Nikon recently released the first development announcement for their new flagship mirrorless Z9. We don’t know much about it officially, and the rumor mill has somewhat more information (which may or may not be accurate). There is also one official image of the camera, with a disclaimer that its appearance may change prior to launch. It looks like a D5 or D6, with a smaller, differently shaped prism hump (of course, on the mirrorless Z9, that’s not a prism). One little-noticed feature is that, unless the basic body style changes, the Z9 is substantially smaller than a D6 or any other dual-gripped camera. Nikon has not published dimensions, but the 24-70mm f2.8 Nikkor mounted on the camera is a released lens of known size, and it actually has the filter diameter (82mm) engraved visibly on the front of the lens. Given that the lens is 82mm in front diameter, the height and width of the camera are each slightly over 130mm. Since the front of the lens is somewhat closer to the viewer than the camera body, it’s not exact – the real number could be anywhere between 130 and 140mm, and the height (including the prism) is very close to the width.
A D6 is 160x163mm, a 1Dx III is 158x168mm, a GFX 100 is 156x144mm and even an Olympus E-M1x with a much smaller sensor is 144×147 mm. A D850 is 146x124mm and a 5D mk IV is 151x116mm. Yes, this is a dual-grip camera, but it’s closer in size to a midsize DSLR than to most of the dual-grip monsters – it’s about an inch smaller in each dimension than a D6. If it’s at the bottom of the possible size range, it could actually be no wider than a Z7, with the grip added on to the bottom. On the one hand, it’s likely to be a substantially less unwieldy camera than a D6 – assuming it’s not thicker than the D6 (and the short flange distance suggests it’s actually likely to be thinner), it’s probably quite a bit lighter. On the other hand, we’ll have to wait for handling reports to see whether the grips are cramped for people with larger hands – since it’s a dual-grip design, you can’t curl an extra finger around the bottom of the body. As far as the actual specifications go, Nikon has only released two details. One is that it shoots some form of 8K video, and the other is that it uses a stacked sensor, which implies very high readout speeds. Since it’s an 8K camera, the width of the sensor is a minimum of 7680 pixels, assuming the minimum, Quad-UHD definition of 8K. If it shoots true cinematic 8K, the sensor is at least 8192 pixels wide. In either case, the sensor could be wider, but not narrower… The sensor aspect ratio is 3:2, since the only interchangeable-lens cameras that aren’t 3:2 are Micro 4/3 (4:3), most medium format cameras (usually 4:3, occasionally 5:4) and some dedicated movie cameras (most commonly 16:9, with other possibilities)), a 7680 pixel sensor would be just under 40 MP, and an 8192 pixel sensor would be just under 45 MP.
Rumored resolutions for the Z9 include ~45 MP, which would very likely be the Z7 resolution of 8256×5504 pixels – the Z9 sensor might be a stacked variant of Nikon’s longtime high-resolution sensor that goes back to the D850. If it were to be slightly higher resolution, it could be a variant of the stacked sensor in Sony’s A1 (8640×5760 pixels), or it could just BE the A1 sensor, assuming Sony would sell that sensor. Which of those resolutions it is makes absolutely no difference – 400 pixels on a ~8000 pixel wide image is immaterial. The Z7 sensor is an excellent performer, and the initial reports on the A1 sensor, which I haven’t used, are that it is also an excellent performer. There are also rumors floating around of a ~60 MP sensor for the Z9, which strikes me as unlikely. The most logical way to get there would be a stacked variant on the A7r IV sensor – but if Sony were going to do that, why not use it on the A1 as well? They used a new sensor, with a new resolution, for the A1, which implies that there is something they didn’t like about stacking memory behind the A7r IV sensor – perhaps just that the design didn’t lend itself to the modification?
Nikon has used three kinds of sensors over the years, and they often claim internal designs, even when that is only partially true, or occasionally almost entirely false, All of Nikon, Canon and Sony have sensor design teams – Nikon and Canon even go a step further back in the process and build the lithography machines that make the sensors (although Nikon rarely, if ever uses the machines they build to fabricate their own sensors – a Nikon-designed sensor may be made on a Nikon stepper, but at someone else’s factory). No camera manufacturer besides those three actually designs sensors anymore – Fujifilm used to, and they still specify highly customized sensor “toppings” – their X-Trans color filter array goes on a standard (and excellent) Sony sensor.
The first category of sensors Nikon has used is entirely standard Sony sensors. Plenty of Nikons have a sensor plucked from the Sony catalog, although with Nikon processing and color science that can make it look quite different from the same sensor in a Sony or Pentax camera. There’s not only nothing wrong with Sony sensors, there’s often quite a lot right with them – and Nikon tends to implement them well. The D800/D800e sensor is probably exactly the same unit used in the original Sony A7r and the Pentax K1, and there have been quite a few Nikons using off the shelf 24 MP Sony sensors (both APS-C and full-frame). They have largely been great cameras, and the ones that have had significant flaws have tended to be related to cost-cutting, not image quality. If the Z9 sensor is “just” the A1 sensor, all the early reviews are showing that this is nothing at all to be disappointed about – it’s a great sensor… If a Z9 is an A1 in a dual-grip body, taking Nikon lenses, and with Nikon color science, that is exactly the camera a lot of Nikon lens owning pros are waiting for to retire their D5 (or even D4s) or relegate it to being a secondary body. A lot of Sony-lensed pros will buy an A1 (or several), and they’ll be just as happy with those.
