News/Views
Fujifilm Pushes APS-C To New Levels
Today at the annual Fujifilm X Summit—this year's being the 10th anniversary of the XF mount—Fujifilm introduced new products that push their APS-C efforts into new territory. Specifically, one camera and two lenses.
The camera is the X-H2S, a 26mp camera incorporating a new X-Trans image sensor that's stacked. This marks the first appearance of a stacked image sensor in the APS-C realm (Canon, Nikon, and Sony APS-C sensors are either FSI or BSI these days, but not stacked). Stacked image sensors are expensive to produce, with the primary benefit being "speed."
The X-H2S uses that speed in several ways. First, the electronic shutter of the X-H2S generates up to 40 fps (with limitations; mechanical shutter is 15 fps with near unlimited buffer). Second, for video the relevant specs are 6.2K/30P video and up to 60P for DCI 4K video. Subject detection in the autofocus system is also another area that improves with that new speedy image sensor and processor, now including animal and other object tracking, as well as 120Hz refresh.
I noted cost in passing, so let's go right to the bottom line: the X-H2S also now becomes the most expensive APS-C camera on the market at US$2500 (body only). Add some of the accessories (US$1000 file transfer grip for Ethernet use, US$400 vertical grip, US$200 fan) and it can get pricey.
Other X-H2S features that are sure to catch your interest:
- The X-H2S has Mode dial and forgoes the dials only interface of many previous X models. The X-H2S is more "DSLR-like" in design. As I've noted many times, the all-dials interface works for slower, more contemplative work, but if you want speed in handling, the button+dial DSLR designs are what perfected that.
- As with most of the high-end video capable mirrorless models these days, the X-H2S has a fully adjustable Rear LCD, which rotates out from the camera as well as tilting.
- Fujifilm has ironically progressed to the type of Focus Mode button/switch that Nikon has (mostly) recently abandoned.
- A US$200 optional cooling fan mounts onto the back of X-H2S should you need to push it to its max, particularly on the video side.
- Data storage is either CFexpress Type B or UHS-II SD (one slot each), or via a special file transfer grip to external media.
- The viewfinder is 5.76m dots, though it has an eye point of 24mm.
Overall, the X-H2s easily establishes a new top for mirrorless APS-C capabilities, though, again, at a price. Canon's R7 is the "poor person's" top APS-C choice, providing a reasonable subset of state-of-the-art APS-C (review coming soon).
The key new lens that was introduced is the 150-600mm f/5.6-8. Thing is, a "speed camera" really is truly only necessary in sports and wildlife photography, and you need lenses to support that. Up to this point, the primary lenses that Fujifilm has in the XF mount that work well for that are the 50-140mm f/2.8, 100-400mm f/4.5-5.6, and 200mm f/2, which really isn't a broad enough set to support a camera such as the X-H2S. The new US$2000 150-600mm f/5.6-8 extends the reach of the APS-C system, but I'm not sure that it fully satisfies the needs in the XF lineup. What we're not seeing is compact, light lens choices in the telephoto range (ala Nikon's PF series), and I'd think that really needs to show up in order to fend off what's happening in the full frame world the X-H2S will need to compete against.
To a large degree, the X-H2S and 150-600mm f/5.6-8 are sort of a natural offshoot of the Nikon D500 and 200-500mm f/5.6 success. However, many of us D500 users moved on, particularly once the Z9 and a wide series of PF lenses produced a total size/weight package that out-performed the older DSLR. While the Z9 is a heavier body, the 300mm f/4, soon to be 400m f/4.5, 500mm f/5.6, and 800mm f/6.3 PF lenses are light, short, and reasonably fast in aperture. Should Nikon move the Z9 features down into smaller cameras soon, Fujifilm will have a tough sale with the X-H2S in terms of winning over traditional mount users (Canon EF, Nikon F).
The second new lens is the 18-120mm f/4 (effectively 28-180mm). This lens is a power mid-range zoom.
A second camera was previewed with the X-H2S, the X-H2. The X-H2 goes yet another route for APS-C: a mammoth pixel count. Another new image sensor shows up in the X-H2, providing 40mp and 8K video, but it's not a stacked sensor. 40mp APS-C puts us close to 100mp full frame, and thus diffraction becomes a real issue, as would be any even small lens misalignment or even less than state-of-the-art optical design. Many of us have already found clear lens resolution issues with the 100mp GFX cameras, which use a sensor two sizes larger, so 40mp APS-C is definitely not going to come without birthing pains.
