(commentary)
Having just spent a fair amount of time shooting with various state-of-the-art DSLR and mirrorless cameras in my month away from the Internet (new reviews coming), I think I have a fairly good sense of where we are today and where things are headed.
Let's look at the components of cameras:
- Sensor and Imaging Processing — These are essentially the same for DSLR and mirrorless now. While you may see differences due to sensor size, those differences are now basically almost totally attributable to sensor size only, and even then we have lots of image processing tweaking that disguises even that. This is easiest to see in the Sony line: the same APS sensor in a NEX or Alpha DSLR produces basically the same results. I don't expect this to change. Moreover, we're getting near the end of clearly visible gains due to sensor technology itself, at least as long as we stick with Bayer filtration. Thus, I would argue that in the future you don't make a choice between DSLR or mirrorless based upon sensor and imaging ASIC unless the technology of sensors gets disrupted in some way (e.g. post Bayer), and even then it's likely we'll see the same thing across both types of cameras very quickly.
- Viewing System — The optical (OVF viewfinders) versus electronic (EVF electronic viewfinders) debate is an intense one. Clearly some people prefer one over the other. But let's not lose sight of a few things (pardon the pun): neither OVF nor EVF is perfect. OVF over the years has been optimized for brightness and overlays, which comes at the expense of seeing focus detail. Position of the eye at the viewfinder is also a variable for OVF (try it: while looking at an edge slide your eye from one side of the viewfinder to the other; so much for 100% view). OVFs will be dimmer with slower lenses, too. EVFs need more brightness finesse than they currently have (too easily kill night vision in dawn/dusk shooting). In terms of pixels, we're already at SVGA levels with most EVFs, while a few current ones are at XGA levels. In 2014 we'll have EVFs appear at about SXGA levels (1280 x 1024, or >4m dots). EVFs definitely show 100% view, and they can be as bright for an f/5.6 lens as for an f/1.4 one. Still, today we're in an area where OVFs are mature and accepted, while EVFs are maturing and not as accepted. Within a couple of years, I doubt that we'll see a lot of resistance to EVFs, though.
- Focus System — This is another area where we have "mature versus maturing." The phase detect systems in DSLRs are highly mature. Quite frankly, today's Nikon DSLR focus isn't really tangibly better than where we were 5 years ago. Yes, there have been some small changes (f/8 minimum instead of f/5.6, for instance). But technically, the partial-silver dual mirror system used in the Nikon DSLRs is about maxed out. When mirrorless cameras were all contrast detect AF with non state-of-the-art sensors, there was a dramatic difference even in single servo focus between DSLRs and mirrorless. Today, not so much, though continuous autofocus performance still isn't at DSLR levels. But again, within a couple of years at most I doubt we'll be saying that anymore. The Nikon 1 and the Olympus E-M1 are very competent at continuous focus, though not perfect. This will only get better, I think.
- Shutter, Frame Rates, and Buffer — I've grouped these together because they form the core of how well a camera can handle continuous action (focus also factors here, as focus has to be fast even for non-continuous action and has a precision need, which is why I separated it out). There's no reason that a DSLR would be any different than a mirrorless camera in this area, other than that the designers made a conscious choice to trim costs. High-end DSLRs and mirrorless are both going to be pretty similar here. Indeed, an E-M1 has a deeper raw buffer than the D7100 ;~).
- Size — While the current state of affairs gives us mirrorless cameras that are smaller than DSLRs, there's no technical reason for that to be the case for width and height of the camera. The primary gain of removing the mirror is depth of the camera. But current DSLR users won't switch to mirrorless easily if they believe they have to give up their perceived investment in lenses. Nearly 100m Nikkor lenses in gear closets is a lot of market inertia. Thus, I suspect we'll see smaller DSLRs (but with the same overall thickness at the mount), but when Canon and Nikon DSLRs really go mirrorless, they won't change size in depth at all (e.g. will retain the mount). Meanwhile, the mirrorless cameras have been growing in size! Turns out that the sophisticated user wants to be able to hold and control the camera with two hands, just like always. So I don't see size differential of the camera itself being a major factor. On the other hand, smaller sensor cameras produce smaller useful lens sets (notice I didn't write "equivalent lens sets"). So mirrorless systems with smaller sensor sizes (e.g. Nikon 1, m4/3) will tend to have a perceived overall "system" size advantage. As I've been writing since 2009: m4/3 became my backcountry kit for this reason: far smaller pack size and weight for my photographic needs (I don't need super fast lenses for landscapes).