The second category of Nikon sensors is “Nikonized” Sony sensors. Nikon sometimes takes a Sony sensor and makes a number of modifications to it. Sometimes, the modifications are just in the toppings – the D3x used the original Sony 24 MP FF sensor from the A900, but it had a somewhat unusual set of filters on it, which may have contributed to the D3x’s high cost, but there was a noticeable difference in image quality in favor of the Nikon. The D810 used a close relative of the common Sony 36 MP sensor, but with a lower gain to allow ISO 64 – is this actually slightly different silicon, or the same sensor with different driver electronics? The most significantly Nikonized of recent Sony-derived sensors is the 45.7 MP, ISO 64 sensor in the D850 and Z7. It’s a tiny (and generally unnoticeable) bit higher-resolution than the 42.4 MP sensor Sony has used in a bunch of their own products, and the more noticeable difference is that it supports ISO 64, offering some extra low-ISO dynamic range over the more common 42.4 MP sensor. It’s (arguably) half a generation newer than the 42.4 MP sensor, and is a slightly better performer. Of course, the 42.4 MP sensor is no longer Sony’s flagship – the 61 MP A7r IV sensor is (or the 50 MP A1 sensor).
Occasionally, Nikon will strike out on their own with a unique sensor design that doesn’t seem to be related to anything else. The 20 MP APS-C sensor first seen in the D500 is an example of this. Nobody else has released a sensor of the same resolution, and it took a long time before we saw another APS-C sensor as fast as that one. Whether or not there is Sony (or anyone else’s) technology in it, this is a significantly custom Nikon sensor, probably designed at Nikon. They’ve used it (or a variant) several times since the D500, and it’s a very capable sensor.
In the end, it almost doesn’t matter which category the Z9 sensor falls in. The A1 sensor seems to be an excellent performer, and Nikon would certainly tune the color science to Nikonish tastes. If they “Nikonized” it by more than that, it would probably be an improvement, although it could also mean a more expensive camera. Many unique Nikon sensor designs have also been excellent performers. Even the much maligned (for low resolution) D2h was actually a very good performer for its day. It had REALLY nice pixels, although it didn’t have a ton of them…
Beyond the sensor, we’ll have to see – nobody’s handled a Z9 yet. It’s rumored to use the largest, highest-performance battery we’ve yet seen on a mirrorless camera, which could mean that it’s VERY fast. Most mirrorless cameras (and DSLRs) use a battery that is nominally in the 7.6-7.8 volt range (when fully charged, they read around 8.2-8.4 volts). A few small cameras use 3.6-3.9 volt batteries instead. The standard batteries use two lithium-ion or lithium-polymer cells (the minor differences in voltage are due to the precise chemistry of the cells – there are a few different formulae, and the voltages are similar but not identical). The lower voltages (usually found in phones and compact cameras, but occasionally in small interchangeable-lens cameras) are single cells. Triple-cell batteries nominally around 11 volts show up in Nikon’s and Canon’s big sports DSLRs, in auxiliary battery packs for a few other DSLRs (especially from Nikon), and in Fujifilm’s medium-format line. There is speculation that the Z9 will use Nikon’s EN-EL 18 series triple-cell batteries – if it does, it will be the first mirrorless camera outside of medium format to run at the higher voltage. A higher-voltage battery would allow for larger, faster motors in the body, and potentially for driving lens-based AF motors faster (I don’t know if lenses, especially the big telephotos where it would be the most important, can accept multiple voltages).
Nikons traditionally handle very well, but we won’t know anything about that for sure until reviewers handle the camera. Similarly, there are very good reasons to expect that focus performance will be very good, and quite possibly better than that – but there’s no way to KNOW that without seeing the camera. Image quality, too, should be superb, but we won’t know until we see it. There are a lot of reasons to be optimistic about the Z9, but very little real knowledge.
I’ve covered Sigma’s new fp L in more detail in the companion article dedicated to Sigma, but it’s worth a paragraph here. Basically, it’s the A7r IV sensor in the Sigma fp body. It has the strange compromises of the Sigma fp – it’s a viewfinderless, electronic shutter only camera with no image stabilization that uses a tiny battery and has no grip. By the time you add an accessory finder (yay, Sigma did release an electronic viewfinder this time) and a grip, it’s not much smaller and lighter than a more versatile camera. What if one could learn to shoot with a camera as bare-bones as a base fp L, though? It’s more than 1/3 lighter than anything else with close to the same potential image quality… It’s built like a tank, and its very simplicity may mean there’s less to go wrong in the field. Can it go farther into the wilderness than anything else that can bring back the same images?
