Roger's Corner

Mirrorless, Mirrorless on the Wall . . . . Part II

Published August 24, 2012

And Now Some Meaningless Speculation

These are simply my observations and speculations. I have absolutely no real information about any manufacturer’s plans. If I did I would have to sign a nondisclosure agreement and couldn’t write this stuff. But I like to speculate. It’s sort of like playing Fantasy Football with camera manufacturers.

I’m going to split this up by the various players. To me the most impressive and enjoyable thing in this entire area is that there is a LOT of variation. The companies are really trying different things. Not all of them are going to work out, of course, but some of them may provide us some really new and different tools to work with.

Micro 4/3


As the most mature of the mirrorless formats, m4/3 may be the most interesting at the moment. For the next year or two, at least, micro 4/3 has the better lens lineup. In the long run the advantage of larger sensors might give other makers an advantage. Or maybe not.  Advances in sensor and camera technology certainly seem to have kept m4/3 in the ballpark so far.

I expect micro 4/3 companies to be the first to price bodies more aggressively to win market share. Both manufacturers have introduced high-end lenses, so the classic model of ‘get the new consumer and they’ll upgrade through the system’ begins to make sense. Panasonic may be the more aggressive of the two simply because of the image stabilization situation. A Panasonic camera owner has incentive to buy Panasonic lenses with in-lens stabilization. An Olympus camera owner has less incentive to buy Olympus lenses since all lenses use the in-camera stabilization when shot on Olympus.

The ‘stay with your brand’ lens trend gets a bit more of a boost because these two makers also seem the most aggressive about putting firmware distortion correction (and who knows what else) into their cameras for their own lenses. Those who shoot jpgs already can tell the difference. All other things being equal, using a lens by the same manufacturer as the camera shows less distortion in a lot of cases. I’m OK with this in jpgs, but I’m not enthusiastic about doing it to my raw images. I’ll handle that myself, thank you.

I have some vague hope that this may lead to built-in lens adjustment and correction for things other than distortion and chromatic aberration. I’m not certain that it would be practical to correct coma distortion, for example, in-camera but it would be cool if they could. With or without more correction, though, I expect we’ll continue to see that Olympus lenses give better results on Olympus bodies than on Panasonic bodies. And vice-versa. It’s not a big deal, but all other things being equal, it’s a consideration.

So I think micro 4/3 format remains the more popular of the mirrorless formats for the next year or two. Panasonic and Olympus continue their quiet infighting to keep their own customers ‘brand specific’ while they present a smiling ‘united 4/3 consortium’ public face. This should be more fun than the early seasons of Survivor.


image courtesy Sony USA


The NEX cameras are the smallest and arguably the coolest of the mirrorless group. As I’ve written before, Sony is the 500-pound gorilla of sensor technology, so I expect NEX cameras will continue to lead in, well, let’s call it potential image quality. I’ve given up all hope, though, that they will ever understand the same principal that’s applied to the best quality photography for 170 years now hasn’t changed: It’s mostly about the lenses.

NEX Fanboys, go ahead and line up to slam me now: The NEX lens lineup currently consists of two reasonably good primes, one excellent but overpriced Zeiss-branded prime, and various flotsam and jetsam. Thom Hogan clearly pointed out Sony’s emphasis: they’ve now announced seven NEX cameras to go with their seven NEX lenses. 2011 and 2012 obviously have been the year of the X in the camera world — like everyone agreed to have a date code on their gear. Sony’s letter should be Y. Like, Y can’t you make some lenses?

With those superb cameras and that awesome peaking filter that makes it so easy to manually focus, Sony must be the most popular company in the camera business right now. I’m sure Sony is making decent profits on their NEX cameras, but I suspect other companies are making even more profits selling things that go in front of NEX cameras.

For example, Sony charged me $700 for an excellent NEX-5n camera. Novoflex charged $267 for a simple adapter to let me shoot M-mount lenses on it. Who do you think made the most profit on my NEX purchases? At last count I can buy adapters to shoot 11 different lens mounts on an NEX. (Just search for ‘NEX adapter’ at B&H. You get 6 pages of listings.) We see it here at Lensrentals, too. We rent more adapters for NEX cameras than we do NEX lenses.

The positive outcome of this for NEX shooters is that the third-party lens manufacturers may end up rescuing the NEX lens lineup. The Sigma 30 and 19mm primes are about as good as any native mount lenses available for NEX cameras at an awesome price. (Micro 4/3 shooters greeted those same lenses with “Yawn. . . more good primes. . . yawn”.) Tamron and Tokina apparently are jumping on board, too.