- Power — An often unmentioned aspect of the DSLR versus mirrorless debate. Those EVFs suck power that an OVF doesn't. If you build the mirrorless camera to differentiate in size (e.g. m4/3 until the GH3, Nikon 1, NEX until the A3000), you also start to limit the battery size you can put into the camera. I have to say I was spoiled by the D3 battery life. Those huge batteries lasted me days, even in constant shooting situations. The current higher-end Nikon DSLRs aren't all that bad, either. The EN-EL15's are lasting me over 1000 shots per charge most of the time on OVF cameras. One of the reasons why so many Olympus E-M5 users opted for the vertical grip (making their camera near DSLR sized) was to get two batteries available. Even then, an E-M5 with two batteries doesn't outlast my D800 with a single battery in constant shooting situations. Not even close. The good news is that we're likely to see the camera makers continue to address power. That'll come both in electronics design (lower power needs) and in battery advancements, I think. But this is actually the area where I currently believe that traditional DSLRs have their biggest advantage over mirrorless at the moment. I carry a lot of batteries with me in the backcountry (and a solar panel for emergency charging) when I shoot with the E-M5. Contrast that with my recent Botswana Photo Workshop: I used the same battery in my D800 for over 6 days before I needed to swap out.
- Features — I think it should be clear that there's no reason why DSLR and mirrorless camera features couldn't (shouldn't) be the same. It amuses me that the camera makers spend so much time on this aspect of cameras and manage to do so many things that irritate the existing user base. The reason why the Fujifilm X series cameras resonated so much with the serious photographer world is that they didn't try to reinvent the basic control system for the fifty-second time; instead, they reverted back to one that was known to work well. Go figure.
- Advances — Clearly, the way photography is actually practiced has changed considerably from the old shoot, process, print, deliver workflow of film SLRs. Funny thing is, the cameras haven't actually acknowledged those changes much, if any. The simple illustration of that is how easy it is to get an image you took on a smartphone to where you want others to see it. Now try that same thing with a DSLR or mirrorless camera. At best, clumsy. At worst, it involves a desktop computer. So it should be evident that there are changes that could be made to cameras that better facilitated what the photographer wants to accomplish and how they work. Again, like Features, this is an area where neither DSLR nor mirrorless has any advantage. Technically, they're both so lost in this respect, that they have a current disadvantage to your phone.
So where does all this put us? Tally it up. No advantage in sensor, image processing, shutter, frame rates, buffer, features, or potential advances to either DSLR or mirrorless. Clear disadvantage at the moment to mirrorless in power. Gap closing rapidly in viewing system and focus. And Size doesn't really have to be a relevant issue other than sensor size choice and all that implies.
Conclusion: we're actually very near the point where DSLRs will simply transition to mirrorless for most users. And not that far from the point where DSLRs will go mirrorless for all users.
So if I'm Nikon, what does a future D3xxx look like? Smaller width and height, same sized depth and grip. minimum 2.4m dot EVF with phase detect autofocus on sensor. Needs a bigger battery! No difference in features or controls, but attention to the things I noted in Advances (e.g. built-in WiFi or Bluetooth and a real iOS/Android app that's capable of doing what you instruct from the camera. Whoa. What did I just write? Yep, the opposite of what all the camera companies are trying to do [e.g. just get the image over to the smartphone and then have the user pick up the smartphone and tell it what to do next]). Advantages to Nikon: no change in lens offerings, lower cost to produce, line longevity, stops mirrorless from siphoning bottom users. Advantages to consumer: simpler camera with fewer moving parts to break, uses same lenses they had (legacy support), delivers more and possibly for less.
Of course, with D3100 bodies still sitting on dealer shelves, Nikon has a bit of a different problem in getting to the next generation DSLR: they still haven't sold out the previous generation and have an over supply of the current generation.