The other interesting recent release, and a camera I’m actually itching to try out for landscape use, including the significant percentage of my work done while backpacking, may, at first, seem a strange choice for lightweight landscape… It’s the GFX 100S. Yes, it means no zoom lenses (too heavy), and only a couple of the primes are light enough. On the other hand, it’s a heck of a lot lighter than backpacking with an 8×10” view camera, and that is what it might offer the image quality of! It looks like there is really very little to work around apart from the weight. It has excellent IBIS, a well thought out battery system, and everything is very well sealed. The 100S and the tiny 50mm prime are well within what I’d consider a backpackable combination, and the 30mm may be as well. I have a request in to Fujifilm for a tester in the couple months before I set out – we’ll see what they say! Am I really crazy enough to hike several hundred miles on the Pacific Crest Trail carrying a medium format camera? If I am, it means the GFX 100S is a pretty unusual camera… Much of that terrain is the kind of place where 40×60” prints are real possibilities, and the best sensor in the world may be worth having. In most ways, the GFX 100S is a relatively poor performer. Set for absolute maximum quality (and how else would you use it), it’s a 3 fps camera. When did we last see one of those? Any $500 Rebel or D3500 can shoot faster. Its autofocus, while much faster than most medium-format cameras, is slow by most modern standards. It’s certainly, and irrefutably, a deeply specialized camera – but one of its specialties is landscape photography at extremely high image qualities. Not even the most dedicated photographer of the natural world will be able to make a GFX 100S his or her only camera – unless they never photograph an animal, and are content with only midrange focal lengths.
It is also a camera that cries out for print, and, specifically, for very large prints where the extreme detail it can produce is noticeable. If I have the opportunity to review one, I will be most interested to see whether 24×36” prints are sufficient to reveal the difference in detail between the full-frame pixel monsters and this most extreme of pixel monsters. It will also be interesting to see if there’s any difference at all in a 16×24” print. If you look at a really good print from large format – say an Ansel Adams original – there’s a “life” to the image that isn’t there in a print from a smaller piece of film. It operates right on the edge of human vision – it’s micro-detail that we can’t quite see, but we sort of perceive. There is a difference in life between a 24×36” print produced from 24 MP and 50-60 MP, and, under close inspection, that difference is there at 16×24”, although you don’t see it at a casual glance. At 300 dpi, the native print size of the GFX 100S is 30×40” (a tiny bit smaller, but so close that it’s probably best to print without reduction or enlargement and call it ~30×40). It is also capable of printing a native ~16×20” print at 600 dpi.
Is this the camera that the double-resolution modes on printers are designed for? Every printer review says that they offer little benefit (and the only time I’ve ever seen a benefit is with earlier printers where using 600 dpi input turned on other quality features) – but that has always involved using either small prints or upsized files. Will a native 600 dpi 16×20” print have some of that 8×10” film magic? Or will it be impossible to tell from any other very good print? I’ve never printed a file like that, so I don’t know… Will it require different sharpening and other handling from lower-resolution files? In one sense, it’s just an oversized version of the A7r IV sensor – the pixels are the same (although the color isn’t, since Fujifilm uses different filters and algorithms – even though the GFX line isn’t X-Trans). In another sense, it’s very unusual, in that you are almost always going to be downsizing the file to print – we’re used to upsizing. Will it make a difference? Where are the limits of our visual system?
Finally, we had a very odd camera introduction right as this was going to press… First of all, it’s a high-end APS-C DSLR (and we haven’t seen one of those from ANY manufacturer in a while), and secondly, it’s a Pentax! If you count the Canon EOS 90D as a high-end camera (it has the 7D mkII above it, so it’s arguable), we last saw a high-end APS-C DSLR introduced almost two years ago, as the Canon appeared in August of 2019. Nikon’s D7500 was released in August 2017, but was actually a step back from the older D7200 in some ways, while being an upgrade in others. The last unarguable high-end APS-C DSLR released was Nikon’s D500, way back in January of 2016! We haven’t seen ANY Pentax DSLR released since the K1 mk II, a minor upgrade to the original K1 in early 2018, and the most recent really new model was the KP in January of 2017. One could be forgiven for forgetting that Pentax even makes cameras. There had long been rumors of a new flagship APS-C model from Pentax, but they had been balanced by rumors that Pentax was quitting the camera business altogether. In the end of March, 2021, Pentax finally introduced the K3 mk III, their new flagship APS-C model. Why release a $2000 APS-C DSLR in 2021, especially one with a rarely seen lens mount? It’s the most expensive APS-C DSLR on the market, and tied with the special finish versions of Fujifilm’s X-Pro 3 for the most expensive mainstream APS-C camera of all (not counting Leicas, because not only are they very expensive, but collector versions distort Leica pricing). It’s the first DSLR we’ve seen with the newest Sony 26 MP sensor that Fujifilm has been using lately. That’s a great sensor – arguably the best APS-C sensor on the market, with the only competitor being the 32 MP Canon sensor – the 26 MP is clearly the best yet in the long line of ~24 MP Sony sensors. In all other specs, it’s extremely capable – somewhere in between the Nikon D500 and what the D500 would be if it were revised in 2021. It’s a higher end camera than the Canon 90D, because the 90D is actually one model from the top of Canon’s APS-C range – but the 7D mk II is ancient. Of course, it had better be incredibly capable, for what Pentax is charging.
Pentax has an unusual place in the market – sales of new Pentax equipment are tiny – the best estimate I’ve seen is 40,000 bodies annually, while Fujifilm sells about half a million. Nikon and Sony both sell around a million, and Canon sells just over 2.5 million bodies annually. Pentax sales are almost certainly smaller than Leica even in terms of cameras and lenses per year (and a small fraction of the value), They are probably selling fewer cameras per year than Fujifilm sells medium format GFXs alone. On the other hand, the supply of vintage Pentax lenses is huge – behind only Canon, Nikon and arguably Minolta/Sony A mount (not E mount).
