Mirrorless Camera News and Commentary

News and commentary about the mirrorless camera world (latest on top). Hover or tap on News/Views in the menu bar above to see the full list of recent articles as well as folders containing all older ones dating back to 2011.

Another Problem for the Duopoly

Korean optics manufacturer Samyang and German optics maker Zeiss are starting to show a new problem for the "closed shops." 

What do I mean by that? Well, neither Canon nor Nikon have ever really shared their mount communications and complete specifications with others. Canon's EF mount has been reverse engineered with some reliability, but Nikon's F mount has proven much more difficult to lock down. For a long time in the DSLR era, having those closed mounts meant that Canon and Nikon not only dominated the camera body volume, but also the autofocus lens volume as well. 

Thing is, Samyang, Zeiss, and many others used to be pretty much just manual focus lens makers partly because of that. They didn't have access to the secret juice that made the Canon EF and Nikon F lenses focus so well and reliably (let alone provide the EXIF data and manage advanced things like AF Fine Tune), and thus those third-party lens companies didn't ever enter into the competitive AF lens arena for the two major DSLR makers. 

Yes, some companies managed to play the reverse engineer game successfully. In particular, Sigma, Tamron, and Tokina. Tokina had several former Nikon optics engineers in its lens group, though, while Tamron has apparently had a few OEM lens deals with various camera makers, including Nikon. It's not difficult to guess how the genie got out of the bottle for those two. Meanwhile, Sigma has been the one doing the most exhaustive reverse engineering, and with the Nikon bodies in particular you can see how very small communication changes in the mount often caught them with an incompatibility they had to fix. 

But all that is changing in mirrorless. 

The m4/3 partners were the first to allow others to openly license mount information. Then Sony did the same thing with the E mount. Fujifilm seems to be a bit more restrictive, but has apparently shared information with Zeiss. 

The result is that Samyang—formerly a company that only made manual focus lenses—now has five autofocus lenses for the FE mount, with another about to be announced. Zeiss—formerly another company that only made manual focus lenses and primarily for the Canikon DSLR duopoly—now has four autofocus lenses for the FE mount (and Fujifilm mount) with another about to be announced.

Whether Canon and Nikon like it or not, the mirrorless push has finally put a dent in their dominance over autofocus lenses for ILC cameras, and this problem is only going to get worse for them. 

The irony is that Nikon really started with lenses, not camera bodies. Indeed, Nikon supplied glass for early Canon cameras. Canon quickly got into the lens business themselves, and eventually the two dominated ILC cameras and lenses, with as much as a 75% market share between them. At this point, Canon has produced over 100m EF-mount lenses, and Nikon nearly as many F-mount lenses. That's a huge installed base of lenses that tends to keep people "in the mount."

But again, all that is changing now. 

It started with mount adapters (e.g. using a Canon EF lens on a Sony FE body), which helped ease the financial pain for some in transitioning to mirrorless. But now we're seeing large lens sets develop in the mirrorless world, and they're not exclusively from one manufacturer for the m4/3 and FE mounts. 

I'll be curious to see how long the duopoly lasts, and whether they can break through the competition that mirrorless has enabled. Lenses, like everything in the camera world, have been experiencing a drop in volume (CIPA shipment numbers):

  • 2013: 26.7m units
  • 2014: 22.9m
  • 2015: 21.7m
  • 2016: 19.2m
  • 2017: 19.2m

That decline has not been as steep as the camera declines, but that's partly because it's also been fueled by mount switching as some people have moved from DSLR to mirrorless. Plus those pesky third-party lens makers now are showing some leverage in the market as they produce autofocus and more sophisticated lenses for the mirrorless market, which is not currently declining in volume.

Advantages and disadvantages accrue to the last players to transition in a market switch. The advantages tend to revolve around being able to see what works and what doesn't. The disadvantages tend to resolve around the fact that the early movers can establish footholds that are difficult to displace.

Lenses are one of those footholds in mirrorless. As I've pointed out before, we're piling up quite a few mirrorless lens options now. Not so much at the telephoto end or more exotic designs such as tilt-shift lenses, but it likely won't be long before those things tend to completely fill in, too. 

Yet today we have Canon sitting on just seven EOS M consumer lenses, most of which are targeted very low, and Nikon sitting on nine CX lenses, none of which are likely to be useful to them in the future as they abandon the Nikon 1 lineup that uses them. 

If I were an executive at Canon or Nikon, I'd be worried silly about lens availability as I try to move into the mirrorless ILC market. I'd be worried that proprietary won't cut it against the more open systems that are now well established. In particular, I'd be worried if I were a Nikon executive that the very foundation of the company—optical engineering—doesn't seem to be really coming to play in a market they need to do well in (mirrorless). 

We live in interesting times. The next year or two will tell us a lot about what we need to know about mirrorless lenses. Let's hope that Canon and Nikon actually have a story to tell there.

The Nikon Mirrorless Wish List

I’ve been communicating with a lot of Nikon DSLR (and some former) users lately about what Nikon might do with their eventual re-entry into the mirrorless camera market. (The Nikon 1 lineup last had any changes three years ago, and recently we’ve been seeing a lot of  “discontinued” or “not available” listings online; even NikonUSA’s store lists every Nikon 1 camera as “temporarily unavailable".)

As I’ve noted in previous articles, based upon sources in Japan I’m pretty sure that Nikon prototyped pretty much every mirrorless possibility we can think of. I’ve received credible reports of all kinds of designs, mules, prototypes, and mockups, varying with everything from sensor size to lens mount to sensor type to design type to feature set. Clearly Nikon has been trying to make sure it considered all possibilities before making any final decision. Returning to mirrorless is an important milestone for them and their future in cameras, and I think it’s clear they want to try to get it “right."

That said, I’ve become more and more curious about what the Nikon DSLR customer thinks Nikon should be doing in their return to mirrorless. Obviously Nikon will do something, but what do the current Nikon shooters expect them to do? The current group of D7xxx and up Nikon DSLR shooters is critically important to hold onto as a Nikon customer moving forward. So I’ve been actively engaging that Nikon DSLR crowd to discuss mirrorless with me.

It seems we have some points of contention and some points of agreement. 

One clear point of contention seems to be whether or not Nikon should make a DX entry mirrorless system. This is a tricky bit, as most of the people I’ve talked to that are making comments about this already own cameras (well) above the entry point, thus they have no personal interest in such lower-end DX cameras. I’m just going to dismiss that contention by saying this: Nikon needs a DX-M system that can at least hold its own against the current Canon EOS M system. I just don’t see that they have a lot of choice in that, as the DX DSLRs are where ILC volume is fading for Nikon and causing market share loss, and Nikon needs entry points that are not on the expensive side if they really want to stop the epic contraction they’re experiencing. 

So, while most Nikon DSLRs users are conflicted about DX-M, I’m not: Nikon needs to do it, do it soon, do it right, and have a compelling story about it when they do. We’re talking here about cameras that will live in the D3400 space, maybe up to the D5600 space. Cameras that compete with products such as the recent Fujifilm X-T100 and X-T20, the EOS M’s, and whatever Sony decides to do with their future APS-C offerings (at least below the A6500, which is a high-end APS-C in my opinion). Thing is, you just can’t ignore the under US$1000 space, nor the near-pocketable camera space. You also can’t make a full frame camera that meets either of those two criteria, so one choice has to be APS-C, thus what I call DX-M.

I’ll return to that thought in a bit. 

The other major point of contention among Nikon DSLR owners that I'm finding concerns any full frame mirrorless camera Nikon might produce. Nikon DSLR users are completely split on one thing: it’s almost a 50/50 “make a new mount” versus “use the existing mount” split as far as I can tell.

The “make a new mount” side seems to have bought the line that mirrorless is always smaller and lighter because of the mirror removal, so you can almost effectively proxy this contentious point by asking what the size of a full frame Nikon mirrorless should be. If the answer is “needs the DSLR ergonomics with perhaps with some weight trimming” that person will almost always say “use the existing mount.” If the answer is “has to be small and light to match the Sony A7"—which isn’t exactly that small and light anymore folks (the A7 grew from 474g to 650g in three generations)—that person will almost always say “create a new mount.”