Which brings me to this: Fujifilm still seems to be pursuing the "big numbers" plan for moving cameras forward. I'm not at all convinced that that approach is the winning one long term. Short term, sure, it gives you bragging rights that wins over the easily convinced. But I'm one that wants to see real improvement in results. For example, for me the 61mp Sony A7R Mark IV was a disappointment. In many ways, the 50mp A1 actually performed better in terms of ultimate image quality, and I still believe that the real top of the image quality heap is defined by the 45mp Nikon D850 and Z7 II (almost equalled by the Z9, but not at base ISO). And the "sweet spot" in terms of image quality for most people would still be 24-30mp full frame, which translates down to something like 12mp APS-C. Thus, I'm skeptical of a 40mp APS-C camera's true usefulness.
That said, Fujifilm's commitment to APS-C over full frame doesn't leave them a lot of options. We'll see how they deliver once the new cameras and lenses become available.
The Dead-end of Canon M Has Been Reached
Well, I've been writing this for some time: Canon's M-mount didn't make a lot of sense once the RF-mount was introduced, as the incompatibility between the two Canon mirrorless mounts didn't provide for a logical upsell or upgrade position from crop sensor to full frame. Both Nikon and Sony use the same mount (Z and E, respectively) for both their APS-C and full frame cameras, which is a much more logical and traditional approach.
With the introduction of the Canon R7 and Canon R10 and the rumored discontinuation of the M6 Mark II, the M-mount now appears to be on life support. No new M-mount lenses have appeared in a some time, though third parties have filled in some of that gap. Both the new R7 and new R10 models introduced today use Canon APS-C sensors, but with an RF lens mount up front, not an M-mount. While both cameras are more upscale and DSLR-like than the M's have been, I think it's only a matter of time before Canon works their way down to the M50 style camera in the RF mount. Again, rumors have it that the M6 Mark II has been discontinued—and the M5 disappeared long ago—and I suspect we'll see discontinuation slowly work its way downward in the M lineup until nothing is left.
The US$1499 R7 appears to be the long-rumored 7D Mark II "replacement" (in mirrorless form). This 32.5mp camera is 30 fps (electronic shutter, 15 fps mechanical), and supports 4K/60P video, though the 60P is at line-skipped or cropped. Autofocus is dual-pixel and has similar subject recognition and tracking to the R3. The image sensor is stabilized (up to 7 stops CIPA with certain lenses) and we get an articulating Rear LCD. The camera uses two UHS-II card slots and is powered by the LP-E6NH battery. Overall, the R7 body is not particularly downsized from the full frame R sizes, though it does lose some modest breadth, height, and weight from the full frame models. Once again Canon is playing with ergonomics, though, with a large thumbwheel now appearing at the top of the rear.
Meanwhile, the US$979 R10 is the new RF entry-level camera, sporting a 24mp dual-pixel image sensor that supports 4K/30P video normally, or 60P with a 1.56x crop, and a maximum of 23 fps (electronic shutter; 15 fps mechanical). The focus system is basically the same as on the R7, again derived from the R3 advances. One UHS-II card slot, and the R10 uses the LP-E17 battery. The R10 body is more downsized and lighter than the existing R bodies, though you might not immediately notice that from photos.
Along with the R7/R10 announcement we get two new (oddly specified) RF-S lenses: the 18-45mm f/3.5-6.3, and the 18-150mm f/3.5-6.3. I write "oddly specified" because 18mm on Canon's 1.6x APS-C format means these work out to be 29-72mm and 29-240mm effective focal lengths. In other words, neither of the new kit lenses give you very much of a wide angle, a trait that Canon's been getting away with for some time in their video and crop sensor lineups, but now one that appears headed towards formalization (the M lineup has a 15-45mm lens, which would be 24-72mm effective ;~). Even the recent RF-mount 16mm f/2.8 STM doesn't quite get you to 24mm effective on the new cameras.
Canon, of course, says that APS-C cameras in the M-line and DSLR line will continue, but one naturally wonders how much lifespan those models have. The news here is that the R7/R10 trump all those models in (1) video capability, (2) focus performance, and (3) frame rate (and for many, buffer). The DSLR-styling of the new models makes the Sony A6### models look as oddball as they are in ergonomics.