The problem for the K3 mk III is that those vintage lenses tend to be full frame, and Pentax’s own full frame K1 mk II is actually cheaper than the list price of the K3 mk III, while being a more desirable body for most vintage lenses because you don’t have to worry about the crop factor. There seem to be only two situations where the K3 mk III might make sense. One is the rare photographer with a collection of great Pentax APS-C lenses. If you’ve been shooting Pentax APS-C for ages, you just might have two or more valuable lenses that only work on a Pentax APS-C body (if it’s one lens, it’s probably worth switching unless you LOVE that lens). The second situation is if you have a long Pentax telephoto and prefer APS-C to extend its reach. The sensor in the K3 mk III is almost twice as dense as the one in the K1 mk II.
For almost everyone else, there’s a better option for less money. If you have, or want to work with vintage Pentax glass, a K1 mk II or perhaps a lens adapter (there’s a real market for a really good Pentax adapter, probably going to FE-mount bodies) is probably a better bet. If you want a ~24 MP camera and are open to various manufacturers, there are more mainstream choices from Canon, Nikon, Sony and Fujifilm. Unless you really want a DSLR, a Fujifilm X-T4 with the same sensor as the new Pentax offers a great selection of current, APS-C dedicated glass. If you aren’t actively trying to avoid full-frame (perhaps to escape bulky lenses), there are various $2000 full-frame options from Sony, Nikon and Canon. With careful lens selection, a full-frame outfit with a couple of lenses might actually be lighter than a K3 mk III kit. If you are set on both APS-C and a DSLR, a Canon 90D or a Nikon D500 are generally far better choices due to support and selection of modern lenses.
If the K3 mk III were a $1300-$1500 camera (and, to be fair, maybe that’s where it’ll settle), it would have a market. Built like a tank, great SLR viewfinder, encourages experimentation with vintage glass. At $2000, it faces too much competition – from its own K1 mk II stablemate, from the X-T4, from the Canon 90D and the elderly Nikon D500, and from all manner of full-frame cameras in the same price range.
Where is the market now, as we can see the vaccinated light at the end of the COVID tunnel? Canon’s huge market share is due largely to Rebels and EOS-M bodies that there are few decent lenses for. Both Rebels and EOS-M are highly vulnerable to smartphones – but EOS-R has never looked stronger. The EOS-M system continues to look crazy – no growth path to EOS-R, and no lenses that appeal to more serious photographers. There are consistent rumors of a higher end APS-C mirrorless camera with an RF mount instead of EF-M. It actually makes sense Canon is slowly discontinuing DSLRs and lenses, and they are unlikely to ever update another DSLR – unless it’s a cost-cutting update to a Rebel. O
n the other hand, they’re introducing RF lenses like crazy, and the reviews generally look great. Canon is more vulnerable than anyone to further losses in volume, and they know it – but they have a gigantic cushion, and they’ll lose much more volume than they will profit, because the volume will come from the bottom of their lineup, while the EOS-R business continues to grow as more and more loyal Canon users transition from their 5D mk IIIs and IVs to EOS R5s… We’re almost certainly going to see at least one new high-end EOS-R body in the next year, and possibly two. The obvious one is a competitor for the Sony A1 and upcoming Nikon Z9 – fast, high resolution, powerful and expensive. Will it show up at the Tokyo Olympics? What will Canon do with the body design? Sony chose to make the A1 petite and let photographers choose whether or not to add a grip, while Nikon seems to have built the grip in…
The second subject of persistent rumors is an ultra-high resolution body – somewhere between 80 and 100+ MP. Can it be done? If it can be, will the resolution improvement be worth any tradeoffs in noise, color or other image characteristics? 100 MP medium-format files are excellent, but those pixels are small, but not tiny – the 100 MP medium-format sensor is just a physically larger version of the A7r IV sensor. An 80+ MP sensor without increasing the physical size of the sensor will have tiny pixels – in the size range of Micro 4/3’s Same Old Sensor, or even smaller. If that’s the noise and dynamic range performance, the resolution isn’t worth it for most uses. Of course, the Same Old Sensor is also old – a modern sensor with the same pixel pitch might well outperform it. Canon’s 32 MP APS-C sensor is in the same pixel density range as the ultra-high resolution sensor would have to be, and it’s a better performer than the Same Old Sensor – but not as good a performer as the Sony 26 MP sensor in newer Fujifilm cameras…
The sensor Canon really needs is the one for the “EOS-R1”. Sony (and soon, Nikon) have set the bar at around 50 MP, with 8K video and super-fast stills performance. The wildcard here is that the EOS-R5 sensor in a bigger or better cooled body could BE the EOS-R1 sensor. We don’t know just what it would take to make the R5 perform like an A1 or what a Z9 might be – the specs are actually quite similar, but Canon has had overheating issues. Is there a bulky, or just expensive way to get full performance out of what they’ve got? If they can solve the heat management with either the R5 sensor or a new one, probably a variant on what’s in the R5, I have a huge amount of confidence that Canon can get a full-blown pro RF body right… They’ve been making top pro DSLRs (and film SLRs before that) for decades. They have great reduced size raw technology, which will be critical on this camera – newspaper sports photographers don’t want to be shooting 50 MP all the time!