The logical outcome would be that Sony realizes other companies are making good money putting lenses in front of NEX cameras, comes out with a broad lineup of good lenses, and dominates the mirrorless market. But I just used the words ‘logical’ and ‘Sony’ in the same sentence, which is never a good idea. So my guess is that Sony continues to make great cameras and in 2013 Zeiss announces their ZNEX line of lenses. You heard it here first 🙂


courtesy fujifilm products


I put Fuji next just because they seem to be the anti-Sony. They made certain to have a few decent lenses available when they introduced the X Pro 1, and they put out a very specific road map about exactly what is coming and when. (Sony doesn’t have a roadmap; they have a smoke-and-mirror graph.) Instead of tiny, ultra-modern looking cameras they went to great pains to make a larger-than-necessary, classic design body.

For that matter, Fuji is sort of the ‘anti-‘ company in general. Everyone else is making small, interchangeable lens cameras; we’ll make fixed-lens cameras with SLR-sized sensors. Everyone else makes tiny, modern looking interchangeable lens cameras; we’ll purposely make ours big and retro. Everyone else uses Bayer arrays; let’s invent a different array.

I like that ‘let’s do what no one else is doing’ philosophy. They drew the line at letters, though. Everyone else has an X, so Fuji had to have an X, too. They didn’t dare to be too different.

Unfortunately, Fuji took being different a step too far with ‘everyone else’s cameras has good autofocus, let’s make one that can’t very autofocus very well’. I love the X-Pro 1, actually. I like the lenses they brought out with it. But if I need to manually focus, I’d rather shoot Zeiss or Leica lenses on an adapter. I had hoped there would be a firmware upgrade to fix the AF, but it now seems we’ll have to wait for the X-Pro 2 to see if Fuji can do AF.

Don’t get me wrong, the camera is quite usable and makes very nice images. But if you’re the most expensive camera in your class, ‘almost as good as the others’ doesn’t cut it with most people. On the other hand, a core of very serious photographers are snapping up X-Pro 1 systems and are very, very happy with them. I didn’t care for it myself, but seeing the images people are making with it, there’s no question it’s an excellent system.

I find it difficult to make a Fuji prediction. It looks like they are aiming to be the ‘high-end’ mirrorless system. (Maybe it’s no coincidence they designed their camera to be the same size as an M9.) But every Fuji release has left me thinking, “Who would have expected that?” So I expect Fuji will do something I don’t expect them to do.


The Samsung NX system is easy to sum up: Really good camera. Really good lenses. Horrid marketing with a poor grasp of what serious photographers are interested in. If you go to their current website you can see a badly-underexposed-plasced-on-confusing-mirror-surface picture of a bunch of lenses and cameras  (I’m sure my ‘underexposed’ is their ‘cutting edge artsy’ but I’d like to be able to see more than the silhouette off the lenses).

I had to search around the website for quite a while find any actual information about the lenses, though. It turns out they’re tucked in the ‘accessories’ section with camera cases, add-on microphones, and such. It’s a shame, really, because lenses could be the strongest selling point of this system. I suspect marketing is easier to fix than autofocus systems, though, so hopefully they’ll get things together.


The Samsung NX official lineup photo. Copyright Samsung USA


Samsung’s future in mirrorless simply comes down to whether they want to be a player or not. If they do they’ll be a very major player. Samsung is a huge company with incredible resources. They’ve got a very good camera and very good lenses already.  They’ve got tons of sensor patents, including a lot for back-illuminated sensors, that should keep them on the cutting edge for upcoming model releases.

Since no one knows about their mirrorless cameras, just a touch of reasonably good marketing could grab them a significant market share.  (I think ‘good camera marketing’ is probably an oxymoron. But sooner or later someone is bound to do it by accident.) On the other hand, their Chairman might forget they even have a camera division and it all might fade away.

I have this bizarre scenario rattling around in my head like a bad dream: Samsung settles their patent lawsuit with Apple by giving them the camera division. Apple then buys Sony, and in no time at all we have Apple-branded NEX cameras with Apple-branded Samsung lenses in front and a retina display LCD on the back. The resulting camera would make the $1799 Fuji X-Pro look like a bargain. And I’m not sure I’m ready for an all-chrome-and-glass camera with no buttons that you control by waving your hand over the LCD to select the AF point, blinking to change the aperture, and wiggle your nose to snap the picture. Not to mention you’d have to buy the f/2.8 app if you want to shoot wider apertures than f/4.