Both sides on the mount argument do agree on one thing, though: new mount or old mount, if there’s a drawback to using a legacy Nikkor on this future full frame Nikon mirrorless camera, Nikon has made a mistake. Pretty much any drawback will trigger disappointment it seems. That’s going to be a very high bar for Nikon to get over. I have no doubt that a recent F-mount type E AF-P lens will do just fine on any mirrorless camera Nikon makes. However, each step back in the mount history reveals something that will be harder to accomplish. To wit:

  • Do all the AF-S lens motors work well enough, particularly in Single Servo mode, where we probably want a contrast detect final step?
  • Will the D lenses that require a screw drive motor in the camera still work, and like the above, will they work well enough?
  • Will aperture rings in older lenses still work to control aperture?
  • Will AI indexing actually be supported?

Nikon might be able to finesse things by having a perfect F-mount adapter for a new mount Nikon mirrorless. Might. Everyone I ask about this tends to be highly skeptical about how far back Nikon can support the legacy lenses with a new mirrorless system. It also seems that each person has their own “breaking point” in this regard. That’s usually expressed as “If my X lens no longer works I’ll be looking at other options in the future, such as the Sony cameras.” (Where X defines a lens in one of the four tiers I put in bullets just above.)

One other point of contention that I keep encountering: whether a new Nikon full frame mirrorless camera should be entry or pro. Is it a 24mp competitor to the A7m3, or is it to be a 45mp+ replacement for the still out-of-stock D850?

I’ve been consistently surprised by the ones that only want the pro side of things and are eager to replace their D850. I actually don’t get it. I’ve written and still believe that the D850 is a better all around camera than the Sony A7Rm3; I fail to see what Nikon could do to make a mirrorless D850 today that tops the DSLR D850. 

There’s a reason why the entry versus pro question comes up over and over in my discussions. And it’s a third group that brings that into better focus:

At least a few Nikon DSLR users take a different approach to defining a new mirrorless product, totally ignoring sensor specs or entry versus pro. I also need to point out that Goto-san’s remarks in China last year echo this: make a full frame camera that supplements the DSLRs, not replaces them. 

To understand that, you have to consider the Nikon Df. While it’s a DSLR, the Df didn’t really fit into Nikon’s DSLR lineup in any easily definable way. Nikon themselves had a difficult time marketing it, eventually picking the “it’s different, and more like our old film cameras” story. The Df didn’t sell all that well—somewhere in the 50-100k units lifetime—partly because it was an awkward marketing ploy (no video capabilities, really a cheap D610 body underneath, no true support for the legacy manual focus lenses that Nikon suggested were great to use with the camera, and more). 

That said, there were customers that bought the marketing line that the Df was something different than the traditional DSLR. Indeed, I think all of us Nikon DSLR owners wanted the Df to be even more distinguished from a traditional DSLR. Where was the expanded abilities to deal with the old manual focus lenses, for instance? Focus peaking, anyone?

So what would make a full frame Nikon mirrorless camera distinguished from the DSLR brethren? 

As I note in an article over on dslrbodies this week, it wouldn’t be telephoto lenses. Once you get to telephoto focal length usage, the DSLRs rock over the mirrorless cameras, and probably will continue to do so for some time. Moreover, the DSLRs have tremendous lens selection, though getting better and better optical designs has also made a lot of the lenses grow in size. 

You’re probably seeing where I’m headed.

Many of the DSLR owners I talked to want a supplemental camera system, and they specifically want that supplemental system to be small and light. Indeed, that’s why they’ve been sampling mirrorless cameras from other makers. You might remember that this is how I got into m4/3 back in 2009: supplement my telephoto-laden DSLRs on safari with something small to shoot mid-range.

Over a decade ago I suggested that Nikon needed to make an FM3D. Basically the small and simple FM3a film SLR size and style body, but brought fully into the digital era. Along with this, Nikon would also need to make a modest set of small lenses to complete the “small and light” package: 24mm, 35mm, 50mm, and 85mm primes that are near pancake in size, maybe an 18-35mm and 24-85mm zoom that were as compact as possible. The Df felt like a half-hearted attempt to get to that. What a clear subset of the Nikon DSLR crowd wants is a full-on attempt to get to that FM3D idea. 

The interesting thing is that there seems to be full agreement over one thing among pretty much all the Nikon DSLR owners I’ve talked to about future mirrorless: smaller and lighter is a requirement. That should seem obvious, because if that isn’t achieved, then what’s wrong with just continuing to refine the DSLR? 

I was shooting yesterday testing out a Nikkor 19mm PC-E lens with the Nikon D850 DSLR. Since that lens is manual focus, I was using Live View and focus peaking (a feature that the D850 has and older D8xx DSLRs don’t). Another Nikon shooter (D810) noticed me and came over to see what I was doing, and immediately took to the focus peaking on Live View. Yes, all you mirrorless readers, the DSLRs can continue to make strides in Live View shooting. Which is the thing that most of you find useful with mirrorless cameras (though via an EVF; but even that isn’t necessarily ruled out for DSLRs, as one can easily imagine a hybrid DSLR viewfinder, and several have been patented).

Nikon was slammed on the Web with negative publicity when they withdrew the DL cameras and never produced them (reminder: 1” compact cameras styled with DSLR-type controls). At present, Nikon has no true small and light camera option available (remember, the Nikon 1 isn’t actually available anywhere that I can find, including directly from NikonUSA). What Nikon DSLR shooters are telling me is that they want such a camera. Indeed, that they expect any mirrorless option Nikon comes out with will be clearly smaller and lighter than the DSLRs. 

So where does that leave us (and Nikon)? 

Nikon’s in a tricky spot, and they’re terrible at marketing (comparatively to the other camera makers, none of which is all that spot-on in marketing themselves). Moreover, Nikon has trimmed advertising so much to the bone it is now completely missing in action. That means that whatever mirrorless camera they launch pretty much has to go viral on its own based upon its features, abilities, and design. And that means they need the Nikon faithful to gush over it.

Here’s what I think Nikon should do:

  • DX-M — EVF and EVF-less models that target the X-A5/X-T100 and M5/M6/M100 and which will eventually sit in the D3400/D5600 price points (it can be a bit above that on launch if the camera is well done). Such a product needs a wide, mid-range, and telephoto zoom kit lens set that’s small, a superzoom, and yes, two or three compact and faster aperture primes. It would help if these had serious video creds. Goal: small and compact entry cameras. 
  • FX-M — an EVF model that’s supplemental to the DSLR line initially. It needs to sit where the D610/D750 live in terms of price point. Such a product needs a wide angle zoom (that 18-50mm DL idea sure sounds great here if it can be made compact enough). And it needs some small primes and an F-mount adapter for everything else. If this camera is really bringing video features along at state-of-the-art levels, those small primes also need to be video-friendly (which is tricky, since they’d almost certainly be fly-by-wire focus). Goal: the day-DSLR-replacment for travel, family, and fooling around. Let the DSLRs still handle the heavy lifting for the serious shooter. 

But this leaves out something. I see people—and camera companies—get far too carried away with basic spec lists and specific product features. 

Here’s what I personally will be looking for in anything Nikon does in mirrorless: does it have Nikon’s DNA

Nikon themselves don’t always seem to understand what that is, so let me explain several things that make Nikons Nikons:

  • Best in class still image quality. They’ve missed the mark a few times along the way, but it’s been clear from day one in the DSLR era that Nikon has targeted this with every product. I’d argue that today the D850 is the best-in-class high pixel count full frame camera, the D5 at or near the best-in class high ISO camera, the D500/D7500 the at or near the best-in class APS-C (DX) cameras. Heck, the four year old D750 actually does a very good job of keeping close to the just released A7m3, despite all of Sony’s purported sensor tweaking. 
  • Best ergonomics. Guigiaro’s SLR/DSLR design language for Nikon was dead on. As I’ve explained many times, it’s all about right hand and finger position versus major controls. Nikon’s deviated from this over many decades very rarely, and at their own expense when they did (witness the Nikon 1). Finger over the shutter release, always. Hand position stable while making control changes, always. Fingers find key controls easily, always. Eye at the viewfinder when making changes, as much as possible. Coupled with a menu system—now with excellent touch control—that is pretty well organized and easy to navigate, what you learn on a lower end Nikon DSLR generally translates very well to the higher end cameras, so you’re not relearning things. (There’s an exception, but it would be easily fixed if Nikon’s engineers would ever talk to users.)
  • State of the art technology. Nikon equivocates sometimes on this one (I’m looking at you Nikon Wi-Fi), but Nikon usually opts for a new and yet-to-be proven technology when it makes sense. For example, XQD. The ultimate problem as we’ve moved forward in digital cameras almost always resolves around bandwidth: moving data around. If you go back and look at who was first at multi-channel readouts and pretty much any other bandwidth-related advance, you’ll find Nikon's hands are dirty. They understand moving many bits of data around while not compromising anything while doing that. 
  • Intelligence. People forget that Nikon was pioneering in so many near-AI things we take for granted today in our cameras. Nikon’s early matrix metering adding intelligence from multiple sources along the way and is still rarely matched by others. Focusing algorithms now do things we never thought possible on a DSLR. Things that seemed like science fiction when they were first proposed tend to be commonplace now in Nikon products (e.g. the recent Automatic Picture Controls that respond to the scene rather than you having to pick which Picture Control to use). 
  • Legacy support. Yes, we’ve had hiccups with this along the way, too, but it is pretty amazing that you can still pull old Nikon gear out of the closet and use it with current cameras. I mean really old gear. My 40-year old 58mm f/1.2 NOCT lens works just fine on my D850. My 30-year old remote cable works just fine, too. It’s really difficult to say that about any other camera maker. Of course, mirrorless is another big changeover like DSLRs were, and it’s a big task for Nikon to keep legacy support going when you encounter such big hurdles. 

I’m sure I could come up with other things I could add to that list, but those are the five seminal strings of DNA I see in pretty much all Nikon gear, and which I as a Nikon user have appreciated over all these years (I can now say that I’m at 54 years of using Nikon gear even though I do supplement that with other companies' products these days). 

I’d argue that those five strings of DNA have to be in anything Nikon does in their re-entry into mirrorless cameras for Nikon to succeed. Regardless whether it is DX, FX, consumer, pro, 24mp, 45mp+, new mount, old mount, or any of the other ideas that Nikon DSLR shooters shared with me. 

Ball is in your court, Nikon. 

The Fujifilm X-A5 Gets a Big Brother

Fujifilm today announced the X-T100. This new camera is mostly the X-A5 internals coupled with an EVF and the three-way tilting LCD from the more DSLR-like Fujifilm bodies.

bythom fujifilm xt100

Let's get one thing out of the way fast: I love the X-A5's images. The 24mp Bayer sensor in the X-A5—and now in the X-T100—produces really nice, clean, detailed images that work buttery smooth in most raw converters. What it doesn't do is real 4K video (only 15 fps), or really fast image handling (only 6 fps with a small buffer). Some of that latter bit is probably the UHS-1 slot on the X-A5 and X-T100, but still, the point I'm trying to make here is that these two bodies are not the ones you want for fast action or high end video. The X-T100 does get phase detect autofocus on the sensor, though, which should improve on the slightly sluggish performance of the X-A5.

At US$599 for the body, or US$699 for the body plus the excellent 15-45mm kit lens, the X-T100 looks like a real useful addition to the consumer side of the X system and gives people another relatively small mirrorless option to choose instead of a compact camera with a smaller sensor (unless you want a shirt pocket camera, in which case you need to go with the smaller sensor cameras). The X-T100 comes in black only, or two different forms of panda-type style (light top over dark bottom, one of which is called champaign).

The EVF is only 2.36m dot, as you might expect in such a lower-priced mirrorless product, and the camera has a top shutter speed of 1/4000 and a flash sync speed of only 1/180, the type of cost reduction effort we usually see in the lower-priced cameras. The X-T100 does surprisingly retain the built-in flash of the X-A5 (GPN of 7m at base ISO).

I’m betting that this is going to be a pivotal camera for Fujifilm and quickly attract a following. I also predict that the number one complaint will become the lack of a front grip bump (the X-A5 has one), though there is an option for an add-on. The X-T100 comes with a little screw-in grip. 

Compared to the X-T20, Fujifilm’s previous lowest-cost DSLR-type mirrorless camera: you don’t get an X-Trans sensor, you have no way to do back button AF, you don’t get a full 30/2/25/24 fps 4K, you don’t get UHS-II support, and surprisingly, the X-T100 is a bit heavier than the X-T20. On the flip side, the X-T20 doesn’t have the X-T100’s nice multi-swivel LCD design. I think a lot of folk are going to get confused by Fujifilm’s multiple cameras with narrow price points and feature differentiations. 

Support this site by purchasing from the following advertiser:



Is m4/3 Still a Viable Choice?

Updates at bottom of article


After my two presentations at B&H last week, I had someone come up to me and ask a seemingly simple question: would I still invest in m4/3 gear?

Since I was one of the first pros that began a dalliance with mirrorless starting with the original m4/3 products back in 2009, that's a fair and useful question. Indeed, it was mostly because of m4/3 that I started this sansmirror.com site in the first place: I could clearly see benefits to mirrorless cameras that would inevitably make them mainstream.

It seems I'm getting questions similar to the one my presentation attendee asked more and more, though. Some of it has to do with insecurity and paranoia. But some of it has to do with market reality. 

Here's the thing: smartphones will just keep nibbling upwards in terms of knocking off dedicated cameras. It should be clear to everyone that the iPhone back in 2007 kicked off a trend that pretty did the same thing to lower end digital cameras that instant cameras (e.g. Polaroid) and disposable cameras did to the lower end film camera offerings. 

At the lowest common denominator type of image—call it a snapshot if you want—there's little reason to own a dedicated camera. The smartphones (and tablets) can achieve a reasonable, basic image that is easily shared, and have been able to do that for some time. 

But those same smartphones keep moving upwards in capability and performance. First it was with megapixel count, then it was with additional features like image stabilization, next we got faster lenses and bigger sensors that worked better in low light, and eventually we got multiple cameras and some basic computational photography capabilities starting to show up and make those small sensors seem to perform more like larger ones. 

Thing is, all these things rolled into one continue to push the capabilities of smartphone cameras upward. We've got one black and white 20mp sensor coupled with color sensors in the Huawei P20 Pro. A true monochrome sensor has a higher native ISO—because it doesn't have color filters over the image sensor—producing better low light resolution and less noise. Couple that with computational photography techniques melding a second or even third sensor producing things like color information and/or depth information, and we get a boost in terms of image quality that nibbles off another layer of dedicated cameras in terms of viability.

At this point, I'd say we're getting smartphones near the level of image quality that the 2/3" compacts were able to produce, and I have little doubt we'll eventually see smartphones that can top a 20mp 1" sensor digital camera, all else equal. 

Thus the insecurity. 

m4/3 is one stop above that 1" sensor camera. Living with m4/3 is a bit like seeing the storm on the horizon headed your way, and wondering if someone is going to come improve your house somehow so that the storm never gets there (or at least so that you can weather the storm). 

Another problem is that both Olympus and Panasonic really want to go upscale. We now have US$2000+ m4/3 cameras, and lenses that easily break into the four-figure mark, too. So another level of insecurity is that you might be pouring lots of money into your m4/3 system only to see it overrun by the smart crowd. 

I don't think that likely. Not at the top end of m4/3. But would I be buying into the lower end of m4/3 these days like I did back starting in 2009 (e.g. the E-PL# type of camera)? No. The Fujifilm X-A5 has a bigger, better sensor, and some excellent small lenses at a very affordable price: I'd buy there for a compact-like camera. Or maybe a Canon EOS M or Sony A6xxx. 

It's the E-M1m2 and GH5/G9 type of m4/3 camera—coupled with some excellent lenses—that would keep an m4/3 photographer clearly away from what the slowly rising smartphone crowd will be able to attain in the foreseeable future. 

bythom thesqueeze

My "Squeeze" slide from a 2009 presentation to Japanese companies

Unfortunately, that presents another problem. With Sony putting the A7m3 at the US$2000 mark, m4/3 is finding itself in exactly the position I cautioned about back in 2010. What I call "the squeeze." At the bottom you have smartphones nibbling upwards, while at the top end you have the big camera makers nibbling downwards wanting to own the DSLR and mirrorless world that's left. That leaves very little territory for compacts and the nearby m4/3 to occupy and own outright. 