One of Canon’s great strengths is that the adapters from EF lenses to EOS-R bodies are excellent (and there are no old screw-drive lenses to worry about). Photographers who’ve been shooting Canon for years have every reason to continue to do so, and the EOS-R system is a great choice for people growing from a phone. Canon’s problem is that they’re really offering a jump straight to full-frame – their APS-C offerings are either DSLRs they’re showing little interest in or else EOS-M bodies with only cheap lens options. Here’s where that APS-C body with an RF mount comes in… Once you hit full-frame, their line is very robust with a great choice of bodies and lenses – the only one they really need to add is the R1.
Nikon has already lost much of the volume Canon has yet to lose – they have been hemorrhaging low-end DSLR sales for a while now. What they’re left with is a million cameras a year of mostly serious volume in the midrange and above. A lot of Nikon’s remaining sales are relatively expensive cameras, bought by people who won’t give it up for a phone… How many are D3500s and such that are still vulnerable, though? Nikon has very little strategy below full-frame – the Z50 is a quietly excellent camera, but there are only a couple of dedicated lenses, and, while it also takes full-frame Z lenses, neither the body nor the lenses is stabilized in that pairing (FF Z cameras all have IBIS, so only a couple of the lenses have lens-based stabilization). They have a couple of nice APS-C DSLRs, but they are getting old – will we ever see updates? Nikon’s actually in better shape than Canon in APS-C mirrorless (in the West – EF-M is more popular in Asia) – at least the Z50 offers a growth path…
Nikon seems to be more interested in their remaining DSLRs than Canon is. They aren’t discontinuing bodies or lenses, and there are persistent rumors of a couple of new DSLRs. The obvious one is a D850 upgrade. Minimally, they would do exactly what they did to go from the D750 to the D780 – add the Z-series features, many of them only working in live view. They could go farther, in either of two directions. One option would be a true hybrid – they’d have to figure out a hybrid viewfinder that could work as either an optical prism finder or an EVF. Fujifilm has done it with the X-Pro series, but that’s not a reflex prism… Is it possible on a DSLR? Is it worth it? Could they figure out some way to stow the mirror out of the way such that it could have a hybrid F/Z mount? Of course it would have to be EVF only with Z lenses – they couldn’t clear the mirror. A physical piece would have to come on and off to make the conversion. Again, is it possible? Worth it?
The second option to go farther is to upgrade the sensor… To what? A Nikonized version of the A7r IV sensor? Is that worth doing? It would be an improvement (assuming it kept ISO 64), but is it a big enough improvement to be worth the effort? A really high-MP sensor? Same question as for Canon – is it possible, and are the IQ benefits there? I trust Nikon (really Sony, who would almost certainly make the base sensor) more than Canon on this, simply because of their track record with high-resolution sensors. From the D3x sensor in 2009 to the present, Sony sensors, which have often found their way into Nikon bodies as well, have been the best there is. If anybody can do it, Sony can, and Nikon is a likely launch customer. If that sort of sensor shows up, expect it in a Nikon DSLR, a Z body and an A7r body, all at around the same time – the Sony body may use a slightly different version of the sensor, but the underlying technology will be the same. If any one of the three shows up, it’s a decent guide to what the other two will have within the year.
The second DSLR they could update is one last iteration on a top APS-C model. It would probably land between the D7500 and D500 in specifications, updated to the present day, and replace both. It would probably be joined by a “Z70” with the same sensor and very similar features, either simultaneously or within a few months either way. The question to ask here is “why not just buy a Z70 with the FTZ”? Choosing the mirrorless body offers both F and Z lens compatibility, and the Z70 will probably have an excellent viewfinder. There may be enough DSLR enthusiasts, especially (often older) photographers who’ve decided not to move to mirrorless to make it worthwhile – but remember that the D550 is a last camera, and the Z70 isn’t. When the inevitable Z90 appears, there won’t be a D580 to join it. My best guess (and I could be wrong) is that we won’t see a D3600 or a D5700. That’s the Z30’s job – if we do see a last low-end DSLR, look for a more or less pure cost-cut. Most owners of the little DSLRs have only one or two lenses, bought in a kit with the camera – sell the lenses with the body, because the Z body will come with native equivalents (which already exist). For those who have more, there’s the FTZ.
Once the Z9 arrives, the full-frame Z body lineup is pretty complete. They might slip in a Z8 with a really high resolution sensor, or that might be the job of the Z7 and successors. There’s room below the Z5, but not much – they can’t really delete IBIS, because too few of the lenses have VR – and stabilization is most important at the lower end. The lens roadmap looks very solid (as do Canon’s, Sony’s and Fujifilm’s), although not all of it’s out yet, and they’ve had some supply issues.
Sony is in great shape – if you start off with a 24 MP A7 III or A7C, you have three (and a half) ways to go for your next camera. You could go for performance, either keeping the same resolution with an A9 series body or going for performance AND resolution with the A1. You could explore video with an A7s III, an FX 3 or a higher-end FX camera. You could choose ultimate resolution in the A7r series. Whichever way you go (and Sony’s hoping you’ll choose more than one), your lenses follow you. Other than the FX series, they all operate very similarly. Sonys can be quirky, but all full-frame Sonys are quirky in the same ways. Their lens lineup covers just about everything (yet again, where’s that 300mm f2.8 G-Master, Sony?), with a lot of multiple options and some nice telephoto choices where other mirrorless lineups tend to be weaker or rely on adapters. The best of them are really excellent, and there are an increasing number of great choices – many older, weaker lenses now have a newer option either replacing them or as an alternative. The glaring weak lens (again, I’ve called this out before) is the 24-70mm f4 “Zeiss”. There is nothing newer that replaces or augments it, since both the 24-70mm f2.8 G-Master and the 24-105mm f4 G are substantially bigger and heavier. Please, Sony, come out with a 24-70mm f4 G that’s at least as good as the Nikkor, and no heavier!
