The common theme you see now is that we're in a DSLR-to-mirrorless transition period. I would agree, but that doesn't mean that DSLRs go away completely, nor does it mean everything will go away. After all, Nikon is still building a high-end film SLR body and selling it (F6).
Transition, therefore, means "the majority of users" transfer from DSLR to mirrorless, and probably over a period of at least three or four years (has to do with update cycles, disposable income, age, and much more). But four years from now, there will still be DSLR users.
Both Canon and Nikon have said they'll continue to build DSLR products while building out their new mirrorless systems. Moreover, Sony is still selling their DSLR (SLT) products, despite not having introduced one for years (the A99m2 body was the last one in 2016, and the last Alpha mount lenses were introduced in 2015).
So let's talk about what's likely to go away in DSLR land, what's likely to stay available, and why.
Canon
It's no secret that the APS-C DSLRs are Canon's biggest bane at the moment. By my count, Canon still has 11 such models on the market they're trying to sell, including multiple generations of some products. Clearly, that will change. Canon executives themselves say so.
The rumors are that the 7D, 77D, and 80D will sort of merge into one model soon, and I don't see Canon abandoing the small, light, and recently introduced SL3 (which is my favorite Canon APS-C DSLR, by the way). The question is whether or not Canon needs more DSLR models with the crop sensor.
I look back to the film-SLR-to-DSLR transition, and I'd say no. The higher end models hang on for a bit during transitions, the lower end models quickly transition and disappear, though there's usually one consumer model that hangs on for a time. Thus, I'd be surprised if we see more Rebels (Kisses) enter the world. And I'd be extremely surprised if we saw more than one.
The last EF-S lens (for APS-C DSLRs from Canon) was introduced in 2017. I don't expect anything particularly new in their crop sensor lens lineup, either. The fact that Canon's lens factory now seems focused on M and RF plus some EF updates by itself says that APS-C DSLR is on last call at Canon.
In full frame, Canon currently has five DSLRs still actively selling. The 1DX will get a Mark III revision, no doubt. It's the Tokyo 2020 Olympics model that's been brewing for several years. The rest? No, I'm not seeing significant updates likely there, though maybe Canon will take the 5D to Mark V with a new generation sensor. I see the 6Dm2 as a dead end (due to the RP) and the 5DS/R as additional dead ends (due to their poor showing in the market and the next RF model rumored to be coming at the end of the year).
The fact that Canon is shooting so high with RF glass early on seems to imply that they aren't going to sit on the full frame DSLRs for performance (other than the 1DXm3). So expect considerable weening in full frame DSLRs.
Curiously, Canon is getting caught out with their Cinema EOS models. Those rely on the EF mount (you can also get a PL version, but that's a smaller subset of the Cinema user base). Sony already is single mount for both their still and video cameras (E mount), and Panasonic is soon to announce an L mount pro video camera, so it looks like they're taking the Sony approach, too. So, canary in the coal mine: watch Cinema EOS. The minute it transitions to RF, that's the day that the EF lens parade stops. Until then, Canon has no choice but to continue pushing EF lenses and updates. I just don't see anything going away in their EF lens lineup any time soon.
So, at Canon: APS-C DSLR is closing up shop, full frame DSLR is still open. The interesting thing is that a huge installed base of Canon APS-C DSLR users exist. If Canon can get their products and marketing messages lined up correctly, they have a large opportunity for moving a significant portion of that base to something new (M or RF). But that's sort of been my point about the mismatch of M and RF: the messaging isn't right.
Nikon
Likewise, Nikon has felt the decline of the consumer DSLR, what Nikon calls consumer DX. The D3500 is Nikon's best selling crop sensor DSLR, and by far. It's doing that mostly on price, which is, of course, a problem in and of itself given how far it has dropped.
About a decade ago Nikon executives said that they had to be prepared to sell a US$400 DSLR. Guess what? They are. And that's still not helping them any, as even D3500 volume is reduced substantially now.