OK, back to reality. My best guess is Samsung continues to make an excellent camera system that not many people know about. As the mirrorless market grows I could see Samsung NX becoming the ‘cult system’ of mirrorless — not as widely popular as the others but with a deeply committed user base that absolutely love it. I probably would have joined the cult myself except for my desire to shoot some Leica lenses on adapters. (Because their flange-to-sensor distance is a bit longer than most mirrorless cameras, the Samsung system can’t be adapted to Leica-M lenses.)

Pentax / Ricoh

Also available in yellow!!! Images Pentax USA


Speaking of cult systems, this marriage should be interesting. Ricoh has definitely done some very different things with the GXR series cameras. They obviously are interested in the market, or they wouldn’t have bought Pentax. For the moment, at least, we’ll consider the Pentax K-01 as their mirrorless entry, but some would consider the Ricoh GR system a member, too.

The K-01, though, brings a whole new concept to mirrorless. It doesn’t change the lens mount and backfocus distance at all from the standard Pentax K system SLR. This has one obvious disadvantage; the camera will remain thicker than its competitors. The K-01 is 2.3 inches thick and weighs 17 ounces, compared to the NEX-7’s 1.7 inch thickness and 10 ounce weight. The lenses are going to be a bit larger, too.

The obvious advantage, of course, is immediate access to a huge and excellent lens lineup. The K-01 can shoot more lenses with autofocus in native mount than all of the other mirrorless systems combined.

A couple of less obvious advantages to such a system are worth mentioning. While Pentax doesn’t currently offer any full-frame cameras, most of the Pentax lenses are capable of covering a full frame. The newer Pentax cameras are using Sony sensors; so full-frame sensor technology is available to them if they want it. And the flange-to-sensor distance of the K-01 (45.46mm) is similar to that of existing full-frame cameras, meaning sensor microlens arrays probably would not need to be changed significantly. Just a thought.

The second less advantage is perhaps less obvious but even more interesting. It goes back to something I mentioned in Part 1: Mirrorless systems cost a lot less to make. Let’s assume that on-sensor phase detection AF soon becomes as fast and effective as the current separate-sensor phase detection AF. That, and a good electronic viewfinder might eliminate most of the advantages of SLR cameras.

Pentax has a small but very loyal group of users who love their lenses. Pentax hasn’t been growing SLR market share for some time now. Perhaps Pentax is going to migrate their entire system to a mirrorless format. The top end cameras would have built in viewfinders and high-end phase detection AF on their sensors. They would only be slightly smaller than current SLRs but would cost significantly less to produce. Intro level cameras would be smaller, have no viewfinder, and throw in a small built-in flash.

The entire lineup, in other words, would be mirrorless with no SLRs left. Depending on their preferences they could price the cameras lower than competitors SLRs and draw users into that huge lens lineup. Or they could price them similar to competitors SLRs and be more profitable.


Courtesy Nikon USA. My apologies for using this. I just couldn’t help it. Sort of like driving by a car wreck.


I was tempted to say Nidak, because I think Nikon is doing a bit of a Kodak. In part I, I mentioned that it would be awfully tempting, if you depended on cameras for the largest portion of your revenue, to try hard to keep the status quo. Certainly Nikon has done everything possible to make certain that no one who was remotely interested in purchasing an SLR would consider their mirrorless offering as an alternative. Sort of like Kodak developing the first digital cameras and going “get away from that, film is where the money is!!”

Before you go all fanboy on me, I do understand Nikon is in possibly the most difficult position of any of the manufacturers. Camera phones are eating up point-and-shoot sales, so Nikon is getting hurt there. They don’t have an electronics, medical imaging, or even large video imaging division to help them out. SLR cameras and lenses are where their business lives and breathes. On top of that they’ve taken some major production hits between flooding and tsunamis. They deserve all the credit in the world that they’re still thriving.

It’s totally understandable, having withstood all that, that Nikon would be very conservative. They’ve positioned their mirrorless offering to hopefully keep the top-end point-and-shoot customer while making absolutely, positively certain that no one considering one of their SLRs would buy their mirrorless camera instead. It’s probably no coincidence that Nikon currently offers more ‘consumer level’ SLRs than anyone: D3100, D3200, D5100, D7000, and D90s are all available, creating hundreds of ‘which camera do I buy’ threads on various forums.

Nikon clearly markets the J1/V1 as point and shoot upgrades, not as a near-SLR cameras. The formula seemed to work, and they sold well, at least initially. But they certainly aren’t competing with most of the other cameras in this discussion, and I see no signs that Nikon plans to do so in the future. Personally I think the old saying, “If you don’t eat your own lunch, someone else will” applies. But perhaps Nikon has positioned themselves perfectly and this mirrorless camera fad will pass away.