Being a division of the same company that makes the best sensors around is a nice advantage for a camera company! Sony will never lack for sensors, and there’s only one that they really need (and another in the “wouldn’t it be nice” category). The real need is a replacement for the 24 MP sensor in the A7 III, A7C and many predecessors. That’s a great sensor, but it’s also been around forever. An A7 IV with a new sensor and the new menus would cause a lot of folks who are presently using older 24 MP A7 bodies of various sorts to upgrade. Many of them would also get the pleasant surprise of just how nice a battery the NP-FZ100 is – they’ve been using FW50s. They’d get current-generation Sony AF – e.g. some of the best AF in the business. The new sensor is almost secondary, getting people to upgrade old cameras where it’s really other features that matter. The difference between an original, slow-focusing, mini-battery A7 and a current Sony is night and day, but people see “24 MP” and figure it’s the same.
The “wouldn’t it be nice” sensor is an ultra-high resolution model. Will it make a difference? Only time (and tests with GFXs) will tell. The A7r IV sensor is very, very, very good at high resolutions – will something in the 80-100 MP range be any better? Are our printers that good? Is our technique that good? Are our eyes that good? I accidentally made a 24×36” print from an A7r IV file that was shot at ISO 6400 recently (I thought it was 1600). Using DxO’s DeepPRIME noise reduction, the file looked terrific, so I sent it to the printer. The print looks great – you can tell it’s high-ISO if you get up close to it, but not easily on the wall. It was only after it came rolling off the printer that I looked at the metadata and realized how high the ISO was. It looks about as good as a 24 MP APS-C print (from an X-T2) of the same size at base ISO (200). Part of that is DeepPRIME – since DxO is allergic to X-Trans, I haven’t had a chance to try it on my Fujifilm files.
If our present top sensors are THAT capable, what more will an even better sensor bring us? What I’m looking for, if anything, is that feeling of detail you can almost, but not quite, see that you get from a print from large format. The ISO 6400 print doesn’t have it, and neither do big prints from APS-C. A 24×36” print from a Z7 at ISO 64 or an A7r IV at ISO 100 does – would a hypothetical higher-resolution camera have even more?
Sony’s APS-C line is a weakness, but of course (fill in manufacturer name, other than Fujifilm)’s APS-C line is a weakness. A lot of very similar bodies with the same sensor, and a lot of undistinguished lenses with a few better options mixed in. By the time you get high enough in the line to get some highly desirable features found on the a6600, notably IBIS and the FZ-100 battery, you are close to the price of an A7C, which not only adds better weather sealing, but is (more importantly) full-frame, allowing Sony’s better lenses to be used at their designed coverages. The much cheaper a6100 doesn’t stand up well to the similarly priced Fujifilm X-E4, largely due to lens selection. If you want to use Sony’s excellent full-frame lenses, it’s worth waiting for a sale on a Sony full-frame body to go with them. Whenever the A7 IV comes out will be a great time to buy an A7 III or A7C for a good price. The older A7 II has been as low as $1000 at times, and I would not be at all surprised to see the A7 III or A7C get down that low by the end of the year.
The fourth major player is Fujifilm, always marching to the beat of their own drum. While everyone else’s tune is “Full-frame is where it’s at”, Fujifilm is jumping straight from APS-C to small medium format. The huge advantage of Fujifilm’s approach is that they have a wonderful selection of APS-C lenses that are made (and sized) for APS-C. The disadvantage is that they really offer one sensor of interest to Luminous readers (they use some older APS-C sensors in entry-level cameras), other than one used in a $6000 camera and a $10,000 camera. I’m ignoring the 50 MP medium-format sensor because it offers so few advantages over the 45-60 MP class full-frame sensors. The gains from the larger sensor are given up because it’s a 2014 sensor… The 50 MP cameras offer very similar image quality to something like a Z7 or an A7r IV, but in a slightly larger, considerably slower body that takes considerably larger lenses. That’s a pretty niche market. The 100 MP sensor is in a different class, but that starts at $6000.
Fujifilm’s APS-C line, as experienced by serious photographers, uses the 26.1 MP Sony sensor with a lot of special Fujifilm sauce in the color rendering, most notably X-Trans color filtration. It’s probably the best sub full-frame sensor on the market today, and outside of Fujifilm, you’ll find it in one Pentax DSLR (Sony, inexplicably, doesn’t yet use it in their own APS-C line). It’s the smallest of a series of four sensors that all use the same 3.76 µm pixels, just more or less of them on different sized pieces of silicon. Its stablemates are the 61 MP A7r IV (and Sigma fp L) sensor, the 100 MP GFX 100/100S sensor and the 150 MP sensor used by the ~$40,000 Phase One IQ4 150. Theoretically, a perfect stitch of 6 images from an $849 X-E4 should be very, very similar to a single image from the IQ4 (color science is one thing that will diiffer).