What's that territory? Bright to moderate light capability with small/light lenses. To me, something like the Pen-F with the f/1.8 primes is a clear example of a kit whose performance you couldn't get from smaller (smartphone, compacts, even 1" sensor) or size you won't get from larger (APS-C and full frame) cameras. There's a sweet spot in there that's relatively narrow. 

Panasonic, of course, has as escape valve: video. The GH5 and GH5s are highly competent 4K video cameras with features and performance that actually take them well into the pro video realm. The GH5s now directly meets the BBC and European broadcast standards out of the box. And in the pro video realm, a GH5s is a relatively inexpensive choice. 

And that was my basic answer to the person who asked the question: I don't have any problems continuing to put time and energy and money into m4/3 on the video side. Where I do start to have issues is how much time and energy and money I want to put into m4/3 on the still image side. 

Don't take this to mean that I'm saying that m4/3 is dead. No, it's not. But Olympus, for example, has shown no real ability to grow ILC volume—they're stuck at 500k units a year and have been for some time—and I believe that's mostly due to the squeeze problem. The Canon/Nikon/Sony trio is still likely to dominate the top end of the camera market, and they'll push down from above. The smartphones and maybe some related products—consider a simpler version of the Light—are absolutely going to keep pressing upwards. 

That's leaving very little room for 2/3", 1", and m4/3 sensor products, I think. At 2/3" the two products we have left are basically "tough" cameras that can go underwater, freeze, and be dropped, plus the superduperwuperzooms like the Coolpix P900. At 1" we're mostly headed for shirt pockets with things like the Sony RX100 and Panasonic PSZ200. 

m4/3 is still a viable alternative in the middle between those products and the inevitable downward push of full frame (coupled with its little brother APS-C). But more and more, I'm considering my m4/3 usage really more niche than mainstream. 

Update: As usual with articles that take a stance like this, I got both support and pushback. I'll address some of the pushback here:

1. m4/3 has a place with telephoto, which smartphones can't do, and with a size advantage that larger sensor cameras can't match. To a degree, that's true. The usual point of reference I'm given here is the Olympus 300mm f/4 (as compared to the Canikon 600mm f/4). True, a 300mm lens is smaller and lighter than a 600mm lens. But we're also giving up two stops of equivalence with m4/3, so we'd really be comparing to full frame 600mm f/8. This latter is important, because the ones that truly want long, fast lenses tend to be wildlife and sports shooters. Both want to keep their shutter speeds above 1/1000 most of the time. You can do that with m4/3 in Sunny 16 conditions, but as light dwindles (edge of day to night conditions), the m4/3 shooter starts to struggle with image noise. Heck, even the full frame shooter often struggles. Focus discrimination also becomes a problem for moving subjects with smaller sensors using on-sensor PD techniques. So, yes, I do believe that m4/3 has a place in long lens use in a small kit, but I've also found it to be a limitation in terms of when I can shoot reliably. Is that niche defensible? Well, take a D7500 and the Nikkor 300mm f/4E PF versus an E-M1m2 with the 300mm f/4: the D7500 kit is 1395g and 450mm f/5.6 equivalent, the E-M1m2 kit is 2049g and 600mm f/8 equivalent. Add the TC-14EIII to the Nikon kit and we're at 630mm f/8 equivalent and we're still very significantly lighter, with better focus performance, particularly in low light. So those that think the m4/3 telephoto advantage is a defensible niche are over-arguing their case. 

2. APS-C is also getting squeezed. No doubt. But that doesn't make things any easier for m4/3 ;~). APS-C is trying to move down in many ways (note the D7500 example above, or witness how small and light Canon pushed the EOS M5 down to). And as they try to move down, m4/3 is right in the way.

3. Lots of lens choice, and the lenses are smaller. True to a large degree, and this is what has kept m4/3 in the game in the first place. They were first in mirrorless to have a complete lens set, and they have often emphasized small size (witness the Olympus 9-18mm, or many of the f/1.7 and slower primes). This is one of the things that attracted me to m4/3 in the first place, particularly for long backcountry hiking. But the assumption of a lot of those pushing back on this article has been that we won't see full lens sets or lens minimization in the larger sensor formats. Both are wrong, as the growing Fujifilm lens set is starting to show. Even in full frame we're seeing some smaller items appear (the Sony 12-24mm f/4 is smaller than its competitive full frame wide angle zooms). This is another trend that will increase over time, as the demand for smaller/lighter continues to grow.

4. m4/3 will continue to innovate. No doubt. But m4/3 engineering resources are massively dwarfed by smartphone camera engineering resources at Apple, Google, and Samsung. Massively. I was stunned to learn how many deep R&D engineers Apple has on photography at the moment. And remember, Apple also designs state-of-the-art CPUs that rival Intel's, GPUs, Imaging ASICs, and yes, image sensors. In the past year Apple has sold 210m iPhones and Olympus 500k mirrorless cameras, and at a higher average price. Who do you think has the engineering, cashflow, and ROI advantage to keep pushing things like computational photography forward? Yeah, it's going to happen in smartphones first and faster. Again, the article is about the "squeeze," not the death of m4/3. The m4/3 partners have their work cut out for them to hold onto a niche, and no doubt they will do just that. But that niche is under severe pressure from all sides.

And, as usual, people are misquoting me. I didn't say m4/3 would die. I didn't say m4/3 would go away any time soon. I didn't say you can't take excellent photos with an m4/3 product. I'm only pointing out that m4/3 sits right in the crosshairs of competing forces. Given that both Olympus and Panasonic did not hold serve in mirrorless against just the Fujifilm/Sony incursions (both lost significant market share), that will be even tougher when Canon/Nikon are fully in the market and adding their weight to the push from above.

I'll also point out that the original intent of the article was to answer a reader question ("should I continue to invest in m4/3?"). I didn't actually answer that question ;~). I simply pointed out the market dynamics and let readers decide for themselves. That some of you feel that nothing of significance is changing and your answer is "yes, I'll continue to invest" is certainly what Olympus and Panasonic want to hear. But I've also heard from a number of you that say "no, I'm going to wait a bit and see how things play out." 

We have a lot of choice now in mirrorless, and it's only going to get better. Ultimately, the strongest choices will win out, and it won't always be for technical reasons. 


The Rumored Claims Are Wrong

Updated to add Fuji Rumors link.

It's that time of year. With most of the Japanese camera companies having now reported their fiscal year results, CIPA having provided their 2018 estimates and early revisions, and some actual real data from the year starting to accumulate, we've got lots of market share rumors and assertions and claims running around the Internet. Happens every year. 

The big one this year is the totally unverified rumor that an internal Fujifilm memo says by "2021 Fujifilm predicts a decrease in market share for Canon and Nikon of 50%." This originally appeared on Fuji Rumors, I believe (link to the article). I haven't added it to my Claims to Remember list because it isn't a verifiable claim. It's hearsay on a rumor site with a vested interest in Fujifilm's success.

But let's tackle this with some real data. Let's assume that DSLR sales continue to plunge by 15% a year until 2021. Let's further assume that mirrorless sales grow by 20% a year during the same period. Funny thing is, that would mean that overall ILC sales would still be about the same as they are now (somewhere around 11m units a year). (Note that DSLR sales have not been declining by 15%, and mirrorless sales have only exceeded 20% growth once. So this analysis is more heavily weighted toward mirrorless than the current numbers suggest.)

Canon and Nikon currently have 95% of 7.5m units (DSLRs). Fujifilm, Canon, Olympus, Panasonic, and Sony have about the same 95% of 3.9m units (mirrorless). Canon specifically claims in their audited financial information that they're at 48% of all ILC and should reach 50% this calendar year. That's 5.5m of the 11m units, DSLR and mirrorless. 

So, for Fujifilm's rumored claim to be true, Canon would have to hit 25% ILC sales, or 2.75m units. That alone would require them to lose DSLR market share to Nikon in the next three years! It would also assume that Canon had no significant mirrorless market share. Neither of those things seem at all likely. It also ignores what surveys of current Canon and Nikon owners show: many of those owning Canikon products are waiting for those two companies to produce more significant mirrorless models. 

The other problem with the Fujifilm rumor is that Sony is another factor. Sony sacrificed market share as it transitioned from DSLR to SLT to mirrorless to full frame mirrorless. They've now created a solid full frame lineup that's showing real growth in sales, though full frame is a smaller portion of overall ILC sales. I have no doubt they'll now defend their full frame products and start to backfill the entry level, crop-sensor products again.