What's missing, of course, is lens choice. Sure, you can mount full frame RF lenses on the R7 and R10, and you can mount old EF lenses via an adapter. But in terms of compelling APS-C lenses, Canon is starting with...well, two not overly compelling kit lenses. This has been my complaint about Canon's APS-C efforts since the beginning: lens choice has been minimal and not particularly engaging. The R7 and R10 are now up against Fujifilm's complete lens lineup as well as Sony's about-to-grow lineup. At least the two new EF-S bodies seem reasonably compelling, but exactly what do you want to stick on the front? Yikes the EF-S line looks lean. Even when I consider regular RF lenses I only find three more that "feel" right for the new bodies.
That said, I'm sure that the R7 and R10 will do well. Against Nikon's Z50 the R10 looks better, though the Z50's lenses look better. Against Sony's A6###, Canon can now claim to have caught up (and perhaps passed) the tech giant in most ways, though again the lens situation needs a lot of work. Against Fujifilm...let's wait until next week before we tackle that question, as Fujifilm is about to make their own APS-C announcements.
Support this site by purchasing from the following advertiser:
The Coming Fast Body APS-C War
If rumors are to be believed, Fujifilm will announce an X-H2s on May 31 and Canon will announce the R7 on May 24. Both are "fast" APS-C bodies, in the sense that they will provide better than Nikon D500 frame rates and more pixels. (I believe that the Nikon D500 is still the current reigning APS-C body, despite being five years old at this point.) The R7 is rumored to be 32mp at 15 fps mechanical, 30 fps electronic. The X-H2s is rumored to use a 26mp stacked sensor, which also implies high frame rates, rumored to be as high as 40 fps electronic. Both new cameras will offer more pixels and speed than the current king of the APS-C hill, the Nikon D500, it appears.
It seems that the birders among you are about to have many interesting choices (OMDS OM-1, Fujifilm X-H2s, Canon R7, Nikon 800mm PF, etc.). Lenses play a huge role for the birding crowd, which is why m4/3 and that Nikon 800mm f/6.3 PF are in that list. On an APS-C body, you'd have 1200mm effective in a really small, competent package with Nikon's latest (plus they have an even smaller and lighter 400mm f/4.5 coming soon). So is Nikon going to join the fast APS-C body war? I'd say not likely this year, but a Z7 III would be effectively a 20mp APS-C body using DX crop and likely to have fast frame rates, too.
The question mark seems to be Sony, who currently doesn't seem to have a clear APS-C plan, yet has a whole lineup of such cameras that could use updating.
Still, I anticipate a heck of lot of "noise" about what's the best birding approach come the end of the year. Here's a preview of what we're likely to hear as the Internet wraps its head around the upcoming offerings:
- Canon RF. Reasonable body choices (R5 in crop mode, R7), slowish lenses?
- Fujifilm XF: State-of-the-art APS-C body choices, not a lot of lens choice?
- m4/3: Great lens choice, not so state-of-the-art focus performance?
- Nikon Z: Excellent lens choices, but where's the (non Z9) DX body?
- Sony: Just choose FE?
One of the things about catering to higher-end enthusiasts is that their chosen specialty aspects start to play a large role in their buying choices. That enthusiast may be solely speciality (wildlife, birds, sports, street, macro, travel, landscape, etc.), or they may be multi-dimensional (some or all of that previous list, plus others). Either way, a camera company has to have both entry and higher level choices for, well...everything, lest they be pigeon-holed as only a speciality brand. The problem I see is that none of the Japanese companies can really point to a broad, deep, full lineup except maybe the m4/3 companies. Even there I see issues, though.
I'm all for better cameras and more choice. The problem I see is that each of the camera companies has a "liability" that makes it easier for the Internet to generate heated arguments over "what's best."
You may have noted that I continually update my "all-around camera" choices. In the long-run, I've found it better to have a truly solid all-around camera than specialty cameras. In the mirrorless world right now, I'd judge the top three all-around cameras to be the Canon R5, Nikon Z9, and Sony A1 (not in that order ;~). So as the new cameras come to market later this month, I'll be looking at whether they provide something that the current all-around choices don't, and how broadly that extends.
From the details I'm hearing, I might have to add an APS-C camera to my "all-around" choices come this summer. But I'd caution you to look not just at camera, but also to lens choices for it, because you don't get to the all-around capability without both.
Does Mirrorless Make You Better?
There's a misguided notion that the only reason why DSLRs are getting abandoned and mirrorless emphasized is that this is what the camera makers want to happen. To some degree, I'm guilty of adding fuel to that fire by pointing out—as far back as a decade ago—that mirrorless was probably inevitable as it allows the camera makers to take out complexity in parts and manufacturing.
But the notion that the camera makers themselves are in full control of the transition is, at the minimum, not complete, and potentially misleading, as well.