Canon USA

This entire section is really presumptuous of me. Canon has barely announced the EOS-M system, only a few preproduction models have been reviewed, and here I am making predictions. I guess my summary would be that Canon hasn’t buried their head in the sand totally, but they certainly aren’t taking an ultra-aggressive approach either.

I give them kudos for using an APS-C sensor (I had feared it would be a G1X upgrade). In fact, they’ve basically crammed a T4i into a mirrorless camera, which is a good thing; the T4i is a nice camera. EOS-M appears to be a good video platform that may give the Panasonic GH-2/ GH-3 a run for mirrorless video dominance. And while it looks a bit boxy, they’ve made it quite small — less than 10 ounces and 1.27 inches deep. That’s a bit lighter but roughly the same size as an Olympus OM-D.

So Canon has released a mirrorless camera nearly as capable as its entry-level SLR. That’s pretty aggressive in one way. It’s pretty conservative in another — I bet development costs were a lot lower than bringing out a whole new system. I’ll also bet there was a parallel development going on in-house around the G1X sensor, so I doubly applaud the larger sensor choice for their mirrorless release.

Canon is filing on-sensor phase detection patents right and left. That is a win-win situation, too. Canon is aggressive about video and on-sensor phase detection is a video advantage that will certainly make its way to other cameras.

So it seems we’ve got Canon being Canon: not taking a huge risk, investing in research that could have multiple uses, and carefully differentiating this product from that product. There is no option for adding an electronic viewfinder on the EOS-M, for example, so it’s clearly not going to eat the T4i’s lunch completely.

The lack of lens options puzzles me a bit, though. It may be that more lenses are right around the corner. The recent 40mm pancake made me first think that some ‘co development’ might be going on in the lens division, too. But then it occurred to me that maybe ‘not too huge on an adapter’ might be enough to get people to use the new, small primes Canon just released on the EOS-M. Wouldn’t that be convenient? People would already have a few lenses if they decide to move up to standard SLRs.

So maybe all we’ll see is a limited native lens lineup, some push to use smaller SLR lenses on adapters, and an easy move-up when you’re ready. I hope not. I don’t think that formula works, even if the native lenses are fairly small.

My thoughts are that Canon will be Canon: careful, thorough, conservative, perhaps even boring, but excellent. They’ll dip their toes into this market, let others take the big risks, and then catch up with whatever becomes most successful. They’ll snap up some market share for certain, especially among users invested in Canon lenses and among those wanting strong video capabilities. The only thing I see really differentiating them might be a better phase-detection AF system than the other mirrorless cameras. But better autofocus is always a very arguable thing when comparing brands.


In case you can’t tell, I’m enjoying the whole mirrorless camera development thing immensely. In the SLR world a big announcement is a new Nikon or Canon lens, a new camera with a few more megapixels, or results that show the newest camera has an extra 0.3 stops of dynamic range.

Over on the mirrorless side entirely new camera systems are being released. The number of lenses available for a given mount can double in a year’s time. When a new lens comes out for your format you aren’t wondering if it’s much better than the one it replaced because it didn’t replace anything; it’s an entirely new option.

Right now, mirrorless is where the innovators live. They’re taking big risks that the ‘old school’ camera companies can’t.

A lot of you will think I’ve been very negative about Canon and Nikon. I have, but that’s because I like them. I don’t want the true camera companies to find out in a decade their technology is behind the times, their average user is 60 years old, and their lenses are bought more by collectors than active photographers.

Don’t think it could happen? Go research who the dominant camera makers of the 1960s, 1970s, 1980s, etc. were. Canon and Nikon are just the most recent leaders. A decade ago the video world had barely heard of RED. Now Canon, Sony, Panasonic and all the others are trying to catch up to them.

Canon and Nikon currently have the largest R&D departments and better control of their own destiny (their camera divisions are major parts of the company, not afterthoughts). But with most of the innovation coming from the other guys, the inevitable will happen eventually.  Someone is going to eat their lunch for them unless they eat their lunch themselves.

The answer seems rather simple. Set up an autonomous mirrorless division with their own budget and tell them to go dominate the mirrorless market. Tell them to not give a damn about taking market share from the company’s SLRs. Do that and the resources and knowledge that Canon or Nikon have available could create some amazing new cameras.

But if they have to get everything approved in the boardroom, corporate infighting and turf protection will make certain they develop a camera that offends no other section of the company. And that will mean they’ve created cameras that don’t worry any other company. They’ll still probably dominate the SLR market for the foreseeable future. But right now, at least, that doesn’t seem to be where the growth is.



Roger Cicala

August 2012





Author: Roger Cicala

I’m Roger and I am the founder of Hailed as one of the optic nerds here, I enjoy shooting collimated light through 30X microscope objectives in my spare time. When I do take real pictures I like using something different: a Medium format, or Pentax K1, or a Sony RX1R.

