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.

Fujifilm Pro Services

Fujifilm announced a pro services plan for the GFX in the US, which will begin on May 1st.

bythom fps


This pay-to-play plan is similar to other professional services, in that it includes cleaning services, faster and discounted repairs, and loaner cameras.

There are some odd bits and pieces. Two that caught my eye are the requirement that you join within 30 days of buying the gear in question, and the fact that Fujifilm disclaims providing refunds for almost any kind of dissolution of service. Also, you Alaska and Hawaii folk are not eligible for the service; only continental US street addresses are allowed.

One thing to note: the combination of three-year extended warranty and joining the Fujifilm Professional Services is highly discounted (US$599 instead of US$798). That option doesn't appear on the main page—and I don't have a GFX camera to register so can't see the subsequent pages—so I'm not sure how easy it is to sign up for that option.

Sony Goes Further Upscale with the A9

bythom sony a9

Sony set a new record today in their announcement of the new A9 mirrorless camera. Oh, not the shutter speed, frame rate, or other stuff you might be thinking of; the record they set is in the number of footnotes needed in the first 88 words: 9 (actually 10, since they later footnote something that appeared in those first 88 words).

I always get curious when I see footnotes. Sony is making a lot of claims with this new camera, and the footnotes tell the fuller story that you might miss if you were to read the top level specs.

But let’s get to the basics first. The A9 is a US$4500, full frame, 24mp mirrorless camera that’s pretty much geared for speed. Frame rate? 20 fps. Silent shutter speed? 1/32,000 (though there’s a footnote). 241-frame raw buffer (another footnote). The A9 has dual SD slots (another footnote). A much bigger FZ100 battery that nets a 450 shot CIPA average. A better EVF than anything Sony has done previously, with 3.7m dots running at 120 fps (another footnote). From a spec sheet standpoint, there really isn’t a lot to complain about. Sony’s covered all the bases to make a top end camera. The A9 should ship next month.

The body itself seems like a slightly beefed up A7 Mark II series body, with a couple of changes and additions Sony A7 users will like (a focus joystick, which is absolutely necessary with 693 autofocus points scattered across almost the whole frame (93%), a dedicated focus setting dial, etc.). There’s a two-battery vertical grip option, and a plethora a multiple-purpose buttons on a weather-resistant body.

Inside we’ve got a new 24mp 36x24mm sensor that’s tricked out with BSI (back side illumination) and stacked electronics and gapless microlenses all with copper and aluminum wiring for efficiency. Sony makes a point at the beginning of the press release to call attention to the fact they’re the world’s largest sensor maker (“…a worldwide leader in digital imaging and the world’s largest image sensor manufacturer…”). It’s likely the stacked electronics that enable many of the speed aspects of the camera. That’s because in electronic shutter mode there apparently is no (or little) rolling shutter effect, and in 4K video mode the sensor pulls a full frame 6K data set that the electronics downsize to 4K. It takes lots of on-sensor or off-sensor bandwidth to do both those things, and it appears much of this is being done on-sensor.

Behind the sensor is a 5-axis image stabilizer that has a claimed 5-stop CIPA rating with a 50mm lens on two axes (another of those pesky footnotes).

So about those footnotes. Some are ignorable, a few are important. For instance, while the buffer and frame per second calculations are correct for one card slot, the second card slot is not UHS-II. Why camera makers think this is a good thing to have differing slots when they are constantly performing integrity checks on the disk tables on cards, I don’t know. Basically you’re always limited by the slowest card in the camera. So shooting to both cards at the extremes of what the camera is designed to do is likely to have some downgrading effect, probably mostly on buffer.

Likewise, that 120 fps viewfinder is actually 60 fps if you want the uninterrupted view with the electronic shutter. Not a terrible deal, but bragging about a spec in one place and then disclaiming for a critical use elsewhere takes away some of the excitement. Likewise, the 1/32000 shutter speed is only available in the Manual and Shutter Priority exposure modes (otherwise you’re limited to 1/16000). Note also that at apertures beyond f/11 focus doesn’t track with the electronic shutter, too.