The cheapest way to get into Fujifilm’s best sensor (outside of medium format) is through the $849 X-E4. It’s a pretty basic rangefinder-styled camera, but it has decent AF, an X-mount with a great selection of lenses on the front, and a great sensor in the back. For $999, an X-S10 trades in the retro rangefinder style for the controls of a tiny, but modern DSLR, and it adds image stabilization. A $1699 X-T4 offers “best of both worlds” controls, weather sealing and exceptional build quality, dual UHS-II card slots, a slew of features (if the X-T4 doesn’t have it, you probably don’t need it) and a nice big battery. It costs as much as a lower-midrange full-frame mirrorless, but it can do everything one can do, too. Image quality is VERY close to 24 MP full-frame. The only real drawback is that if you’re at 24 MP full-frame from Sony, Canon or Nikon, there are several places to go – you can pick a video-centric body, a pixel monster or a sports and action camera. The X-T4 is pretty darned good at all three, but it doesn’t have specialized stablemates.
Fujifilm’s APS-C lens lineup is excellent and comprehensive. Like Sony’s full-frame lineup, it has only a couple of holes. The big one is any prime longer than 300mm equivalent. While Sony is missing a 300mm f2.8, but DOES have a 400mm f2.8 and a 600mm f4, Fujifilm has the opposite problem. They have an excellent 200mm f2 (the APS-C equivalent of a 300mm f2.8), but nothing longer. They have a good selection of tele zooms, but just the one big, fast prime (and a couple of teleconverters for it). Apart from that, everything else missing is arguably even more exotic – there’s no fisheye (although there is an 8-16mm zoom if you want to get really wide), and there aren’t any tilt-shift lenses.
Unless you’re printing big, you don’t need more resolution. If you share your images digitally, the X-T4 (and really any other modern camera) will outperform not only your monitor, but everybody’s monitor who might see them. The X-T4 is very, very close in many aspects of performance to Apple’s $5000 XDR Display – the camera has slightly more resolution in the short dimension, they are nearly identical the long way, and the color gamut is such that the Apple display should just about display everything the X-T4 can capture. It will happily outperform anything less than a very elite monitor (or an 8K OLED TV). Digital capture has simply advanced to the point where we can’t see what our cameras can capture on most monitors, and, even if you’re looking at really elite monitors, most ways we share images won’t transmit the image at full resolution. All social media platforms, for example, compress images heavily, to the point where what you used to shoot it doesn’t matter. If you’re displaying on an elite monitor or an 8K OLED TV, sending the images from a computer (or possibly a memory key) connected directly to the screen, then the camera might matter, as it certainly does for printing. In just about any other situation, there is some step in the image transmission pipeline that converts to a poor-quality JPEG, and the differences between decent cameras are erased.
Fujifilm’s other line is the medium-format GFX system. You pay in weight and cost for the big sensor, especially in that lenses are significantly bulker than their full-frame equivalents. It’s probably not worth it for the 50 MP version most of the time. What’s more interesting is the 100 MP options – expensive, but the detail they can capture borders on the absurd. I had a GFX 100 in for review (see A Tale of Two Fujis), and the images were unbelievable, but the body was bulky and hard to use. The new GFX 100S looks like a big improvement, and I’d love to try one with some well-chosen lenses.
Beyond these four, I have questions about the viability of most other camera companies in the long term. Since I last looked at in the beginning of the year, L-mount looks a little better than it did, due to the S5 being a more conventional body than the big S1 series, and due to Sigma’s increasing support for the mount. I’d now put L-mount as more likely to survive than Micro 4/3 – it’s a full-frame mount looking for a niche, but it has Panasonic’s video expertise, Leica and Sigma’s tendency to play with odd bodies to support it. It actually has three fairly complete lens lines, with such oddities as three 50mm f1.4 lenses (and I wouldn’t be at all surprised to see a fourth in the next year or so when Sigma releases a mirrorless-specific DG DN Art lens to supplement the present stretched DSLR lens) What other mount can get you a full-on video rig, a full-frame camera that looks like a bar of soap, and a $6800 design exercise?
Micro 4/3 still has a number of problems, starting with the Same Old Sensor, and with the effective apertures of most lenses being really slow (because of the 2-stop difference in depth of field and the noisy sensor). A summary of Micro 4/3’s problems is that the E-M1x is actually heavier than the GFX100S… A 20 MP camera featuring a sensor that was a poor performer in 2016 is bigger and heavier than a camera with 6.4 times the sensor area, five times the pixels and three stops of additional dynamic range. Of course, nobody’s going to choose between those two cameras – one is made for action, and the other for ultimate image quality – but… The more “normal” E-M1 mk III is the size of a Sony A1 and only a little lighter (of course, it’s MUCH cheaper) – and those are legitimate competitors designed for the same type of photography. As other mounts come closer to the durability and image stabilization of the best Micro 4/3 bodies (and exceed the previously standout video performance) Micro 4/3 is stuck competing on price, and Fujifilm (among others) is a tough competitor with the higher-end bodies – much better image quality, similar or better performance, for about the same money with cheaper lenses when you compare on effective aperture.
