I admire Leica, it has been a long time since their value proposition made any sense. For half the price of a Summilux, with equal or better optical performance, the ZM is a no-brainer on the M-mount.
I am not sure how well ZM comparisons translate to other mounts. Since I mentioned Otus lenses: about half the cost, about half the weight and maybe 90-95% the build and performance? The quality/price ratio must be decent! Meaningfully more so than other cheaper lenses? Dunno. We’re well beyond sufficiency and into diminishing returns in this league at any rate.
An F/2.8 Otus would interest me. I’d certainly like to see how much smaller and cheaper a no-compromise ‘day-time’ lens could be made.
Sounds about right – the challenge with SLR lenses is the long flange distances mean a lot more optical gymnastics has to go into ensuring telecentricity and acceptable corners. I’ve mentioned slower (but no less perfect) lenses to Zeiss before, but was told they didn’t think the market was sufficient.
I wonder if that is an education problem?
Perhaps to joe consumer, prime lenses = fast, therefore I wont buy a prime lens if it is not fast? I suspect many photographers are guilty of that way of thinking as well.
Aside from composition choices (subject isolation), surely as noise performance continues to improve, the requirement to use fast lenses aught to decrease… Oh well…
Wouldn’t it be nice if Voigtlander could find a profitable niche rebooting its APO-Lanthar series!
Could well be – especially if the lens is expensive. Like it or not, consumers still make up the bulk of sales, not pros.
I suspect the problem with the APO-Lanthars wasn’t the price or profitability, it was that they were too close to the ZFs…
I do enjoy your reviews; no one can eviscerate a piece of imaging equipment better than you Ming 🙂 I do think Olympus has given the industry a nice nudge forward with respect to this sensor shift technology. If one doesn’t try to implement it beyond its current capabilities, it can be quite useful and even enjoyable to use. I have been able to coax some fairly decent, if not excellent hi-res macro shots with the Oly 60mm and the aggressive sharpening you mention. We will hopefully see this technology improve when Sony and others join the fray – Diglloyd is gushing over the future potential of sensor shift.
Well, at least you know it’s honest because I bought it…
Yes, I do believe your feelings toward Olympus and this product are very honest 😉
Let me clarify: the local principals, not HQ. And despite that…I still bought it. I guess I am precisely the kind of idiot they want after all!
Thanks for the honest and thorough review Ming. Question; did you like (within it’s obvious limitations) the performance of the Panasonic 35-100mm f4-5.6 in the 75-100mm range?
It’s a surprisingly handy lens, given its size and price. Only good at two apertures though – 5.6-8 at that range, limited by diffraction and the lens’ design. Make sure you have enough light 🙂
Thanks Ming, I plan to use it in good light (fast primes in low light) so no problem there.
While back in Boston a few weeks ago for a conference at MIT, I shot a night Red Sox games with this Pana 35-100/4 lens and an E-P5 at ISO 400. It did rather well, although the corners do lag a bit. Burst mode action shots of the plate and first base action from well up in the stands were surprisingly sharp crisp. I was quite surprised and pleased, given the low cost and its minimal size and weight.
It’s actually not that difficult to design a moderate speed lens which doesn’t have to covers a large area; fast/perfect/large format – that’s where things become challenging.
Hi Ming. Any idea what the maximum recording time is? AFAIK, all cameras coming to the EU have a maximum of <30min because of taxing reasons (an additional… 0,1% or something), but there are rumors that if someone just puts a Japanese firmware this limitation does not exist. It is extremely difficult to find this information on any fora, for the simple reason that very few people would be shooting more than 30 minutes of continuous video. It would require external power supply, etc. But for seminar recording it is indispensable.
Thanks for your time.
Mine appears to be limited only by card size – I don’t see a countdown anywhere. That said, I have not ever run a camera >30min for a single take…
Battery life is actually less of an issue than you might think. I have shot a 8h operation – with about 5h of actual video time – on a single battery, with a bit of charge remaining at the end. These are surprisingly power-efficient cameras, especially compared to the Nikons (1-1.5h of video and a larger battery is toast).
Aah, a very special and select selection to illustrate!
( And S. Dali can go hide himself.)
And thanks for the revealing review!
Still hoping for an affordable higher resolution OM-D-quality-IBIS camera…
– – –
[ My _temporary_ solution (for a longer reach) is the now cheap EOS-M + MagicLantern (for focus peaking and zebra clipping warning) with an EF-S 55-250 IS STM,
rather usable with a screen loupe, but quick only in auto modes – only one dial. The shutter is fairly quiet.]