More so than Canon, I think Nikon makes a complete restart with crop sensor cameras. The D3500 is weak and low margin now, the D5600 is missing in action, the D7500 needed a huge discount to get back into the "selling" column, and the D500 never really delivered on sales like the D300 did.
I'd bet that all four DX DSLRs would go away immediately if it weren't for two things: (1) parts commitments, and (2) the need to moderate the contraction. To get costs down (and gross profit margin up), Nikon has always been very aggressive about volume commitments. It's why they've reused sensors across so many models for so many years. I'd bet that the D7500 can't go away soon because it's what will get Nikon to their sensor commitment number.
As for moderating the contraction: Nikon doesn't have a crop sensor future that they've defined and announced yet. Way back in May 2017 I outlined the options that Nikon had for mirrorless. They picked my #4 as their first step, but they haven't yet tipped their hand as to what they'll do with crop sensors. That was my #3 options:
a. use the existing DX mount
b. create a new crop sensor mount
Neither seems out of the question, surprisingly. The line of DX AF-P lenses mean that Nikon has a small set of wide angle to telephoto lenses that would perform quite well on #3a. Meanwhile, if #3b is "use the Z mount," that doesn't seem to play into what crop sensor must be now: a much smaller and lighter offering than full frame (or something very high end, ala the D500). So #3b could be a play like Canon's EOS M: a different mount than their full frame mount. Personally, I think that would be a wrong play.
Frankly, Nikon has no perfect option for crop sensor at the moment, which may be why they're taking their time to try to get whatever they do there as "right as possible."
This is a long-winded way of saying "DX DSLRs go away relatively fast, but we don't know what replaces them, if anything."
The FX full frame DSLR, however, is another story. The D5 will be replaced with a D6, for the same reasons Canon will do the 1DXm3: wanting something top end for the 2020 Tokyo Olympics coupled with the fact that they started design work on these new cameras back in 2015/6.
The D850 still sells well, and I believe that Nikon will likely update it again (the D850 is itself the third update in the series). I've long thought that the D750 will get an update, and I can see such an update as being a "bridge" between DSLR and mirrorless: use the Z6 sensor and logic and you get a highly competent and silent DSLR using Live View. You also get the video improvements and more. Throw in Sensor VR, and I'd say you have a life extension machine (that would be the D750's life, not yours ;~).
So I'm not expecting a lot to change in the full frame end of Nikon DSLR lineup. The D610 and Df seem destined to go away, though.
In terms of lenses, as much as Nikon likes to say they're an optics company, they always seem to be a "constrained" optics company. They don't launch a lot of new lenses, and now they seem committed to pushing out Z lenses at a pace that consumes their constrained limits. We'll probably get an FX leftover design or update or two, but I'm not holding my breath. Realistically, that's fine, as the Nikkor F-mount lineup is pretty complete and compelling as it is.
What's likely going to go away first in full frame are the screw-drive lenses. That would be the 14mm f/2.8D, 16mm f/2.8D, 20mm f/2.8D, 24mm f/2.8D, 28mm f/2.8D, 35mm f/1.4, 35mm f/2D, 50mm f/1.4D, 60mm f/2.8D, 80-200mm f/2.8D, 105mm f/2D, 135mm f/2D, 180mm f/2.8D, and 300mm f/4D. But not quickly. Nikon has this tendency to just keep a lens in stock as long as there's any demand for it.
Which is why the AI-S lenses still exist! They work well on the remaining film SLRs (!) and now on the new Z's. Things die slowly at Nikon. So even though I'm saying that pretty much all the lenses that aren't AF-S/AF-P are in danger of going away, there's no indication that they will do so anytime soon.
