(commentary)
The last few lines of my latest article on mirrorless camera sales growth had these words that appear to have upset some of you F-mount folk: "I still predict that the mirrorless/DSLR world will eventually be one and the same." That provoked a number of "don't take my DSLR away" reactions, most emotional rather than fact based.
While I mostly prefer the OVF (optical viewfinders) that has previously defined DSLRs, their days are numbered. The reasons I prefer OVF probably aren't ones you've even thought of, while most people's reasons for preferring OVF are not supported by the facts.
One common claim is that "OVF viewfinders are brighter." Really? Ever look through one at night? The real usefulness of the OVF is that it is a better match to our eyes, which are adaptive to light: the OVF gives us the light as our eyes would normally see it, not at some arbitrary LED-driven brightness level. Indeed, one of the reasons why I prefer OVF to EVF has to do with that: when I'm shooting at edge of day or into the darkness, I don't want to lose my eye adaption to the ambient light levels. This is the same reason I carry a small, low-power red beam flashlight with me: once my eyes start adapting to lower light levels, I want them to stay there. Our eye/brain doesn't move instantly from cones to rods and vice versa. It can take many minutes to get full adaptation, so I don't want that disruption.
Another claim is that "OVF viewfinders are faster." Technically, yes. There's no electronic delay imposed on the light in an OVF. But...most of the claims about "faster" have to do with the lag between seeing something and the picture being taken. Sorry, but we now have EVF/shutter combinations on mirrorless cameras that are near equal in total lag to the mirror/shutter combinations in DSLRs. Technically, it should be possible for an EVF system to have less lag, as re-establishing the video feed on an EVF should result in shorter "viewfinder blackout" times than it does with an OVF. Put your state-of-the-art mirrorless camera and DSLR into manual focus mode and see what the real "lag" is between them. You might be surprised. And EVF lag will only get better over time.
Yet none of this is the reason why camera makers will eventually drop the mirror in DSLRs. The real reason is simpler: cost reduction and manufacturing simplification. Nowhere is that more visible than in considering the Nikon D3200 and the V2. The D3200 has thousands of parts, the V2 hundreds. Which do you think should have the higher profit margin when sold at the same price? (Why Nikon charges more for a V2 I have no idea. Someone in Japan is dreaming.) Which do you think takes less people to make? Which can be more easily changed when a new part comes along?
This is not an idle consideration. The V2 doesn't need the four mirrors (one partially silvered) to get light into a dedicated autofocus part. The imaging sensor has that built in. Not only does this reduce parts, but it eliminates a time-consuming and fiddly alignment process. In a V2, the primary alignment issue is the lens mount and sensor. In the D3200, it's multiple mirrors, a pentaprism, a focus screen, a metering sensor, an autofocus part, and the sensor/mount alignment.
One of the things about electronic parts versus mechanical parts is that they tend to come down considerably in price with volume production and semiconductor manufacturing advances. Here's one aspect of that you probably haven't realized: solar power is driven by the same technology curves. Some time in the next 10-15 years solar will become the lowest cost form of power because of that constant cost reduction curve that semiconductor-based things enjoy. We're basically two generations of advances away from that point. This is one reason why no one wants to build big multi-billion dollar power plants at the moment. The smart investors can see that your roof is going to be a solar collector and power your house in the future, and that long-distance transport of electricity is inefficient. The real issue with solar power has always been a different one than most people think: storage. You can store oil in a tank, coal in a bin, nuclear in a rod. If you want to power things at night with a solar system, you have to have a way of storing the power you obtain during the day.
But I'm getting a bit off track. Well, maybe not completely: the point is that some changes are inevitable short of a new, yet-to-be-discovered breakthrough short-circuiting them. The inevitable for cameras is that they become less mechanical and more electronic. This was clear to me in 1993, it's even clearer to me today. There's a sub-component to that, though: as things get more electronic, software becomes the most important differential.
Sony is a bit ahead of the curve with its DSLRs: they switched to a fixed mirror and replaced the OVF with an EVF. They'll be switching even more to the all-electronic mirrorless-DSLR-of-the-future in the coming year. Unfortunately, Sony introduced as many disadvantages in their first generation as they did advantages, so the thought that they could use their advanced approach as a way to steal share from Canon and Nikon backfired (nor was it particularly well marketed, either). Still, you can expect that Canon and Nikon will eventually head down this same path, as they can't risk others getting cost/performance advantages over them.
Which brings me to my last point. Some seemed to read my words as saying that the F-mount is doomed. Not at all. Canon and Nikon would be foolish to abandon their existing mounts, just as Pentax, Sony, and Olympus would be. If we ever get to a future where there's no value in your existing lenses, then the whole camera market will be in a complete and utter uproar, and market shares will change and companies disappear overnight. A totally disruptive technology entering the market could do that, I suppose.
Right now, however, I still see mostly iterative technologies on the horizon. If you want to know the future of consumer DSLRs, marry a Sony SLT with a Canon S1L: EVF, on-sensor focus detection, but in a legacy mount with a smaller body size. As usual, Nikon seems to be last to take the steps in the future direction: they've neither electronified their DSLRs (Sony Style) nor have they yet miniaturized them (Canon Style). But I think it's inevitable.