Well, if only Sony would stop compressing their raw files…
Yes.
And make a better IBIS (referring to your review above).
And combine it with the more affordable APS-C sensor.
Or Olympus adopting the possible future Fuji-Panasonic new technology (higher res.?) sensor.
( High resolution for moderate digital zoom between fewer primes – and for large prints.)
I think more likely is the 20MP M4/3 sensor that sony just announced…it is doubtful this is going to improve on dynamic range at all. Or show much of a meaningful increase in resolution over 16MP, for that matter.
Aye.
( And I love the way your watch photography has improved.
Dali can really go hide himself. )
Haha, thanks!
Seems to be harder for them than for me to quit smoking !!
But at least you *want* to, right? That’s got to count for something.
Then smoke the Ming(?) (or perhaps my) way,
a cigar or pipe after a good dinner, or occasionally,
without inhaling.
Shouldn’t influence health too much.
– – and it helps keeping your fingers from your camera when possible shots come to your mind – and they stop doing if it’s a good smoke.
I literally don’t move my cigar away while shooting. Love the dodge and burn natural 🙂
I hope you have a weather.. smoke-sealed camera!
😉
I can imagine smoking influencing shooting positively,
but I can’t really imagine shooting making the smoking more enjoyable?
Well, it doesn’t make it worse. And if you’re doing very long exposure astrophotography…
“.. it doesn’t make it [the smoking?] worse.” : That’s *really* multitasking!
But you will have to avoid distorting the milky way. 🙂
Perhaps with a superlong tele wide open (for lack of dof).
Or by choosing a night with a light breeze – very light to enjoy the look of the smoke.
Saves us renting a smoke machine for the cinematic shots!
o
o
😮
Perhaps they will consider doing so about the same time they start planning a serious sert of e-mount lenses.
(sert) > set , Sorry.
We’ll see Kristian. There’s a few signs in the horizon ‘if’ the Zeiss Batis series are indicating a new aera with a serious RAW output from the E-mount cams. It’s not so that Sony did not hear the message. I think they did.
Thanks for your interesting articles. You ditched the 1st OMD for some time and had been using DSLR and medium format cameras since then most of the time for many months for more superior image resolutions to meet your requirements, probably you had used to carrying them around wherever go go not to miss any shot and for maintaining consistency in your image quality. If that’s the case what’s the point to have the OMD back, if it can’t meet the Ultraprint requirements ? If you are going to use it for casual shooting and teaching, is there other better choice of cameras (you mentioned few in your last paragraph) or just merely out of curiosity and G.A.S. you had the OMD back 😀
I thought I mentioned several times in the review it’s for video work…
Great review. I’ve been wanting to get into MFT for a while(because of size and lens choices) , but I’m unsure of whether or not I should get the E-M5 MKII, the E-M10 or just pick up a used EM-5. The E-M10 is half the price of the MKII right now(and I get $100 cashback as well) with a kit lens, but then again, I want the high end features(because of GAS)
I’d skip the E-M10 because the IBIS system is cut down and not as effective. It’s either a used E-M5 if you’re shooting stills only, which has no issues and pretty much identical quality to the rest, or a new E-M5II if you’re doing video. The only time I’d pick the E-M1 is if you need PDAF on legacy 4/3 lenses.
A friend of mine has the E-M5. Maybe I’ll borrow hers to try ut out before I make a decision.
The EM5 (original version) also suffers from shutter shock. Not sure, Ming, where you got the impression that it wasn’t subject to the problem.
From extensive use of two bodies, testing three additional others, and 50,000+ frames and never seeing it?
Speaking of testing bodies, Ming, did you ever test another E-M1 to see if shutter shock was impacting all of this model, or just a portion of those produced? I know they introduced the firmware updates, but it is my understanding that this does not fully eliminate the issue. Any thoughts?
–Ken
Between my partner and I, we owned four bodies. I spent a day testing an additional 89 bodies at Olympus HQ, which is basically everything they had in stock at the time. They were from four different batches, and two of those bodies were purchased outside the country. All of them exhibited the problem. Frankly, this is something that should be done by the QC department at the factory, not an end purchaser!
Ouch! That is a lot of bodies to test. Thank you for the detailed reply. I have had an E-M1 for some time and while I love the ergonomics, I have found what I believe to be a higher than normal (for me) number of shots that do not seem as crisp as one would expect they should be. Now, I know that user error and lenses are also possible culprits, primarily the former, but I just cannot seem to put my finger on it. I am not ready to completely put the issues on the doorstep of shutter shock, but at least I know it is not an unreasonable culprit.