The thing that's going to surprise you is this: Nikon flash is broken. I wouldn't be surprised if it goes away or gets replaced. Oh, the SB-5000 is probably safe. But all kinds of Nikon flash is now showing senility. The SU-800 doesn't work correctly with the D7500 or the Z6/Z7, for instance. Which means that the macro flash kit is also kaput unless Nikon fixes this problem. The SB-300 and SB-700 also seem out of sync with the cameras now (e.g. the SB-700 doesn't show up in the menu-driven flash options).
Either Nikon is going to redo the flash lineup and fix compatibilities, or they're going to let most of it wilt. Funny thing is, the D5 (and upcoming D6), Z6 and Z7 don't have a built in flash, and don't use some of the features of the current flashes (such as Autofocus Assist), so Nikon flash needs a redo, not a wilting.
Pentax
With no mirrorless strategy in place, all Pentax has is DSLR. Unfortunately, the news isn't good no matter what comes next. If it's "stay the course with DSLRs," that means something like a 2% market share of a fast declining market, which doesn't allow for much investment in new product. If it's "let's do mirrorless," they don't have enough engineering bandwidth to continue to do much with DSLRs.
A recent dpreview interview with Ricoh executives at the CP+ trade show in Japan generated the following quote: "I imagine, in two or three years, some users who bought mirrorless cameras will return to DSLRs or choose to use both systems, because each has its own benefits."
In any transition, there's always a subset of folk who essentially create a small backlash. They go to the new restaurant and sample the new cuisine, don't like it, and then return their old restaurant, at least until it closes.
Thing is, as I've said from the beginning of mirrorless, the camera makers have big incentives to get people to transition. Fewer parts and simpler construction that requires fewer alignment steps means that if they can keep price points close to intact, they can make more profit. And if the market continues contracting, they aren't as exposed to that because they can give up some of their increased profit margin.
If Pentax thinks that their forecast Back to the DSLR movement is going to be significant in size and benefit them, I'd say they're going to be wrong. Pentax is in the wrong place to take advantage of that, as they don't have any mirrorless folk who would backlash ;~). I really don't see a Canon M/RF, Nikon Z, or Sony E owner saying "you know, I liked DSLRs better" and then going out and buying a Pentax DSLR.
A persistent rumor has been going around about the Pentax brand, too: that Hoya only licensed the Pentax name to Ricoh for a specific period of time. Frankly, Ricoh should have consolidated its Ricoh and Pentax cameras under one brand name a long time ago. Lately I've noticed more "Ricoh" executives at trade shows, but fewer "Pentax" ones. That might be because of launches of things like the GR and Theta, but it might be more indicative that the one-brand idea is finally starting to play out internally.
While Pentax was early to the small MF Sony sensor with the 645Z in 2014, they've been quiet since Fujifilm and Hassleblad came on the scene, and Fujifilm clearly wants to own that market, so silence on Pentax's part is not comforting.
Likewise, that the K1 is still on the 36mp Sony sensor and with only a minor change since its appearance in 2016 also isn't comforting. It feels like Pentax is slowly sliding backwards while the other camera companies push forward despite the market contraction.
That said, I don't see anything going away in the Pentax lineup until it all goes away.
But...
Take a deep breath, close your eyes, concentrate on shutting down your senses, think about nothing but your mantra. No, you can't use the mantra Om, because that might make you desire Olympus resurrect the OM series as a full frame mirrorless camera ;~).
If you were to make the decision today to buy your first DSLR camera, you're fine. The best of the bunch are excellent cameras that should last you years, probably a decade. And in the Canikon world, you'll be able to buy lenses and accessories for quite some time. Longer than you camera will last.
If you're making a decision about a camera today based upon some belief that things will be different five or ten years from now—e.g. buying mirrorless now because it is the future—you're a grass is always greener on the other side time traveler.
I've written it many times: the best all-around ILC you can buy today is the Nikon D850, a DSLR. No, the Sony A7Rm3 didn't change that. Nor does the Nikon Z7.
Yes, our DSLR choices are going to start getting more limited moving forward. But the ones that remain are great choices.