As far as anyone can tell, the current situation is this: Canon 50%, Nikon 22%, Sony 14%. That's 86% of all the DSLR/mirrorless units that are being produced, of which Canikon is 72%. The purported internal memo is said to claim that Canikon would therefore be at 36% in three sale years. That's a change from almost 8m units to 3.5m units in three years (using the current CIPA numbers and those 15% decline/20% increase estimates I noted above). I find that unfathomable. 

Moreover, for that to happen, Fujifilm, Olympus, Panasonic, and anyone else picking up that huge market share slice would have to pick up new sale outlets here in the US. That's a costly endeavor, as getting shelf space is not as simple as "we make a better product." 

Realistically, the coming few years are going to see:

  • only Canon and Nikon truly pursuing the remaining DSLR market—which will likely still be 4.6m units in 2021 ;~)
  • coupled with Canon and Nikon fully entering all aspects of the mirrorless market (likely 6.2m units in 2021). 

What you have to look at if you're at all interested in the market share positions is this: mirrorless is currently about 4m units (12-month trailing average) with Canon, Sony, and Olympus being the three biggest players at the moment. Canon is going to add products, Sony is going to add products, and Nikon is going to re-enter. By 2021 that mirrorless market is going to be perhaps 6.2m units, so there's room for both Canon and Sony to expand and Nikon to enter. Expecting any of those three big players to fail completely is just irrational speculation. 

But let's run a scenario: let's assume that Nikon enters and can't get more than 10% of the mirrorless market and can only hold it's 20% of the DSLR market. That is still 1.5m units out of the 10.8m predicted. Let's further assume that Canon expands to 80% of the remaining DSLR market and can only grab 20% of the mirrorless market (below where they appear to be today). That's about 4.9m units. Hmm. Add those together we still have 6.4m units of 10.8m, which is about 60% of the market. The rumor predicts that they'd only have 3.9m units, which is lower than just the DSLR estimate alone (again with 15% decline every year between now and 2021).

I call hogwash. Someone had a fantasy dream and decided to see if they could get others to believe it was predictive. 

The big three ILC companies are Canon, Nikon, and Sony. Indeed, they're the ones with the best fully global presence, with Canon and Sony having other major product categories than cameras that make them fully global, and which they leverage the cameras on. Nothing dramatic is going to happen with the Canon and Sony presence in the camera market. And, if Nikon is truly awake, they'll shore up their gaps, too. 

The safe bet is that Canikony currently have 86% of the ILC market share, and that in 2021 it's pretty likely that they'll have no less than 80%. Can Fujifilm, Olympus, Panasonic, and Pentax nibble a bit off in a flat market? Perhaps. But I'd say that the only way that market shares dramatically alter is if some company has a truly breakthrough product, one that renders even the current best mirrorless products as "not nearly as good." 

There are only three places that such a change could happen: the image sensor, the camera intelligence (CPU/AI), and the camera workflow (integration into the Internet). Who has the biggest sensor R&D? Canikony. Who has the biggest camera intelligence R&D? Canikony. Who has the parts necessary to integrate better with the Internet? Mostly Sony, but this is really a software thing more than a hardware one, so all the Japanese companies have issues here.


Sony Kando 2.0

Caution: semi-connected rambling ahead...

bythom US CA Monterey ZS200 21487


The Urban Dictionary describes “kando” as "one who excels in an activity to a level of such proficiency as to make the opposition and fellow teammates alike appear mentally challenged.” I’m pretty sure that’s not what Sony intends with their use of the word kando.

Indeed, then Sony CEO Kazuo Hirai’s definition was quite different when he presented the term: "Kando translates to mean emotional involvement.  The power to stimulate an emotional response.  The power to make people say, ‘Wow.'  All Sony products must be inspired by a spirit of kando.”

Both definitions was going through my head last week when I attended Kando 2.0, a now annual event that Sony puts on and which has several components to it. 

This year’s Kando was held in Monterey, California, at the Asilomar Conference Center to be exact. It was a risky choice, as weather (fog, mist, overcast) is definitely possible, but it appears Sony picked the perfect week, and the California coast line cooperated for pretty much everything on the schedule. And boy, everything was on the schedule. 

Sony appears to use Kando as a three-pronged event. Sony employees and managers were first to arrive, and there were activities scheduled around company things. Next came many of the Artisans and some invited guests. Sony employees, a few invited press and pros, and the Artisans interacted, with the Artisans teaching and inspiring the Sony team and others, and vice versa. Finally, there was the last two-and-a-half days, where about 200 selected individuals come for a jam-packed schedule of workshops and shooting opportunities. 

Just to give you an idea how diverse the shooting opportunities are at Kando, this year we had: kittens and puppies, food, flowers, time-lapse, landscape, astrophotography, surfing, portraits (including underwater), rock concert, street, long exposure, and a zillion other choices. If you needed gear for something, it was either supplied (e.g. intervalometers and filters for the long exposure workshop) or could be borrowed (camera bodies, lenses, Manfrotto tripods,<tk> lights). Microsoft supplied 30 Surface computers for use, Epson had big printers ready to output what you shot. 

Then there was the swag bag (a Manfrotto photo backpack filled with small goodies). 

bythom kando goodiebag

To get into Kando, you have to submit an essay as to why you should be there, and point to a body of work. I did, and I was accepted. But I noticed most of the 150 that were accepted all had very visible and different kinds of Internet presence (not just in photography, but in other disciplines, as well). It seems clear to me that Sony is doing at least some of their selection process based upon how well the things people heard, learned, felt, or experienced at Kando 2.0 would be amplified out into the far reaches of the Web. 

That’s a nice touch, though not perfectly executed. I’ll have more to say about that later. 

So why did I apply for and end up at Kando 2.0? 

I’m perpetually in learning mode. When I saw the initial list of folk presenting, the types of shooting opportunities, the companies that would be there, and the fact that Sony Professional Services would be there with the full set of Sony loaner gear so I could try out a few of the lenses I haven’t yet used, I saw the opportunity to spend some intense time dipping into the Sony universe as I worked on finishing up two Sony camera and a couple of Sony lens reviews, and to maybe learn some new tricks. Oh, and get inspired shooting things I don’t normally shoot.

It’s when I’m around other photographers who are doing different things than I that my brain tends to ignite with new ideas and creativity. And that tends to spill over into my own shooting and teaching in ways that are unpredictable and sometimes surprising.

For instance, upon arriving, for some reason the first thing I ended up doing was get into a long discussion with the Capture One folk. I have to admit that I’ve probably paid less attention to that program than I should have. As we wove deeper and deeper into the product, I was introduced to a feature I didn’t know existed in Capture One. It’s actually not something that looks like much until you consider the workflow implications. Oh, wait, I was supposed to do a lecture on field workflow at B&H the following week. So we deep ended real quick into field workflow and I realized that—at least for macOS users—a trio of products that included Capture One could make a very compelling field workflow that covered all the things that I say a field workflow needs to deal with. Ruh-roh. Time to make some additions and changes to next week’s presentation ;~). 

But it got deeper and more creative quickly. 

At one point in the discussion, the Capture One demonstrator wanted to show off the grading feature. And he did the thing we all do, all the time, no matter what product we’re using: because the controls are kind of small when docked in the panels, he pulled it off to create a bigger window for the controls. 

The first thing that struck me was this: about a quarter of the image we were working on was now blocked from view. Most of that big window that was created was just black background (as in most post processing UI you keep controls/panels in dark mode so that it’s not distracting you from the image). 

Aha!

Why not just get rid of all the black and make that big dialog transparent? In essence, go to a HUD (heads up display; the Japanese would call this OSD, or on-screen display) type of use? The grading dialog in Capture One is really just three thin, directly-manipulable circles. Why shouldn’t I be able to center those over the area I want to grade (selection) and manipulate the dials directly over the image? Well, uh, because that’s not the way the designers designed it. 

Which led to a discussion of why the Nik U-Point technology was so interesting and useful. You picked a point, the HUD-type display that emanated from that point allowed you to directly manipulate the data within the point’s control. Doesn’t clog up the screen with big dialog boxes/panels.