Thanks,
–Ken
You’re telling me. We thought it was a one-off issue at first because the initial (preproduction) review unit didn’t show the problem, but turns out to be most certainly systemic 😦
Given that I had one of the highest technical hit rates ever (i.e. perfect focus/no camera shake) with the E-M5, the soft blurriness or double imaging seen on the E-M1 was perplexing. It’s gone on the E-M5II.
Given your very careful approach and attention to detail, I’m surprised that you didn’t see any shutter shock with your EM5. I certainly have with mine, and many others have as well. Here’s a particularly insightful and carefully done analysis of shutter shock on the EM1 and EM5 for your edification:
http://www.dpreview.com/forums/post/53456685
I would certainly have mentioned it if I saw it, but generally if you don’t see it in several units chances are it isn’t an issue. Though there’s no way to be sure unless you test *every* camera, which is of course impractical. It was very obvious and consistent with every single E-M1, though. All 80+ cameras I shot in one way or another…
Do you remember, when shooting on the original E-M5, whether or not you kept the image stabilization always on, even when shutter speeds would not have warranted its use? And did you use the anti-shock feature, which also creates at minimum a 1/8 second shutter lag (although this option might have come in a firmware update after your time with the camera…)?
I am definitely struggling with shutter shock on my E-M5, which I’ve had since its release. I am doing my own testing now on how much impact IS on/off and anti-shock on/off is having. And I’ve come to be wary of using any shutter speeds between 1/80 and 1/250… In any case, you seem to have shot a lot more with the E-M5 in your short period than I in all 3 years, so I would very much appreciate your answer to the questions above. Thank you!
Cheers,
Sami
IS was always on, and I never used anti-shock because of the lag introduced. I always ran the latest firmware. It seems I might have gotten lucky with my two cameras, and unlucky with all 80+ E-M1s(!)
This review is definitely very interesting to me, not because I shoot Olympus or plan to, but simply because I had the privilege of watching you shoot with it for the first time and all the frustrating bits and pieces that followed. But I highly doubt there is anybody else who has shot this camera with a stopped down Nikon PC-E (I still recall your Eureka moment) or under as many different conditions as you did for this review.
Well, it was a bit of a faff to get it to stop down – mount on Nikon body, select aperture, press stop down button, unmount without turning camera off so magnetic diaphragm stayed engaged…let’s just say it probably wasn’t necessary other than for curiosity!
I’m not really sure why Olympus didn’t consult photographers or filmmakers enough before releasing this camera. Well I guess the photography bit is ok, but if I was just thinking of a camera for stills I’d stay with the E-M1 or E-M10. For video this was what everyone was hoping for, a much better codec combined with the fantastic IBIS. Unfortunately they should have released it with a 1080p image at least equivalent to the current Panasonic picture and have considered the functionality and firmware a bit more. This camera could have been a runaway success similar to the original E-M5. I think I’ll wait to see if they improve things in the next iteration. Until then I will stay with my unstabilised Panasonic cameras which seem to be better hybrid devices.
They might have done, but it looks like only some of my feedback from the first generation (let alone the E-M1) made it back. No idea if it was filtered locally or at HQ level. I still think this is a better stills camera than the E-M1 though – we do not have any issues with shutter shock and continuous shooting, and IBIS is further improved still.
I think there may still be hope for the video output if we use an external recorder via HDMI out – this option might bear further investigation, but really increases the size of the whole thing.
I was wondering how well this camera would pair with the upcoming Blackmagic Design video assist. Seems like a relative bargain at only $495 and also fairly compact. Might be worth considering.
Good question. Guess we won’t know til there’s actually a product to try…
The Atomos Ninja Star could be a great pairing too. But I have not found a single footage with the pair – and only a single mantion about someone which tried and said that saw no gains (without footage). Could be true – looks like the problem is in the sampling method, not with the codec.
I was looking at that too, but the recording frame rate appears to be limited to 30p. There was one mention of clean 4:2:0 HDMI out on another site, but I haven’t seen anything else. Could possibly be that nobody has tried it because if you’re going to rig one of these things, chances are you don’t mind the size and just went for something with better native video quality on a gimbal?