I think it’s going to take some shooting with the A9 to really get a sense of its ability and how these small disclaimers impact actual shooting. One of the footnotes that concerned me was “25% faster autofocus" than the best A7 model to date. Thing is, I’ve never really complained about the speed at which an A7 focused. It’s the accuracy with which it focuses. Clearly Sony is doing phase detect to initially move the lens and contrast detect to finalize focus. But that finalization is coming off 60 fps data. The thing about the Sony focusing I’ve found is that it is incredibly good at a lot of things, but motion tracking to or from you tends to make for a lot of small misses in continuous shooting. Small misses that net usable images, often because of depth of field, but not tack sharp images.

Sony seems to be trying to say the A9 is a 1DxII or D5 equivalent—especially since they went so far as to copy Nikon’s Ethernet connection, among other things—but the proof is in the shooting. The thing about Nikon’s latest DSLR focus system is that it is fast and precise in tracking rapid and erratic motion. The DSLR phase detect focus design just has more focus discrimination in it due to the geometries involved, so it doesn’t need a second step to verify focus, and the D5/D500 certainly are doing focus calculations far faster than 60 fps with their dedicated sensors and CPU.

Still, any improvement from the A7rII is welcome and definitely will give DSLR users pause. That’s where the lens situation starts to become an issue for Sony, though. We’ve got a 70-200mm f/2.8, but only f/4.5-5.6 70-300mm and now 100-400mm lenses (see below). Sorry Neal (Manowitz, VP of digital imaging at Sony), but I can still “follow and capture the action in ways” with a DSLR that your camera can’t. Make some lenses and your statement has more weight.

So despite all the buzz from Sony that the A9 is a sports-shooter camera, we need to verify that it really does have the necessary focus performance, and Sony needs more lenses for that market ASAP.

I find plenty that’s very appealing in the A9 specs. But I have to wonder from all the footnotes and all the still small-and-hard-to-find buttons whether Sony has fully dialed in a DSLR-killer yet. Testing will tell. Until then, we can all just drool over the plethora of interesting spec points.

bythom sony 100-400

Besides the new camera, Sony also introduced the 100-400mm f/4.5-5.6 GM OSS lens, which covers the full and APS-C frame sizes. While Sony hasn’t yet made a super exotic for the FE mount, this lens is probably right beneath such an effort, being a G Master design with all of Sony’s known flourishes. If the hypothetical MTF charts are predictive of actual performance, this should be a very good lens optically, especially at 400mm.

The 100-400mm lens should be available in June for US$2500.

Support this site by purchasing at the following advertiser:


Mirrorless Nibbles Away at DSLRs

We're getting near an interesting crossroad.

In the ILC market, mirrorless as a percentage of all ILC cameras has slowly been climbing. For 2013 through 2016 the numbers go 19.3%, 23.8%, 25.6%, 27.2%. I'll bet that mirrorless will cross the 30% threshold this year (2017; for the first two months of 2017 we're at 37%, but that probably won't hold).

bythom cipa feb17


Some of that crossover is inevitable. The DSLR duopoly of Canon/Nikon has something like 20 total DSLR products on the market—a bit more if you count some of the previous generations still hanging out in inventory—but the mirrorless world is generating 20 new cameras a year. The likelihood that Canon and Nikon introduce super interesting new DSLRs in any given year is less than the likelihood of a super interesting new mirrorless product being announced. So the basic marketing message is stacked against DSLRs.

But frankly, what more would you want from a DSLR that Canon and Nikon haven't already provided? DSLRs are extremely mature products with very high quality and performance capabilities coupled with near exhaustive feature sets. Sure, we'll see iterations of these products that push them faster, with more pixels, and with even more features. But the real question is how many people would actually find those things useful?

Selling a DSLR today is a marketing problem. I'll give you an example from a different market: I have a very capable 10-year old vehicle that's got fairly low mileage on it and is in excellent shape. It suffices for all the things I need a vehicle for. Why would I want a new one? Well, the auto maker's marketing would have to somehow convince me that there's something missing in my current vehicle that I actually need or would find useful.

Not that those things don't exist. It would be nice to have AirPlay in my vehicle—though I can add that through third-party electronics—and it would be nice to have some of the newest safety changes, particularly the auto braking, lane change warning, blind spot warning, and better back-up systems in the latest vehicles. I'd love to have a hybrid vehicle, too. But have those things risen to a high enough level to get me to abandon a perfectly fine vehicle and pay lots of money for a new one?