Many lenses are very expensive for what they are, given the effective apertures – they like to emphasize “twice as long as it looks” in advertising, but not “also two stops slower than it looks”… Other mounts are getting better at telephoto, so reach is no longer a unique advantage, and cropping a pixel monster to increase reach looks better than the Same Old Sensor. They could make a really compact camera, but things like Fujifilm’s X-S10 or Sigma’s fp and fp L aren’t much bigger than small Micro 4/3 cameras, with huge differences in image quality. Micro 4/3 now REALLY needs a new sensor, and it has to be something the others can’t just use a larger version of (what if they got an organic sensor first, and it had problems scaling to larger sizes?).
Pentax, as mentioned above, is quixotic. The K3 mk III won’t help them unless it comes in much cheaper than its list price. Even if it does, the Pentax audience is limited to people with Pentax lenses (or who want to buy them used by scrounging around in odd places – many camera stores don’t carry as many because of the limited market) – there is no reason to buy into Pentax now. I continue to think that the best possible Pentax product is a first-party mount adapter to put all those lovely lenses on Sony FE (or something else, but FE seems a good match). They just have to be losing money, and a Pentax/Sony joint adapter would support everyone who has the lenses, while opening up a wider selection of bodies without the price premium of very limited production.
Leica will keep on being Leica – they’ll make M-mount cameras just as they always have, and Sigma and Panasonic may provide enough support to keep L-mount viable as well. Even if L-mount were to go away, M-mount is such a weird proposition that Leica will keep making it. A significant percentage of M bodies and lenses sold never leave their boxes – this is especially true of the large number of special editions made especially for collectors, but even many “standard” M-series cameras and lenses end up being collectible. There is no series production camera other than a Leica M that will reliably be worth more than you paid for it in a few years. There are certainly cameras where that’s true, but it would take luck – imagine if you had an original Nikon F still in the box – but you’d have to have realized the significance of the SLR 60 years ago.
In general, the outlook after CP+ and heading into what would be the summer travel season in most years looks similar to that at the beginning of the year. A couple of systems look a little better than they did. The GFX 100S is a more exciting camera than I thought it was going to be – Fujifilm got the size and weight down by more than I thought possible while keeping IBIS. It also came in at the bottom end of the price range I would have considered likely. I thought we might see a camera using the slightly tired 50 MP sensor, with IBIS, around $5000, and a bit bulkier than what we got and I thought we might see another 100 MP camera in the $7000-$9000 range, but would have thought that it would either be bulkier than it is or lacking IBIS. My best guess was a $7500-$8000 GFX100R without IBIS. I was surprised to get something as sleek and (relatively) affordable as the 100S, while retaining IBIS.
The Sony A1 and Nikon Z9 were both relatively predictable – we’re now waiting on the EOS-R1 to join them. Will all three show up at the Tokyo Olympics? Who will support the sports bodies with big telephotos? The big news about the A1 is that it doesn’t seem to have overheating issues, at least so far – when the similarly sized and specced EOS-R5 does.
L-Mount is in better shape than I expected, largely courtesy of Sigma lenses (and to a lesser extent, bodies) and the Panasonic S5. I am encouraged by OM Digital making more noise than I expected about continuing to release cameras and lenses – but still unclear what the path forward is without a new sensor. I am at least as skeptical of Pentax as ever – the list price of the K3 mk III doesn’t help, since no non-hybrid APS-C DSLR should be that expensive, even one with a highly supported mount – let alone one that is primarily a platform to experiment with vintage lenses.
We still have a market with the big four and a bunch of smaller players. Unlike some other writers, I unequivocally include both Nikon and Fujifilm in the big four. Yes, neither one is as in strong a position right now as Canon or Sony, but Fujifilm’s unusual approach and Nikon’s enormous installed base should keep both around for the foreseeable future. I’m not worried about any of the big four, except in the case of a bad acquisition… Canon has the huge market share, but a lot of it is vulnerable Rebels and EOS-M. Their sustainable share is still almost certainly larger than anybody else’s, but it’s not 2.5 times Sony’s or Nikon’s.
One of the most interesting statements on where we’re headed is from Sigma CEO Kazuto Yamaki. [The] photography market is indeed getting smaller – but it won’t be getting smaller infinitely. According to our estimations, it will reach its minimal size next year, in 2021 – and since then it will stay pretty stable. I would compare the ultimate size of this market to the level of camera sales in times of traditional photography, before the great digital revolution. The true number of passionate people, who want to make photos with highquality cameras and lenses is now – and was then – probably very similar. I’m personally even slightly more optimistic than Yamaki-san, for one reason. There are now more people in the world with the financial means and the freedom to take up photography than there were in the film days. He’s probably right as far as percentages go, but the percentage of photographers is similar in a larger universe of possible photographers. There were three major (Canon, Nikon, Minolta (now Sony)) makers of film SLRs, plus a few smaller companies (Pentax, Olympus). There were also a couple of eastern bloc camera makers (Praktica, Zenit, etc…) and three medium format makers of note (Hasselblad, Mamiya (now Phase One), Bronica), plus Leica. There were also three big aftermarket lens makers (Sigma, Tamron, Tokina) and a number of smaller ones led by Carl Zeiss. Even though digital cameras are more complex and need greater economies of scale, a similar market (~5 million cameras/year) will sustain the big four, Sigma’s lenses (and maybe their few cameras) and a couple of other lens firms (plus small companies making manual lenses that aren’t as complex), and Leica in its niche, and perhaps one or two more besides.