Technically, you could do this HUD-thing with Levels, Curves, Grading, HSL, a whole host of things we want to manipulate while we’re looking at the image. Just present me the Curve line overlaying the image and let me directly manipulate it. Eezy peezy. 

From there I went down to the beach to see what was going on and I started thinking about our cameras. Yeah, HUD exists in the EVF/Live View world. Press the Display button on most cameras to toggle through the Ugly-HUD choices. Most of them are variations of putting lots of icons of current settings over the image area. Even when you bring up a histogram, it’s the full histogram box, and it’s taller than it need be. 

What if it were more HUD-like? A wide-but-not-tall histogram that had “tells” at each end to show you that you’ve hit black (0,0,0, or a settable value) or saturation (255,255,255 or a settable value)? The Olympus mirrorless cameras can do this without a histogram, doing a form of peaking display at both ends of the exposure. And of course peaking itself is a form of HUD display, though it doesn’t tell you much about the range of values, just what parts have hit the extreme edge of the exposure capture.

At this point, my mind was racing down the possibilities. 

Some of the mirrorless cameras (Canon and Olympus, for example), allow you to drag your thumb on the LCD to position the focus point while you’re looking through the viewfinder. Good first step, but what if I wanted to set focus settings? (Bringing this back to Kando, I had a brand new Sony camera I needed to set up in my hands at the time.)

Well, many of the Sony lenses have this well-positioned lens function button that is naturally right where your left thumb sits. What if I were dragging my right thumb on the LCD to position focus, but if I pressed and held the lens function button I’d immediately get a Focus HUD centered where my thumb was: now move my thumb up to set AF-S/AF-A/AF-C, or left to set AF/MF, or right to set Area Mode, or down to set Eye/Face detect functions. Since the viewfinder is EVF, that would allow me to control all focus operations with two thumbs while looking through the viewfinder and staying ready to shoot.

Now wait, where are the settings on that new A7m3 I was holding that I needed to set for focus? ;~)

After a lot of menu scrolling and trying to remember what the cryptic descriptions/names meant I eventually got my camera set up the way I wanted it. It didn’t help, of course, that I was in bright sunlight and the LCD/EVF are not set by default to be used in such bright light, so I made slow progress to getting everything set right. Of course, by this time at Kando 2.0 I was already half-way through shooting a five-hour music event that had popped up (I told you I was going to ramble).

bythom US CA Monterey A7III 19588

Access to the musicians was so good it was actually more difficult to shoot them. Literally, I could have stuck a camera in their face (see above). 

bythom US CA Monterey A7III 19550

But because the photographers could get so close, it was difficult to actually take a shot of the whole band—actually, there were five bands—because there were so many photographers in front of the musicians! 

bythom US CA Monterey A7III 19805

Still, it was good warm-up for me, as I was being stressed in lots of different ways: I haven’t/don’t shoot rock concerts, the sun was setting behind the stage, I had lots of other photographers to watch out for and deal with, and, of course, I had a new camera in my hands that wasn’t even set up yet. So I probably wasn’t going to win any awards for what I shot of the musicians, but still, this was fun and it was beginning to stretch my brain, even though I was just off an all-day, cross-country set of jet rides. Exactly the reason why I wanted to experience Sony Kando.

bythom US CA Monterey A7III 20013crop

The next day was a smorgasbord of learning and shooting. In the morning Sony Artisan of Imagery Thibault Roland quickly taught us the way he approaches making really long exposure photographs during the day and it was off down to the beach to try to make it work in practice. (As an aside, don’t use variable ND type filters for this, as they are essentially dual polarizers and you get caught in all kinds of issues with edge-to-edge variation in filtration, particularly when you’re shooting wide. Buy a set of dedicated ND filters that can take you down 12 stops or more when used together and it will be oh so much easier. The chart Thibault gave all of us actually goes down to 16 stops, which is what he needs to get those truly long exposures he shoots during the day)

bythom US CA Monterey A7III 20159

Next up was shooting flowers with Artisan Caroline Jensen. She quickly went through her impromptu handheld style of flower shooting that produces her unique looks, then it was time for us to all go to the Flower Truck—did I mention that Sony was providing everything we needed as subjects?—to pick some flowers to work with and then go off and see what we could get using her techniques and with her advice as we shot. 

bythom US CA Monterey A7III 20071

The final shooting experience for me on that first full day was surfing just before sunset. I came to the beach equipped with a peashooter (70-200mm). So I did what any native Northern Californian is not at all averse to doing: I wandered out into the water. Which turned out to be a good choice, as a couple of times it gave me angles that others weren’t getting sitting back from the water with their longer 100-400mm lenses. You'll note in most of these images the surfers are coming down to me. Good thing I grew up in Northern California and dealt with cold water all my life (Pacific Ocean, Tahoe). 

bythom US CA Monterey A7R3 1706
bythom US CA Monterey A7rIII 21003
bythom US CA Monterey A7rIII 20674
bythom US CA Monterey A7rIII 20560

Now, I’m not going to go through everything I shot at Kando. We’d be here all day. But I did want to point out just how diverse the shooting opportunities were. If you want to see what the pros and Artisans were shooting during the week, try this Kando page. Or try this Kando page to see more of what the participants I was with were doing.

Sony was encouraging everyone to post to social media using a Kando hash tag. That’s not my way of sharing an experience (this article is), but it definitely meant that there was a lot of imagery coming out of Asilomar with Sony branding on it. Indeed,  most of the gear raffles (read on) were drawn from random selection of those that posted with Kando tags. Thus, I probably wouldn't have posted these images on Instagram to avoid the conflict of interest situation anyway.

That’s not to say that Kando was all about shooting. It wasn’t. There were discussions—for example on photography usage on social media—inspirations (Guy Kawasaki’s keynote address), and even focus groups. And there were giveaways, giveaways, giveaways. Seems like at every meal there were raffles going on with Sony handing out things from plane ride shooting with Chris Burkard to Playstations to Sony A7Rm3’s, plus lots of great lenses and accessories. 

I mentioned earlier about the amplification of the event out into the Internet. Curiously, I don’t think that all went as intensely as it could have. The AlphaUniverse Web page was displaying Kandoimagery, and there were all those #SonyKandoTrip hashtags on Instagram and Facebook running around. But this marketing effort felt incomplete to me. You kind of had to know where to go or what to look for if you were going to find the fire hose of images coming out of Monterey from Sony gear. And while some key photo press was here for the second stage of the week before the rest of us arrived, they weren’t amplifying the Kando event as usefully that I could see. Maybe more is coming. 

To me, I think Kando needs more connection to the folk who weren’t here. This would be the time to publish Kando tutorials from the Artisans. It would be the time to tie in raffles for people not at Kano on AlphaUniverse. I felt very “engaged” at the event. I’m not sure anyone not at the event felt engaged.

For me everything at Kando seemed to have surprises lurking. It seems like I knew one heck of a lot of people who were here. Old pro photography buddies like Bob Krist and Patrick Racey-Murphy. Some of the key personnel manning the booths at the exhibition. And yes, Guy Kawasaki recognized me from my Silicon Valley days and he immediately challenged me: “I hope you’re still using a Mac” (my software company was one of the ones that Guy had evangelized and got into the Mac developer program before it was announced). And yes, Guy, I am.

Despite all the older folk I knew, the crowd felt distinctly young, energetic, and enthusiastic. There was tons of energy all over Asilomar and throughout Kando 2.0, and interestingly everyone was engaging in conversations with one another as we bustled from one thing to another or sat for meals. The Sony personnel were ubiquitous and gregarious. And that proved to be infectious. Bravo. Every camera company should be as engaging to their customers as Sony was to those of us at Kando.

I’ve been to a lot of company-sponsored events over the years, but Kando 2.0 has to be at or near the top of the list. Fun, fascinating, fervent, festive, and fruitful. We need more of this type of thing happening in photography. It makes you want to create more photos, and more interesting photographs. It makes you want to stretch yourself and your gear to its limits. 

So yes, Hirai-san, I had an emotional experience, and “wow” isn’t a bad word for it.

Now, about those small buttons on the cameras...

The Problem for Canon and Nikon

Rumors have it that Canon and Nikon will both introduce full frame mirrorless cameras later this year. I believe those rumors to be true. That would put Canikon almost exactly five years behind Sony's original A7/A7R full frame mirrorless camera introduction. 