Posted in Roger's Corner
  • Dirk Uys

    Roger Cicala’s last NEX paragraph: I don’t expect Sony to let Zeiss or anybody else make their lenses for them. Not Sony. They will want to own the ecosystem. But they may have a reason for not putting effort into lenses now.

    Tom Cavanaugh’s comment: For now. For now. I was also wrong in not seeing how useful a large flange back distance is. But I suspect that Sony got bigger than me because they’re brighter than me. Methinks the result is an interim kludge and the problem will go away in NEX II (NEXT?)

    Methinks NEX II will have a concave sensor. Same bodies, new lenses, much less CA, much less distortion. And hard corners from simpler lenses. Only the 500-pound gorilla could do that. It would do to photography what the pneumatic tyre did to land transport.

    I don’t know when. Somebody said within five years.

    Is it asking too much that they also come up with a non-Bayer microlens array that obviates AA filters?

  • Tom Cavanaugh

    The nice thing about the Sony NEX series is the thin body.

    The bad thing about the Sony NEX sseries is the thin body.

    Having the short flange focal distance with the (relatively) large sensor means the lens has to bend light more, so it can project an image onto the entire sensor. This results in greater chromatic aberration and greater field curvature (softer corners.) As a result, any E-Mount lens performs less well and costs more than an equivalent A-Mount lens used on Sony’s Alpha series of cameras. This is a problem that can not go away.

    The only solution is to go with lenses from other mounts, but that means having an autofocus, autoexposure camera, and only using manual focus and manual exposure. The only exceptions are Sony’s own A-Mount to E-Mount adapters which make a NEX camera bigger, heavier, and more expensive than an equivalent Alpha camera.

    I think Sony Corporate decided what the NEX line would be, and Sony Digital Imaging came through the best way it could, but the result is a kludge.

  • Roger Cicala

    Hi Andy,

    I agree with all of what you said, but I was trying to point out that there is some controversy about what might be applied in raw data, not just in-camera jpg. As an OM-D shooter using the 12-35 as my go to lens, I plan on doing a bit of correction in raw. My concern is that all distortion and edge corrections come at a reduction in resolution. I want to decide when that happens in my work. I know a lot of people don’t care, though.


  • Andy Moore

    Hello Roger,

    I’m sure a lot of people have mentioned this “factual correction” on any number of other blog postings where you have mentioned Micro 4/3rd, but I’ll mention it again. The m43 standard includes provision for geometry distortion characteristics to be stored in the lens firmware and to be communicated back to the body. So any lens with electrical connections (even those Sigma primes) on any body will correct for barrel/pincushion.

    There is additional provision for individual manufacturers to store additional lens data in the lens firmware, and have the bodies use that data. So far it seems Olympus hasn’t used this part of the standard. Panasonic have – they store Lateral CA data in their lenses, and their bodies can use this to do in body Lateral CA correction for JPEGs and tag this info in the metadata of RAW files.

    This seems to have lead Panasonic to compromise on how much Lateral CA their lenses create, since it won’t affect those lenses used on Panasonic bodies. This is causing one hell of a stink on the DPReview forums, mainly from EM-5 owners that have bought the 12-35mm X lens and find it fringes all over the place…

  • Esa Tuunanen

    Sony definitely has lots of leading edge tech like sensors, modern non-interlaced video modes and very high resolution EVF.
    But NEX has its own problems. Sure they got the smallest bodies but lenses are anything else than balanced with bodies and NEX-7’s corner crammed viewfinder isn’t very comfortable with its small eye relief/image size… while A77 has lot better viewfinder using same OLED display component.
    Also body/mount might be simply too cramped to include sensor shift based stabilization. That prevents possibility for feature similar to Astro tracing of Pentax utilising sensor rotation which can’t be matched by lens based stabilization.
    With P&S size NEXes built around on screen menus instead of direct controls for activating settings it feels that they’re saving higher end for Alpha SLTs.

    Though if there’s someone protecting analog mechanical era legacy that’s Canon and Nikon. Sure hope others will eat Canikon’s lunch, dinner and supper, that would serve them right for pulling parking brakes of technology for decade to lock people into SLR legacy systems dominated by them.
    Nikon 1 is obvious case and overhyped Canon’s EOS-M is 650D’s electronics coupled to controls and ergonomics of IXUS serie, more precisely to lack of them which doesn’t need any innovation. Which shouldn’t be surprise considering Canon even lacked bridge/prosumer digicams before time of cheap DSLRs.