Thing is, a dedicated camera is not something anyone other than perhaps a photojournalist, an event photographer, or some pros with corporate clients needs. If the working professionals were the only ones buying cameras, the market would be so small that we'd have only two or three companies selling products, and they'd be at least twice the current costs, probably higher.
Two other customers in the market are more important in the long run than the pros: enthusiast/hobbyists, and true consumers are driving the market, particularly when they get convinced they need a camera (e.g. the old Kodak "protect your memories" guilt marketing). Those two groups have far more influence on what happens in terms of moving from one platform (film SLR, DSLR) to another (DSLR, mirrorless). If those groups don't see the benefit to them, such moves don't happen. Or more likely: they are attempted, but don't take root.
DSLRs solved a clear problem for the enthusiast/hobbyist/consumer: suddenly they could near instantly verify that the photograph they took is the one they wanted. We joke about sports photographers "chimping" with the first DSLR cameras, but as it turns out the ones that benefited most from instant image review were the non-pro group. That ability was a clear benefit to them, and they started buying dedicated cameras in record numbers very quickly. That led to a dramatic increase in units sold in the first decade of this century, and probably saved a few camera makers from oblivion along the way.
The question in the headline is essentially this: do mirrorless cameras have that same kind of impact that DSLRs did? Do the enthusiasts/hobbyist/consumer groups get enough clear benefit that they essentially endorse the platform switch from DSLR to mirrorless and generate new sales for the camera companies?
My answer is mixed. Yes in several aspects, no in another.
First, let's start with the benefit side. There have been several, though they're not free from argument as to their importance:
- WYSIWYG — Mirrorless elevates chimping to real time. The viewfinder fairly represents the final image.
- Size/Weight — Taking out parts also let the cameras get smaller and lighter. This crowd never really liked large five-pound necklaces.
- Focus — Focus anywhere in the frame was new and useful. Used correctly, it gets rid of the focus-and-recompose delay.
I'd argue that those three things were enough to make the transition to mirrorless inevitable. It was a slow transition, though, taking over a decade before mirrorless unit volume exceeded DSLR unit volume. However, as more crossed over and began touting the benefits, the transition eventually gained enough momentum that the two dinosaurs of Canon and Nikon had to make the switch, too.
But let me be clear, this wasn't an early round KO by mirrorless. DSLR users are still fighting, though the camera makers have mostly moved on.
In retrospect, the transition point was when Sony gave up on SLT types of DSLRs and eliminated the pellicle mirror. Doing so let them get to all three of the benefits I point out above instead of just one. To Sony's credit, once that Alpha decision was made, they ran with it as fast as they could, and created a lot of marketing noise as they did. Which is how the mostly DSLR user base started hearing about the fact that there might be benefits from a new platform compared to their existing one.
Over the decade, I documented a lot of this. I discovered Samplers, Leakers, and Switchers, all of whom had been dedicated DSLR users, but in their exploration of alternatives eventually became mirrorless users. At one point I found through surveys (mostly of Nikon DSLR users) that leaking was driving about 5% of the camera purchasing going on at the time. But 5% sustained eventually eats up the 100% ;~). The Rule of 72 says that this 5% leak would take only 14 years to become 100%, remarkably close to the time it actually will end up taking.
If it weren't for the supply chain issues going on, I believe that we'd see a faster transition now, and DSLR sales dwindle even faster. That's a bit ironic. I'd have thought that part supply issues would have just had Canon and Nikon leave DSLRs totally behind, but there's something I'm not understanding about the supply chain that's kept them moving enough parts into DSLR production that we're still within sight of the 50/50 transition point. I suspect that it's that to get back on top in mirrorless, Canon and Nikon need the next generation in parts (stacked sensors, faster image processors, etc.), and those have been difficult to get on fabs in quantity, let alone to drive down in price. But it'll happen.
The question mirrorless users should be asking themselves is this: what's the next major transition point? Nikon's Z9 seems to suggest the loss of the mechanical shutter, and I'm pretty sure all the camera makers would love to lose their mechanical IS platforms, too. Cards seem so old school in this age of moving things through the cloud. And, of course, more machine learning can make that WYSIWYG/Focus benefit into WYWITFP (What You Wanted In The First Place).
Life doesn't stand still in technology products. If you're a mirrorless camera owner today, you're in the golden period of mirrorless: you have plenty of choices, they all perform quite well, and they all pretty much live up to the promises that mirrorless was making for you in the first place. The camera makers will continue to refine those products—but not make large advances—right up until the next platform shift occurs.
