It seems a lot of you missed the announcement.
Okay, there was no official announcement, just a clear trend: the bottom end of dedicated cameras is now 24mp. Moreover, with the likely upcoming mirrorless announcements from Canon and Nikon, the true consumer bottom the big makers really want to establish is 24mp full frame (EF, FE, FX) at a price point of around US$2000 initially.
This is posing a problem for quite a few companies. We have three things to discuss here: pixel count, sensor size, and price.
Let's start with the 24mp.
That effectively puts the m4/3 cameras below the bar. Panasonic has the GH5 positioned as a video camera, which is how they're managing to keep their heads above water. But Olympus is now right at the tide line. Even the 1" sensor cameras are tricky, as Nikon's cancellation of the DL series shows: at 20mp with low light issues, can these cameras really hold off the smartphones with many multiple sensors that are now visibly sneaking over the horizon? Panasonic and Sony's response has tended to be the thing Nikon toyed with in Coolpix for so long: longer lenses. (Hmm, didn't Cat Stevens have a song whose lyrics went "Longer lenses are coming to win us, coming to win us..."? Oh, no, that was boats. Never mind.)
So, 20mp is basically just at the waterline at the moment. And its primary advantage over the ever marching forward smartphones would be long focal length ranges. Thus the return of the longer lens on the RX-100 Mark VI and Panasonic's likewise push upward with the ZS2000. (And again: Nikon had it two-thirds right with the DLs. The 18-50 was a long focal range that would keep smartphones at bay for wide angle work, the 24-500 was to be the camera that carried the P900 crown forward. The problem is that marketing and sales thought the 24-85 was going to be the big seller, and it wasn't far from what was already available [RX, GX] or the oncoming smartphones. I still believe it was a mistake cancelling the DLs. It's left Nikon retrenching far higher up the line than anyone else, which implies a future far lower volume of customers. I don't know how Nikon gets entry customers going forward given all the mistakes they've made with Coolpix, KeyMission, DL, and even lower end DX. Eventually, they'll need an answer.)
While I understand how we got to Nikon's 20mp DX siblings (D7500, D500), the pixel count there is decidedly hampering sales. The public has correctly perceived that 24mp is the entry point, and ironically Nikon had a role in establishing that. That said, 20mp APS-C is better than 20mp m4/3 is better than 20mp 1".
I guess I wouldn't argue with you if you argued 20mp was the entry point for dedicated cameras, but evidence seems to indicate most people think it's 24mp now.
Which brings us to sensors.
Every camera company is now spending a lot of money on sensor development. While Sony Semiconductor may be the sensor manufacturer for pretty much everyone except Canon, as volumes came back down everyone went to the most critical aspect of a digital camera to try to distinguish themselves: the sensor/ASIC engines. The problem with this is that APS-C sensor prices stopped going down due to volume; they've gone back up as everyone gets into the engineering tinkering game.
All that R&D being poured into unique sensors has to be paid back by customers at some point. So, yes, you can get some pretty nifty tech in APS-C image engines—from Canon's dual pixel to Fujifilm's X-Trans to Nikon's BSI to Sony's new faster BIONZ—but you're paying more for it. Only Canon right now has the clear volume to spread their APS-C sensor costs across more products, making that new tech tariff a little lower for their users.
Which brings us to price point. US$500-1000 has long been the sweet spot for dedicated cameras. Tens of millions of ILC devices have been sold in that range, though the biggest decline in volume is now occurring there. If only the camera makers could reset the price point to US$2000, they'd be in a better place, especially if they could get that volume moving up. The Canon 6D, Nikon D750, and the Sony A7 have told the big three that yes, with some care, you can get a nice bump in volume at US$2000.
Indeed, we're back to Olympus: putting the E-M1 Mark II at 20mp m4/3 at the US$2000 price point just made those US$2000 24mp full frame cameras look much more compelling. And US$2000 was enough to deal with the higher sensor costs in those full frame cameras.
So we're now in an era I'd describe this way: 24mp, full frame, US$2000 as the primary entry point the camera companies want you to pick, with 24mp, APS-C, ~US$1000 as the fallback for the price conscious. Anything outside of those two has to have a unique reason to exist, something that would make you ignore the three primary attributes I just described.
The RX-100 Mark VI, for instance, tries to make its case by being able to fit in a shirt pocket. The GH5/5s claim to be broadcast level 4K video cameras. The list of "compelling cameras" outside the three attributes goes on similarly.
You'll note that I haven't mentioned DSLR versus mirrorless.
Why? It doesn't matter.
Oh, it matters in the sense that the camera makers would really like to get to mirrorless with global electronic shutter as fast as possible, as it removes complications and cost. But in the photographer sense? No, not really. The Seven Dwarves that moved to the mirrorless first all have managed to get you to believe that smaller, lighter, face focus, and WYSIWIG can't be done in DSLRs, but that's not particularly true. That the DSLR Duopoly has pretty much been lethargic at doing those things is more the case.
Canon did manage to show us that the first two are very possible with the SL models, but they haven't promoted them, probably because more of their lineup is bigger, heavier DSLRs they want to still sell. Nikon manages to do face detection in the D5 series cameras, but they don't seem to know how to market that.
No, the real story isn't DSLR versus mirrorless. As I've written for years, we'll eventually lose the mirror and its complexity, and we'll lose the mechanical shutter and its complexity, too. That's evolutionary and driven by manufacturing issues, not user issues.