    And Fuji’s claim to fame and hype around it is as fresh thinking as corpse of certain Joseph Vissarionovich.
    I consider camera as utilitarian tool and X-Pro 1’s design is made around design limitations of analog mechanical tech and piece of plank form factor.
    Why on Earth should camera’s design be fixated to limitations of single old tech?
    That’s same as making cars according to blueprints of horse wagons!

    Same really for Olympus putting excellent tech in sensor and stabilization into retro OM-D… Those positions of dials didn’t come from studies of best ergonomics but from getting around mechanical limitations of using film.
    Also almost completely centered retro position for lens mount/viewfinder just makes nose hit to body/display.
    Moving lens mount/viewfinder to body’s left edge like in those 2003-2004 era bridge/prosumers which were more innovative than any current mirrorless body would solve that and keep viewfinder lined with lens.

    And Samsung… Their management seems fixated to photography meaning ability to send snapshots directly to Facebook.
    Maybe they’ll make also that Android camera with all the problems of computers and smartphones so that you can play Angry Birds during sending of those snapshots.

  • Roger Cicala


    Great minds think alike 🙂 We’ve actually been looking into some subscription based models for those who just love trying everything.

    I got shot down on my “Lens of the month club” idea, though.


  • After reading this, a thought struck me: This should be a really good time to be running a lens-rental company!

    I ditched my entire Nikon DSLR system after realizing that for my style of photography, mobility trumps a huge accessory lineup. Currently I’ve got a small Micro 4/3 system and I’m happy. But that’s now. The various mirrorless systems roughly are at parity right now in terms of overall picture quality — one a bit ahead here, another a bit ahead there — but as you alluded, within a year or two someone may open up a big lead on the pack. The problem is that we don’t know who!

    This makes me very reluctant to invest much more in M4/3 lenses than I already have… and every other mirrorless user is basically in the same leaky boat. If the system you pick today becomes an also-ran, your lenses are going to be worth zilch on eBay. I don’t mind spending a thousand bucks every couple of years for a significantly better-performing camera… but if I have to write off several thousand dollars worth of lenses at the same time, that’s painful enough to make me think twice.

    So, the next couple of years could be very good for Roger’s business! Maybe you should make it like Adobe Creative Cloud — we could just pay a monthly fee and you’d send us whatever lens we need at the moment for whatever camera we’re using right now…

  • David Bateman

    Roger I think your wrong about what you wrote regarding Pentax. Yes the camera is thick. But you have an anti-Sony NEX camera system in a way if they use it the I think they will use it. The NEX Camera system has very thin cameras with 19mm flange distance. The result is the lenses are huge, some really huge. An example is the 30mm Macro the A mount macro on adapter is the same size as the NEX Macro. You need the extra depth to make the lens work. So generally all sony lenses will be big.

    Now look a Pentax and think Lieca of old. The Pentax body is big, it lets you mount the old lenses with the old lens formulas. But you can design new lenses that are either folding into the body for walking around (like many Lieca folding lenses) or you can design them to be in the body for when shooting as well. The Rear glass element can be designed really close to the sensor (made for APSC) and the overall lens is very small. Also your telephoto lenses that need to be further from the sensor to work, can now fold into the camera for pocketablity when walking around. I really hope Pentax (whom are excellent lens makers) will capitalize on this and make a very small telephoto system.

    The other advantage that Pentax has you do comment on. They can toss a full 135 sensor in the camera and please people that want it. Then fold wide angle say 35mm lens into the body and now you have an amazing system that might shift sales away from the NEX full frame system rumored. Also it gives you clearance for Image stabilization that the Sony can’t offer.

    One last thing, your thoughts and articles are so good. I do check this sight almost daily to see what up. Thank you for your thoughts.

  • Guillaume

    Thank you Roger for reply.

    I re-read “Autofocus Reality Part 3A: Canon Lenses”. As for the standard Canon 50mm f1.4 prime, you measured : LV SD = 9 and AF SD = 34. Ouch !!!

    Well, I may be mistaken but I presume that OM-D EM-5 + Panasonic Leica 25mm f1.4 is better than 5D MArk III + Canon 50mm f1.4. I think this lens is in dire need of replacement…

    For street photography for instance, what’s the point having a 5D Mark III if you misfocused so many shots. Unfortunately, I don’t do solely street photography otherwise I guess the OM-D EM-5 would be the right choice.

    Can’t wait to see LV vs AF standard deviation of Nikon’s 50mm lenses 🙂 That would just be a good reason for me to pick one camera over the other…

  • Roger Cicala


    I wish I knew the answer to that, but I’m still doing the AF research right now. That’s a huge project, and I’m not sure even when it’s done if I’ll know the answer. But I’m going to try to figure that out.