What does five years' head start mean?


50 prime lenses, and 13 zoom lenses. Four camera bodies.

The zooms:

  • 12-24mm f/4
  • 16-35mm f/2.8
  • 16-35mm f/4
  • 24-70mm f/2.8
  • 24-70mm f/4
  • 24-105mm f/4
  • 24-240mm f/3.5-6.3
  • 28-70mm f/3.5-5.6
  • 28-135mm f/4
  • 70-200mm f/2.8
  • 70-200mm f/4
  • 70-300mm f/4.5-5.6
  • 100-400mm f/4.5-5.6

The prime focal lengths (from Laowa, Lensbaby, Mitakon, Samyang, Sigma, Sony, Tokina, Voigtlander, Yasuhara, Zeiss):

  • 180° fisheye
  • 10mm
  • 12mm
  • 14mm
  • 15mm
  • 18mm
  • 20mm
  • 21mm
  • 24mm
  • 28mm
  • 35mm
  • 40mm
  • 50mm
  • 55mm
  • 56mm
  • 70mm
  • 85mm
  • 90mm
  • 100mm
  • 105mm
  • 135mm

These are just the full frame lenses in the FE mount itself (not E or APS-C), not counting lenses you might adapt to it. Yes, some of the primes are manual focus lenses, and some are on the oddball side (Lensbaby's 56mm), but these are all lenses that just mount right up to a Sony A7 or A9 series body. 

Canon and Nikon between them have well over twice the full frame lenses currently available, and third party offerings probably double that again. But those are for the DSLR EF and F mounts. 

It's going to be interesting to watch how Canon and Nikon tackle the lens deficiency problem they enter full frame mirrorless with. They could, of course, just make mirrorless cameras using their current mounts and lenses wouldn't be an issue. That, unfortunately, would put an empty snout on their mirrorless cameras that the Sonys don't have.  (I have pointed out elsewhere that such a snout doesn't have to be empty. It can hold controls, it can have a drop in filter slot, it can still have function.)

The expectation at the moment is that Canon is unlikely to abandon their EF mount. It's also used on their video cameras, for instance, so bifurcating the full frame lens lineup adds a degree of complexity that I'm pretty sure Canon would rather avoid, if possible. Moreover, Canon made the switch to a modern all electronic mount back in the early days of autofocus: the only real benefit to be gained by changing mount now would be in reducing the flange distance. That can reduce the camera depth—but not if you use the current size hand grips DSLRs have for ergonomic purposes—and open up some new smaller lens design possibilities, most notably in the wide angle range. 

Nikon, on the other hand, has a tougher problem. Because they've used the same mount from the beginning, they now have a lot of add-on variations they've bolted to it: AI indexing, screwdrive autofocus support, multiple electronic connection sets, mechanical aperture linkage, electronic aperture linkage, and more. Using the existing mount comes with a lot of extra baggage, much of which Nikon is surely now wanting to leave behind (they've left off AI indexing on the D3xxx, D5xxx, and D7500 DSLRs, they left of screwdriver autofocus support on the D3xxx/D5xxx models, and they left off the mechanical aperture links on the E-type lenses). 

Sony's put up big numbers with their five year head start. Sony has a pretty full lens set available for the A7 and A9 models now, with only one major exception: primes and exotics past 135mm. There is the announced-but-not-yet-shipped 400mm f/2.8, but that's only one of at least a half dozen lenses they're missing above 135mm to fully compete with Canon and Nikon. 

To compete in mirrorless, both Canon and Nikon need a strong story. They're going up against a line of four cameras and 63 lenses, after all. You can't just say "we make better cameras, and here is one camera and a handful of lenses." Sony is clearly offering a lot of choice to mirrorless users interested in full frame. 12mp, 24mp, 42mp. And these are third generation camera designs, too.  Any lens you need (up to a point). 

It's going to be interesting to watch this game play out. Play out it will, because Canon and Nikon can't just sit around any longer. Both are experiencing some erosion in key DSLR product lines, and both would love to have the simpler manufacturing and potentially lower costs of producing mirrorless (versus DSLR). 

My belief is that Sony's biggest threat is Canon. Why? Well, there's Canon's size in the camera market (50% of all interchangeable lens cameras sold are Canons). There's 100m Canon lenses in customer hands. They're also just as good at marketing as Sony is. But the biggest point may be this: so far in surveys I've done it's more likely that a Sony A7/A9 shooter came from Canon than Nikon. Canon wants those customers back, and they'll move aggressively to make that happen. They don't want any erosion in a market they dominate.

Nikon, on the other hand, is an enigma. They're both tough to discount (because of their long-established leading engineering), and they're tough to worry about (because they are shrinking in size while their marketing is not as strong). While they have almost 100m Nikkor lenses in customer hands, it's unclear whether they'll continue that legacy forward. Nikon also seems less concerned about competitors at the moment than they are at stopping the slide in volume they've experienced since 2013. 

Most bets are on Photokina for the game to start. That means August or September product announcements and a lot of hype at the show itself (this year starts September 25th). 

Funny thing is, the game isn't just going to be played by Canon, Nikon, and Sony. I'll wager a bet that Leica has a new full frame camera coming, and that Fujifilm and Hasselblad will take their mirrorless medium frame to 100mp at the same show. 

What that means for Sony is that they're going to be fighting a two-front battle. The A7Rm3 has to hold its own against a GFX100 and X2D, while the A7m3 has to hold its own against Canon, Leica, and Nikon offerings. 

Competition is good. Sony can't stand still. They're going to need to flesh out the features, ergonomics, and performance of the A7 and A9 models even more than they have. Because the competitors are coming. The one big arrow they have in their quiver at the moment, though, is lenses. I don't expect Sony to slow down at releasing new FE lenses.


The War Over Nothing

It’s been interesting to watch the Internet Intensity Engine fire up with the Blackmagic Design 4K Pocket Cinema Camera (4KPCC). Starting the day the 4KPCC was announced, threads all over the Internet began exploding with “Panasonic’s done” messages. 

Apparently the 4KPCC is going to cause everyone to stop buying Panasonic GH5 cameras because this new 4K camera is half the price. 

Nonsense. 

Realistically, these two cameras are different in so many ways that it’s difficult to enumerate them all. Moreover, Blackmagic Design is a bit more focused on film-like looks, while Panasonic is a bit more focused on video-like looks. That alone is enough to distinguish the two. 

Then there’s the smaller swivel LCD on the GH5 versus the big fixed LCD on the 4KPCC, the autofocus abilities (or inabilities), the IS differences, and more, more, more.

To me, I’m likely to try using a 4KPCC on a fixed support as a B camera. I’m more likely to move around with the GH5, and more likely to use it as a prime camera. It’ll be interesting to see if I can cut between them and grade them well enough that they look the same. If so, I’d be very happy owning both. 

The really funny aspect to this is that Panasonic already created a firestorm with the GH5s: which one do you buy, GH5 or GH5s? Again, there are differences, and they’re important.  

These nonsensical X is better than Y posts just gets us back to the “do you really know what you’re doing and what you need” thought. Almost none of those posts declaring a “winner” seem to have much analysis to them on the need side. It’s all emotional fandom. 

Look, I’m happy I have another video camera choice that uses my m4/3 lenses without a crop. I can see a couple of 4KPCCs as secondary cameras in my studio, but probably not as main cameras (basic autofocus need, for one). Ditto in the field. 

But I’m also not going to form any real opinion on this until I’ve had the opportunity to test them. I had real trouble grading the original PCC footage into my timelines with my main cameras. And low light performance wasn’t particularly good, either. 

So I’m just sitting back and enjoying the fan-flamed filmmaker fora follies as pure entertainment. I suggest you do, too.

The Mirrorless Prisoners' Dilemma

Game play has all kinds of interesting applications to real life. It's one reason why I love to play games, especially those that involve strategy over just straight out skill: it makes me think about and test strategies.

One reader reminded me (thank you) after my Attachment Rate article that Canon and Nikon are in a strange form of prisoners' dilemma. The original version of that game is simple: if you and your cohort rat each other out to the cops you both serve 2 years in prison; if you both remain silent, you both serve 1 year in prison; if one of you turns in the other but the other remains silent, the silent one gets 3 years in prison and the other remains free.