The real story is where the entry bar is, and how many players can live with what it is (now and as it evolves further). So let's take a brief look at each company:
- Canon — ILC clears the bar, compacts just below it. In the ILC lineup, Canon is 24mp and up, has plenty of full frame with more coming, and is the one player that can still easily dip below the US$1000 lower bar without hurting themselves.
- Fujifilm — clears the bar, sort of. Fujifilm's main lineup all manages the 24mp part just fine. A lot of their lineup sits at, near, or below the lower US$1000 point. But their sensor is APS-C. Fujifilm avoids the intense full frame competition by jumping to MF. One might say that Fujifilm is cleverly playing just outside where the Big Three are fighting. The problem, of course, is that forces customers to convince themselves that different is better.
- Leica — ILC clears the bar but only with premium pricing, compacts just below it. To Leica's credit, they understood 24mp and full frame as points they needed to hit. They even have paid some attention to the "lower" side of things with APS-C entries, though again with premium pricing. In essence, they've done a good job of protecting their premium position by mimicking the mainstream entry points, but with their higher-end cache.
- Nikon — ILC clears the bar, compacts are crashing into the ground. For the time being, the two 20mp DX DSLRs are close enough due to their other attributes, and Nikon has spent a lot of time defining the attributes in the full frame race. Of course, many feel that Nikon has some tired horses in this race (D3xxx, D5xxx at and below the lower bar) and some missing horses (mirrorless). I don't think that will stay true long. Nikon is a one-thing-at-a-time company, so we simply await what they'll do in a few areas.
- Olympus — below the bar. Olympus really only has two win conditions with a customer: ignore the E-M1 Mark II and buy an older, smaller camera at a lower cost. Olympus is poised to fail at the US$2000 full frame bar. The 1" compacts nibble at the small, light side of things (and why do we not have a m4/3 compact from them?). Too much engineering is going on below the bars—they've got five m4/3 bodies in that space, as if that will bring them more volume. Nope.
- Panasonic — mostly below the bar. DFD has an issue against on-coming PD systems: Panasonic would need to drive bandwidth on the sensor far higher than it is to keep up. But that's the least of Panasonic's worries. Despite very well designed cameras that are also generally well received, they're not 24mp, they're not full frame, and what they have at the US$2000 price point has the same problems as Olympus (unless you're a videographer).
- Pentax — A very unique case. Yes, they're above the bar. But I'd argue that on a lot of key performance attributes—e.g. autofocus—their engineering puts them behind. That's not because they don't have great engineers. It's because their volume now makes it difficult to stay with the innovators. You see little glimpses of what they can do (e.g. pixel shift), but the entire systems are moving behind the Big Three in terms of features, technology, performance, and pace. Of the DSLR players, it's now easy to predict that Pentax would be the last to 24mp, full frame, mirrorless.
- Sony — ILC clears the bar, compacts just below it. That said, Sony has been right up there with Nikon in terms of pressing the 24mp, full frame, US$2000 entry point (and the models above it). They've also been trying to keep those compacts from falling too close to the oncoming smartphones. It's the middle, that US$1000 ILC price point, where Sony has been less than active lately. That allowed both Canon and Fujifilm to get some traction, has kept entry DSLR sales from collapsing more, and a host of other issues that Sony probably needs to eventually fix.
The 24mp, full frame, US$2000 point is going to be interesting to watch. Canon, Nikon, and Sony all sell in the very low six figures of units a year with that. Let's call the total 600k/year for argument's sake. If the bar really is at the right point and the marketing engines work, we should see that number rise significantly. Indeed, if Canon and Nikon both transition their entry point full frame to mirrorless later this year, there may be enough marketing energy after Photokina to spawn a clear uptick in full frame sales (assuming, of course, that all the current trade war nonsense doesn't kill the global economy).
The lower US$1000 (and under) APS-C point is also interesting to watch. Generally, it's been running 10:1 or so to full frame in terms of volume. Thus, fumbles in this space tend to lower your overall ILC market share and give you fewer products over which to average costs.
Thing is, the entry bar changes over time. But it also will tend to remain at one level for awhile, as you need time to pay back R&D. We had a long run at 6mp APS-C, for instance. I suspect we're having a long run with 24mp full frame now.
Eventually, though, tech's relentless push forces you off the old entry bar to a new bar. That's the trickiest aspect of all. No one really seems to have a strong sense of where the bar will go. Computational photography is absolutely in progress in smartphones. It eventually will have to be embraced by the camera companies, too. When? How? Those are the known unknowns.
One final comment for the professional crowd: one of the on-going problems professional photographers have had in the digital age is that competent amateurs often are in the right place at the right time with entry bar cameras (i.e. ones that produce usable images for the media). If the bar is at 24mp full frame, that really means that the pros have to up their game considerably, as 24mp full frame is enough for a two-page magazine spread, even at high ISO values.
Thus, as the entry bar moves up, you see the pros having to move up too. The pro bar is currently at the Sony A7Rm3 and Nikon D850 level. With exceptional shooting and sales skills also required. It helps to have a unique style or look, and it helps to be able to deliver video at the same time as still shoots. Lighting is a way to distinguish yourself. It would also be helpful to be extremely fast at delivery.