That's the DSLR problem in a nutshell. Do I need something better than a D810? Probably not, though I sometimes use a Sony A7rII to have a smaller-than-Nikon-DSLR kit (and as I've noted, it's really the competent and reasonable-sized f/4 zoom set that makes for most of the "smaller" bit), and I'm intrigued-but-not-convinced by the medium format cameras that are appearing.

Mirrorless does add some things that DSLRs are a bit behind the times on. The use of an EVF, for example, provides a more what-you-see-is-what-you-get view of the world on most mirrorless cameras, we also get real-time focus peaking info and exposure histograms, too. All three of those things can be very useful to someone that isn't using their cameras all the time, as they're helpful feedback mechanisms that are faster than "shoot the image, review the image, change the settings, repeat until correct."

But the real issue for Canon and Nikon is this: ultimately it's cheaper to make a mirrorless camera than a DSLR of equal capability. There's no doubt in my mind that both companies will push harder into mirrorless sometime soon. Indeed, the EOS M5/M6 appears to be the first of Canon's truly serious attempt to get a mirrorless product right. Nikon can't be too far behind, though let's hope that we don't see EOS M and M3 type experiments first.

I used the word "crossroad" above and I mean it. It won't happen this year or probably not even next, but I'd guess that mirrorless will have a greater volume than DSLR by 2019 or 2020, somewhere in that range. The exact timeframe will depend upon when Canon and Nikon really ramp seriously in mirrorless and how fast they bring their mirrorless offerings up to DSLR-levels of performance and features.

Don't get me wrong. DSLRs won't go away, particularly at the high end (basically 77D and up for Canon, D7200 and up for Nikon, and especially full frame). Anything you can do in a mirrorless camera you can add to a DSLR, so the claims that mirrorless will someday do things that DSLRs can't is, I think, mostly incorrect. To some degree, Sony's SLTs have already proven that. I'm thinking more along the lines of "hybrid," where the viewfinder itself can flip between optical and EVF. I wouldn't be surprised if the 1DxIII and D6 we get for the 2020 Tokyo Olympics are just that: have it both ways.

So the crossroad is this: ILC slowly becomes mirrorless as mainstream, DSLR as smaller high-end niche.

But the road ahead is narrower.

ILC volume shows no real sign of stabilizing: the numbers go 20m, 17m, 13.8m, 13m, 11.6m. They might hold at around the 11m mark this year partly because of all the delayed entries we didn't get in 2016 due to the sensor shortage. I'd also tend to characterize a fair amount of the volume we're getting now as simple customer iteration coupled with sampling. The customer iteration cycle—the time between when they update a current camera to a new one—is growing; currently somewhere above two camera generations and getting larger. The current customer sampling is a bit higher than it will eventually be as things eventually settle down and people pick their horses for the next race. I would not besurprised to find that at the point where we finally hit the crossroad, the overall ILC volume has dropped to 9m units a year.

At current market shares, that means:

  • Canon — 4.5m units a year (currently somewhere around 5.5m)
  • Nikon — 2.4m units a year (currently somewhere around 3m)
  • Sony — 1.3m units a year (currently somewhere around 1.6m)
  • Everyone else — 800k units a year (Olympus alone claims currently at 400-500k)

Given the market dynamics I mention—switch of mirrorless/DSLR dominance, smaller market—Canon's switch to automated manufacturing and now seemingly competent mirrorless camera offerings means they'll be aggressive at keeping that 50% market share. They have the cost factors already in place to do so, they now only need to show that they know how to expand their mirrorless line correctly and usefully, while protecting themselves in the declining DSLR market.

  • What Canon needs: more models, more lenses

Sony seems to have taken a slightly different approach in the past few years, moving further upscale and less worried about volume. Thus, their market share hasn't really budged much from where they started in ILC, though their profitability certainly has risen. Moreover, Sony still cameras share the E/FE mount with Sony video cameras (as well as some sensors), so they're getting bigger returns for their R&D bucks. Smart move, overall, though they really would like to get a larger piece of the market.

  • What Sony needs: a model that really breaks through and increases market share (I'd argue that is likely to be APS-C, and to do so they'll need more and better E-mount lenses, and they need to keep the price from creeping up too much)

The remainder (less Nikon, which I'll get to last) of the players are in a very tight squeeze. While m4/3 pioneered mirrorless and was the first to build full systems of cameras and lenses, the m4/3 market share has been in decline.Olympus just can't seem to budge above 500k units/year, and the financial/organizational situation there probably has them now entrenched at or near that level. Panasonic just moved the consumer camera group to the Appliances division in an apparent though unspecified as of yet cost-cutting move, which seems to indicate that they, too, are in a financial/organizational situation that's not going to net them growth.