You only truly win if you rat out your associate and they stay quiet. (There are plenty of other variations of this game, and as you'll see, our mirrorless version is more complicated.)

So how's that apply to Canon and Nikon and their likely full frame mirrorless entries? Both companies have to make the same choice (new mount or existing mount), and the outcome will be decided based upon what they pick. For example:

  • Canon and Nikon both pick existing mount: Nothing really changes if the cameras are relatively equal (my assumption throughout this is that a Canon C7 is basically equivalent to a Nikon D7 is basically equivalent to a Sony A7III). Relative ILC market shares stay the same as the DSLR duopoly slide over to mirrorless. Some might say Nikon "loses" in this scenario, as they wouldn't make inroads against Canon, which is clearly something they'd like to do. But it's not really a loss. It's the status quo maintained between those two if they both move in the same fairly short time period. If there's a loser in that scenario, it would be Sony, because a user choosing Sony from scratch basically requires buying all new lens sets, while the huge base of Canikon DSLR owners shifting to mirrorless won't have to. Result prediction: Canon, Nikon, and Sony in that order of ILC market share.
  • Canon and Nikon both pick new mount: Nothing probably changes in the Canon/Nikon market share relationship, but this empowers Sony, who is already there with a large line of lenses that Canikon couldn't duplicate day one. Result prediction: Sony, Canon, then Nikon.
  • Canon picks existing mount, Nikon picks new mount: This is a nightmare scenario for Nikon. Canon users can pick and choose between mirrorless and DSLR and use their lenses on both. Nikon users that want to try mirrorless have to start a new lens collection or use awkward adapters. And again, Sony is already there with a full set of lenses, so it's Nikon that's a clear loser in this scenario. Result prediction: Canon, Sony, then Nikon.
  • Canon picks new mount, Nikon picks existing mount: This is a problem for Canon, as it means that they have to count on adapters to appease their current user base. Result prediction: Nikon, Sony and Canon, with those latter two probably duking it out for second.

Of course, this is an advanced and complicated Prisoners' Dilemma: Canon and Nikon could both choose to opt for both courses: they could make both an existing mount and a new mount mirrorless system! I'm not even sure how to score that scenario, as the number of combinations and permutations rise considerably, and it's unlikely that they'd do so with equal cameras.

But one thing is pretty clear from the above: Nikon has to choose to use the existing F-mount for full frame or they strongly risk losing the number two ILC market share. Canon has slightly more flexibility, both due to their larger established base, but also because their EF lenses have demonstrably worked well on adapters; the same isn't true of F-mount Nikkors because of all the small variations that are in the legacy base coupled with things like where the connectors are.

Of course, the notion of creating a new mount still has its desires: it means you can take everything you learned that you did wrong in the past and fix it once and for all with the proper dimensions and attributes in the new mount. Ah, the Sirens call of tech: start from scratch and all will be well in the end. Many a shipwreck has occurred listening to that call. Of course, other shipwrecks have occurred by not listening to it (Kodak being a prime example). 

Teams of folk at both Canon and Nikon are certainly aware of the above. They've already spent long days analyzing how the choices may or may not pay off for them. If for some reason they've decided to pick a suboptimal choice in the basic prisoners' dilemma problem they face, that also means that there are teams already dedicated to trying to spackle over that problem with marketing and technology messaging. 

So hang on to your horses, mirrorless world. We're about to see the big rodeo players trying to ride the bull. 

Attachment Rate

dpreview's recent interview with Fujifilm executives literally started with a comment that should be paid attention to: "Even more impressive is the lens attachment rate, as we’ve sold so many lenses as well."

What is a lens attachment rate, and why is it important?

Basically, attachment rate is the number of lenses sold versus bodies. If you sold only one lens per body, then the attachment rate would be 1. Moreover, there'd be no reason why that camera should be an interchangeable lens camera ;~). 

Historically, the overall attachment rate has been around 1.65 lenses per body sold, with a low of 1.5 and a high of 1.68 over the past 12 years. Here's the pertinent CIPA chart:

bythom cipa attachment

But this number is extremely deceptive. One of the things I discovered in surveying tens of thousands of enthusiast digital camera users is that they upgrade bodies regularly. So consider someone who started with a Nikon D70, then upgraded to a D90, a D7100, and recently to a D7500 (that's an every-other generation upgrade). If they followed the attachment rate, they'd now have 6.4 lenses for their camera on average. What my surveys showed is that this isn't quite right. The on-going updater actually has an average of 8-10 lenses, which would imply an attachment rate of something closer to 2.

At the other end of the extreme, the more casual shooters who just think they need an ILC and then buy a superzoom to use with it don't tend to buy another lens. The superzoom stays stuck on the camera, and they upgrade far less often. I don't have as good information about this group as the regular upgraders, obviously since my sites tend to appeal to someone who is more enthusiast and trying to keep up with their equipment, not the casual consumer. But the attachment rate I measured for this group is something between 1.1 and 1.2. 

Now consider the person that buys into a new mount, which happens a lot with mirrorless. Yes, there's the "adapter route" that many think they're going to take, but I don't see many actually following that usage for long. Instead, what I measure is higher attachment rates. So Fujifilm shouldn't be surprised that they've seen high attachment rates for their new medium format mirrorless camera. The only person that's going to buy that is a serious enthusiast or a pro, and they not only have a higher attachment rate to start with, but they're starting from scratch when they buy a body.

Now why does all this make a difference in mirrorless?

Olympus and Panasonic made a mount switch in 2009 (to m4/3). They filled out a full lineup of lenses (and keep doing so), but I'll bet that they see higher initial attachment rates among switchers (people moving to their mount) than they now do among updaters (people upgrading bodies every couple of generations). Sony made sort of a double switch (E in 2010, FE in 2013). With E they took a mostly consumer approach, mimicking what Canikon did (wrong) with EF-S and DX, and I'll bet that they saw a very low attachment rate, which is probably one reason why we haven't seen a lot of E-only lenses lately. With FE, I'm pretty sure that Sony has a very high attachment rate, because the A7/A9 crowd is a more sophisticated enthusiast/pro and needs new lenses.

So what about Canon and Nikon?

As far as I'm concerned, Canon has made the same mistake with EF-M as Sony did with E. Target low and you get low results. 

Full frame, on the other hand, is going to be interesting. If Canon really does use the EF mount for their full frame as my sources say will happen, that changes the dynamic entirely. Canon already has an installed base of over 100m EF lenses. Canon won't get an attachment rate bump if they use the same mount. 

The same is true for Nikon, as well. Their crop sensor mirrorless prototyping has been all around a new mount (mimicking EF-M). But their full frame prototyping has been both new mount (Sony choice) and old mount (Canon apparent choice). 

This has to be an agonizing decision for Canikon. All those legacy lenses in closets and in use are their strength, and one reason why they dominate DSLR sales (and overall ILC, as well). Keep the mount and they enable their entire user base to pick mirrorless or DSLR without much or any of a lens penalty. But their attachment rates will be low. The number of lenses they're selling now will be about the number of lenses they're selling later, so you won't see either of them proclaiming, as did Fujifilm, to be surprised by lens sales.

On the other hand, a new full frame mirrorless mount from Canikon would mean a higher attachment rate and more new lens sales. But it would also mean that they would be five years behind Sony in terms of lens choice and supply, which means they'd still be enabling people to switch. 

So which will it be? We'll know before the end of the year, I'm pretty sure. If it had been up to me to make the decision, I would have voted to stick with the current mount and just deal with the snout that the empty mirror box would leave behind, and deal with the attachment rate problem a different way (e.g. marketing).

As you might have noticed, the mirrorless cameras are getting bigger. Another quote from that Fujifilm interview at dpreview: "One purpose of us doing the X-H1 is that some customers actually requested a bigger grip and better handling, especially together with bigger lenses like the 100-400mm." Yep. When you put two pound lenses out front, you're going to need a grab handle, not a little soap bar design camera body.

That bodes well for keeping the lens mount in the Canikon full frame mirrorless world. No way do I want to put my 400mm f/2.8 on an adapter on a small body.


text and images © 2018 Thom Hogan
portions Copyright 1999-2017 Thom Hogan-- All Rights Reserved
Follow us on Twitter: @bythom, hashtags #bythom, #sansmirror