  • Guillaume

    Hello Roger !

    Many thanks for your recent articles about focus and mirroless systems. If I am not mistaken, you’re using an OM-D EM-5. I’d like to ask you a question please related to focus.

    Though the followings are different cameras and considerings non-tracking subjects, would you say that the contrast detect AF accuracy through EVF of a Panasonic Leica 25mm f1.4 mounted on a OM-D EM-5 is better than the phase detect AF accuracy through OVF of 50mm f1.4 or 80mm f1.8 mounted on a D800 or 5D Mark III ?

    Latest 5D Mark III’s PDAF is on par with its liveview mode’s CDAF as you observed. On the other hand, when reading reviews of both D800 and 5D Mark III, their CDAF in liveview mode are not stellar. So being on par with them is not being on par with CDAF of OM-D EM-5 I guess… What do you think ?

    Many thanks !


  • My Olympus OM-D just arrived, and the next day I found your “part 1” article, in which you say my 17mm is the worst of the wide-angles …. Well, something had to go against me … I’m otherwise happy. I’ve been out with the 17, have some photos on my blog ( ), tomorrow i’ll go try the 12-50.

    I wanted to support Olympus for doing something good, and hope it helps them survive their recent f-ups.

    I’m disappointed with the software Olympus provides, it can’t make a decent panorama … feeding the same three pics into Photoshop produces a find image, but I have the option of doing it with RAW instead of jpg. I’ve tried the 3D option, but it didn’t do much of anything useful … .it seems to want me to align objects back in their original position. And I’d much rather do it with RAW than with jpeg? have I mentioned RAW is better than jpeg? And DNG is better that ORF.

    Some things are good … for my first foray, I shot in P, from the waist, without aiming or focusing, and wound up with some decent images. Of course they needed some post processing, but it’s a good sign that they survived the processing with decent results.

    As for what you say about Olympus vs Panasonic, in terms of Image Stabilization in the body or in the lens, seems to me a Panasonic body has to be notably better than an Olympus body, since the P will IS with a few lenses, but the O will IS with all lenses.

  • Moo

    Pentax has another mirrorless system, which isn’t mentioned, the Q. Basically the complete opposite of the K-01, with a small sensor and small lens range, but what was (still is?) the smallest MILC body.

  • Roger Cicala


    I liked the G1X sensor very well. I had feared Canon would use that in their mirrorless as a way to keep SLRs separate: you want a full APS=C sensor, you have to buy a full SLR.


  • I have to somewhat disagree on Canon (to an extent). Why did you not like the G1x sensor? It seems to do quite well. And it would fit better as a conservative Canon. And make for more compact lenses.

    It could easily replace their Rebel line while leaving high end APS-C/35mm DSLRs room to breathe. It’d certainly give m43 a run for their money (hopefully not completely wipe them… love my m43 kit). And it buys them time to make a mirrorless 35mm that performs like it’s mirrored counterpart.

  • Set up an autonomous mirrorless division with their own budget and tell them to go dominate the mirrorless market. Tell them to not give a damn about taking market share from the company’s SLRs. Do that and the resources and knowledge that Canon or Nikon have available could create some amazing new cameras.

    You must’ve read Clay Christensen’s The Innovator’s Dilemma—or its ideas have become so pervasive that they’re now part of the intellectual landscape. If you’ve already read it, then ignore this comment; if not, give it a shot, and you’ll see the basic problems that Canon and Nikon have, described in a universal way.

  • I really like the Nikon mirrorless. Isn’t this what up-graders from P&S should want? I hear complaints about the difficulty or fear involved in changing lenses, and the cost of buying lenses. The Nikon offerings are still small, bigger zoom potential, and the image quality is good enough. Nikon’s poor sales makes me wonder if mirrorless is more for those moving down from DSLRs than for those upgrading up to them. Is there really a consumer market for mirrorless?

  • Jochen Demuth

    “And I’m not sure I’m ready for an all-chrome-and-glass camera with no buttons that you control by waving your hand over the LCD to select the AF point, blinking to change the aperture, and wiggle your nose to snap the picture. Not to mention you’d have to buy the f/2.8 app if you want to shoot wider apertures than f/4.”

    I almost spilled my coffee. Roger, you’re so right – this time even outside the camera circus.

  • I agree with William Tracy.
    Thank you Roger. Your are innovative in being journalist. In the same way as Small companies are in front of Canon and Nikon.
    May the curiosity be with you!