The Blackmagic Video assist offers 10 bit 4.2.2 ProResHQ recording in 1080p 24/25/30 & 60 and the E-M5 Mk II outputs YCbCr 4:2:2 via HDMI so it seems like it could be a good match. Regarding the size thing I was figuring that it could be cool to have a long HDMI cable, stick the recorder in a coat pocket and still have a relatively compact shooting experience, with the significant benefit of the wonderful 5 axis IBIS. For the extra few hundred bucks I wonder whether you’d see a significant improvement in video quality. It has to be worth trying.
That does look interesting…
I’m in love with the images Ming. And on top of that they sell cameras!
I am on standby until I see a vast improvement of the sensor performance. Until then I’m good with the EM-1.
Thanks Gerner. It seems that is not what the principals believe though…they think fanboys and blind optimism and higher spec sheets sell cameras 😉
I would say let them believe what they want .. the camera isn’t going to get you one mm of compositional value .. so they’d be better off nursing the good photographers out there and colaborate. Then they would perhaps contribute to a better bottom line for big Oly!
I don’t think they see it that way…
At these days, every camera and lens manufacturer needs some support! 😉
I’d better go buy a spare then!
Ah, the Voice of Independence! Thanks for this review Ming. Very honest and enlightening. Can’t help but feel that some issues can be rectified by FW upgrade. Olympus built a good rep with the E-M1 FW upgrades, so there’s hope for the future.
One question on recommendations: a second body next to an E-M1 with FW 3.1? Another such E-M1 or this E-M5 Mk II?
Thanks Peter. That’s the good and bad news: bad because they shouldn’t have been issues in the first place, good because they’re very easily solvable. And then we go from a 80% camera to a 95% one.
That last question is tough: the E-M1 still has no way to have continuous shooting and shutter-shock free images; the E-M5II lacks PDAF for legacy lenses but has much better video. On the balance of things, I’d be inclined to say E-M5II actually…
I agree on “equipment issues”. They shouldn’t be there in the first place. As discussed before, the manufacturers should stop using its base of followers as beta testers (guinea pigs) for their rushed-to-market equipment. Apparently show schedules, press releases and beating the competition are more important factors than being a reliable brand for their customer base…
On the second cam question: I suspected the E-M5 MkII would have you preference. Currently has mine too. And although I’m not the one to procrastinate, I may wait this time to see what the E-M1 MkII will give us…
Well, the customers will only play the fool once. After that, we vote with our wallets. All that’s left is the shallow hope that there are enough people who have no choice but to buy your product, and frankly, that’s a really naive way to run a business. But hey, what do I know – I’m just a photographer. 🙂
The E-M1 II will certainly be expensive. It’s already not far off a D750…and I know which I’d rather have at that price point. There is no comparison on image quality – stills or video.
Thanks for transcending your difficulties with Olympus and providing a review. I thought when the multi-shot high res feature was given a lot of attention at the time the camera was released that it was a neat trick — but only a trick, not offering a lot of real world value. Even with the shortcomings you’ve disclosed, the E-M5 II looks like a good fit for me. Now to find a way to afford one.
Oh, and despite the built-in glitches and the incredible level of irritation provided by the manufacturer, every time you pick up an Olympus camera it starts spewing forth the most gorgeous images. There seems to be some form of synergy there. I’m sure the added Ming factor is the dominant contributor, but the result is always impressive. Even your earlier Olympus lens reviews show it.
🙂 Thanks!
Do you see yourself keeping the E-M5 MK II for the foreseeable future?
Yes, because it does what I expected it to – no more, though. Also, it’s rather expensive to buy just for a review, no? 🙂
But didn’t you do that with the Fuji XT-1? 😛
*I jest, I am aware that the handling (especially the direction the lenses turn) did not agree with you*
No, I never reviewed it. And yes, I lost quite a bit on that one, too. It wasn’t the lens direction, but the shutter speed and ISO dials…which you cannot change, because they’re engraved…
But it found a good home with us!x
My loss is your gain. Glad it went to a good friend!
Wow, so you feel out of love with the X-T1. Interesting.
Couldn’t get used to the dials spinning the wrong way and no way to reverse that. 15 years of muscle memory has them going the other way. I was missing shots and it was driving me mad.
Yes, and using the other system regularly, and having them work opposite ways was a pain, I am sure! I have tried the Fuji bodies, and even though some people feel they touch their soul or something, I just found them massively painful and fiddly, as someone who shoots all manual. I don’t want the camera to get in my way of doing what I want.
Well, ergonomically they feel great. But hard-coded dials are a blessing and a curse, depending on which way your fingers think is positive. Mine were the opposite direction and they definitely got in the way. Since it was easier to change the camera than my fingers, that’s what I did…