Fujifilm is growing, but they started late and from zero; their growth has someturbulencethey're likely to hit soon. Still, of the mirrorless players, they're clearly the one making good decisions that are leading to increased sales. Because of that, they're dangerous. They could break out from the single-digit market share to double-digits, and that will mostly come at the expense of Canon, Nikon, and/or Sony.

  • What the other players need: a growing market instead of a shrinking one

Finally, we get to the wild card in ILC, and that's Nikon. Nikon is the canary in the coal mine. They're a company that was part of a strong camera duopoly (ortriopoly if we include compact cameras) that is totally exposed to the market as a company. More than half their revenues come from cameras (it peaked near 75% of their revenues, currently around 60%), and virtually all of their profit for a number of years now has come from cameras. If Nikon trips up oncameras, Nikon is at high risk of total business failure.

I'd argue, however, that even if Nikon trips up on cameras, that really doesn't present a lot of long-term benefit to the other players, nor does it really change the dynamics of the industry. It would take a bit of pressure of the declining market off the other players, sure. But consider this: If Nikon were to go "poof" today and disappear completely, just redistributing their market share would mean we would have Canon at 63%, Sony at 18%, the rest with 9%. So all that happens in that disaster scenario is that the duopoly becomes Canon/Sony instead of Canon/Nikon, and that Canon becomes more of the strong-arm.

So critical to the coming crossroads is one thing, and one thing only: what Nikon does in mirrorless. If they botch that, Nikon will shrink as a player and end up, best case, as a far smaller player making just niche DSLRs that continue to appeal only to all those owning those 100m Nikkor lenses in the wild. Worst case, they get absorbed by someone else, as Pentax was in the late film, early DSLR era.

  • So what does Nikon need? A mirrorless hit, basically. It needs to be priced right (<US$1000), it needs to be spec'd right, it needs to have high quality and performance parameters, and it needs—do I hear a buzz?—a real lens set. I'd argue that it needs more than all that, too. It needs to be a camera that can tie in to social networking easily. Easier than SnapBridge ;~).

Meanwhile, Nikon has another problem: what to do about the mirrorless customers they already have. That's not a small group, though it's not a particularly active buying group these days. Still, the J5 and V3 haven't been iterated in two years, no new lenses have appeared, and Nikon has been totally silent in marketing, too, leaving the two-year old marketing messages pretty much alone while the market changes around them.

If Nikon just says bye bye to the Nikon 1, this gets added to them saying bye bye to the DL models, and who knows what else (cough: KeyMission)? Which means that people aren't going to be terribly interested in what Nikon comes up with next outside of DSLRs because it has a higher likelihood of being abandoned.

All that said, Nikon is still the one to watch. The DSLR era got kicked off by Nikon in a big surprise to everyone else with the D1, and that came at the end of period where Nikon's market share was slipping significantly and they were perceived as being "behind." I have no doubt that Nikon has technological prowess to pull off another surprise. The question is do they have the courage to take that risk again?

But if Nikon doesn't come back into the mirrorless market with guns blazing, just who is going to set the ILC world on fire again? Canon's taking the iterative, protective approach. Olympus/Panasonic made an early impact to get things kicked off, but have lost traction, despite some engineering marvels. Medium format is just too expensive to resurrect ILC cameras, and to a lessor degree, so is Sony FE.

So say what you want about Nikon. We should all want them to succeed in the way they succeeded at the end of the film era: with a surprise that kicks off a new period of growth. Or at least manages to halt the decline.


2017 News/Views


2016 News/Views


2015 News/Views


2014 News/Views


2013 News/Views

The news and views for 2013 by month from sansmirror.com:


2012 News/Views

The monthly news and views for 2012 from sansmirror.com:

2011 News/Views

Sansmirror was started in October 2011 as a spin-out of bythom.com. Here are the 31 News/Views stories from the original three-month period:


text and images 2017Thom Hogan
portions Copyright 1999-2016 Thom Hogan-- All Rights Reserved
Follow us on Twitter: @bythom, hashtags #bythom, #sansmirror