  • Roger: The biggest issue of getting phase detection onto the sensor is “how do you do PDAF with rows or crosses of small (pixel sized) PDAF sensors”. And how do you do on-sensor PDAF without messing up the image too much. That’s what all the patents in this area (from Nikon, Canon and Sony (IIRC … or am I making that up?) are about.

    The other issue is impacts on image quality through interpolation over the on-chip PDAF sensors as they make little gaps in the Bayer CFA that have to be “healed” to “fake up” Bayer-like RAW otherwise each RAW convertor software company has to do this (and the companies can avoid modifying their Bayer raw to JPEG software). So RAW is very slight pre-cooked. Nikon came up with a clever scheme for this. I think this may turn out to be the bigger problem in getting on-sensor PDAF into DSLRs. This is also a reason why me might see “flappy mirrors” on the high end cameras for some time to come.

    The size and complexity of the curent PDAF arrays (along with their mechanical alignment and calibration) ust goes away when you are making them on the sensor. They’re just scatter across the sensor as you need them. Sensor read out speed is not a limit.

    The arrays may even get less complex if the cameras use a hybrid of PDAF and CDAF to determine focus. PDAF is useful because it tells you both the magnitude and direction you need to focus in so it’s good for initial focus. CDAF is good for fine focus when you are nearly there (or when the PDAF gives up. Plus CDAF can do face-detect and all the other newer focusing modes (that despite some snearing are very useful focusing modes … as you may have found on your OM-D). I suspect most on-sensor PDAF cameras will use a hybrid approach.

    Once the sensor issues are fixed I think hybrid PDAF/CDAF will follow very quickly. Nikon can already do it on the Nikon 1 (to some extent). Canon seems to be doing less well with the EOS-M. Sony is rumored to perhaps have it on the NEX 6 (we’ll see). But I think the timeframe is 2 to 5 year (max).

  • Greg

    Another excellent write up!

    Regarding the lack of Canon EF-M lens options, I would not expect a large, extensive selection of native EF-M lenses but something more akin to EF-S. The marketing seems to tout the EOS membership of the camera, and by extension access to the entire EF lens system.

    Instead, I would expect a selection of EF-M lenses featuring shorter focal lengths where the shorter flange to sensor distance and/or smaller physical size would be advantageous. Zooms in the 70-200 range or longer will remain EF only as will short telephoto and longer primes.

    I do note that the 2 EF-M lenses announced so far have metal mounts, unlike the kit EF-S, and the 18-55 has superior specs, MTF, features, and higher price compared to the EF-S. This leads me to think that Canon is aiming their mirror-less system at least as high as the T4i, not below it. I also think we could see a few nice EF-M primes (in contrast to the one EF-S macro) in addition to the zooms.

    When Canon fully gets AI servo type phase detection on the sensor, it won’t be conservative.

    Greg H.

  • Roger Cicala

    William guilty on all counts except running a rental house. I’m still the ‘founder’ of Lensrentals but no longer the owner and I don’t do any day-to-day management of the company. (And as you noted, the person who usually proofs is out right now. I’ll try to fix it up myself over the weekend. ) If phase detection moves onto the sensor without loss of accuracy (I think that’s some years away) I wouldn’t be surprised if a lot of manufactures have mirrorless offerings in their SLR mount. The sheer size and complexity of the best phase detection arrays, though, makes me think it’s a large step before we get AI servo type phase detection on the sensor. But it would be cool if I’m really wrong about that.


  • Roger Cicala

    Mostly because I just don’t have a handle on them. I’ve talked about their products a lot in other blogs, but the big thing with Olympus is the scandal / business aspect of the company and I don’t know anything about that. Will they sell off the camera division? That’s probably the biggest question right now and I have no clue about that kind of thing. I did read that they just sold their mobile phone unit, so maybe that makes it less likely they’ll sell of the camera division.

    I don’t think it contributes, but for full disclosure I shoot an OM-D mostly these days. I try to not go Fanboy about what I actually use, so maybe I’ve leaned back too far the other way and minimize Oly somewhat.


  • Ben

    Out of interest why do you keep outskirting around Olympus? Both Part I and Part II you make a few references, but you don’t actually go into any meat.

  • Thanks, as always! And three responses:

    1. This is such an embarrassment. Because there is absolutely no real journalism covering the photography world, an old guy running a rental house is now one of the only real sources of decent, objective information in this realm.

    2. I know this is meant to address the “mirrorless” phenom, but I’d like to see a broader reach. I believe technology is moving to a point where real photographers don’t have flappy mirrors between them and their composition/images. I’d like to see Roger’s take on this.

    3. If you’d like to send your copy to me for basic proofing before posting, please do. This one has some glaring errors.

    Thanks again!

Follow on Feedly