Here's a question I get in a variety of ways: why do we need mirrorless cameras?
Sometimes it's not a question, but an assertion. That assertion is often hugely broad, as in "mirrorless is the future of interchangeable lens cameras."
Pretty much every variation of the question or assertion I receive has at its root an assumption that mirrorless cameras are different than DSLRs in some meaningful way.
I’ve done a lot of thinking about this question in the last seven years. A lot. I’ve also challenged those making the assertion to explain how they got to their conclusion while also trying to ascertain why others were asking the question.
So let’s first start with the thing that’s the same about mirrorless cameras and DSLRs: they both allow you to change lenses. They’re interchangeable lens cameras (ILCs). At the heart of any question or assertion is that fact: we’re talking about ILCs.
Early on in the mirrorless history (2009), the mirrorless competitors were essentially overwhelmed by the legacy lens sets of the DSLR makers. Back at the start of mirrorless, just the Canon and Nikon SLR/DSLR lenses in circulation alone was over 75 million units, and both companies had 60+ lenses in their current catalog. All while mirrorless was really starting from zero in both categories.
That’s one reason why you saw so much “use an adapter” talk early on in mirrorless, even if it meant losing automated functions (exposure and focus). If there was an advantage to mirrorless, it wasn’t lenses early on.
So that’s where we really have to start our discussion of the question/assertion: why would anyone have even thought to start down the mirrorless road?
Well, from the manufacturer standpoint, look at the ones that entered mirrorless early and aggressively: the Seven Dwarves as I called them (Fujifilm, Leica, Olympus, Panasonic, Ricoh, Samsung, Sony). Fujifilm and Sony are actually two of the later entrants in mirrorless, but not before they had first beat their heads against Canon and Nikon trying to make DSLRs.
So the camera makers thought that mirrorless was a chance to create a “different” product that the Duopoly didn’t have, the classic Ries and Trout “create a new market” suggestion for #3 and lower players.
What was different about the early mirrorless cameras? Remember the early Olympus marketing and ads for the E-P1? That camera was supposed to be the new social and blogging tool for the next generation; literally a new kind of “pen”.
That marketing campaign went nowhere, mainly because the connection to the Internet was abysmal and workflow no different than using a DSLR (or even compact camera) back in 2009/2010. What saved Olympus as a mirrorless provider was dogged persistence coupled with the fact that the cameras and lenses they produced were small compared to most common DSLRs, yet competent. That plus the fact that they kept having fire sales on older models as they quickly iterated, making it easy enough for many to sample the product.
Next we went through a period when mirrorless cameras were about size and weight gains compared to DSLRs, and partly because of that their attractiveness to new audiences, particularly women and the Asian markets. I’m not sure that worked out differently. Some people bought into mirrorless because it seemed smaller and lighter, but many were doing so because mirrorless cameras kept selling at reasonable prices in fire sales. You don’t see a lot of fire red and pink cameras out there today, yet that was what was being pushed by camera marketing departments for several years.
So far we have “smaller and lighter.” And maybe “cheaper” if bought on sale. That “cheaper” part, by the way, is one of the reasons why I’ve written for some time that eventually most of the ILC market will head towards mirrorless: there are fewer parts, and especially few mechanical parts in a mirrorless camera compared to a DSLR. In theory, you can build mirrorless cameras with more automation, fewer alignment procedures, and with far fewer parts. Nikon themselves proved that with the Nikon J1 and V1 back in 2011. No digital camera I know of outside of some cheap compact cameras has been made with fewer parts than the 183 in the J1; by comparison, most DSLRs have thousands of parts. (Yet Nikon decided to hand paint every logo on the J1 ;~).
But is there anything else? Yes, I’ve come to the conclusion there is, and it’s the same reason that DSLRs took over from SLRs: solving a user problem. Or at least appearing to.
I’ve been challenging everyone that claims that a mirrorless camera does something their DSLR doesn’t for some time now, and then listening carefully to their answers. The data is now strongly pointed in the same direction, so I’m comfortable that I understand what’s happening.
As usual, it’s a workflow thing ;~). It’s a workflow thing rooted in a user problem.
Let me first explain one reason why DSLRs just skyrocketed away from SLRs so quickly: feedback.
The classic user problem for a film SLR user was this: they’d go out and shoot and have no idea if they had captured what they wanted. They’d drop their film off, and pick it back up an hour or a day or a week later and then look at the results. Dang! Didn’t get that right.
What wasn’t right? Typically exposure or focus or use of shutter speed (not stopping action or having visible camera shake). The problem, of course, is if this happened on a once-in-a-lifetime vacation or event, you could never redo the photo you missed. Worse: unless you had made a record of what you did (e.g. aperture and shutter speed), there wasn’t any way to learn how you failed.
An awful lot of film SLRs got tried by people—especially after they added autofocus—but then those SLRs went into the closet forever when people got frustrated with them (hey Tokyo, that sentence should rivet your attention, because you have the same type of problem today, but let’s move on…).
DSLRs solved the user problem I describe by closing the feedback loop. Take the shot, look at the LCD to see if you got it right. If not, change something and try again until it is what you want. Even though the LCD was often small and dim and didn’t have very many pixels, you could still see the basics well enough to make sure you got a shot you liked. Sure, you may still have missed a moment that won’t recur, but a lot of the mass consumer photography is static and posed and easily recreated in a second (or third, or nth) pass.
One thing I noticed at the start of the century in working with students, both amateur and professional, is that skill sets improved rapidly and dramatically when they switched from film SLR to DSLR. People that never understood exposure and couldn’t master it now did. Shooters that were having handling problems could experiment and figure out how to correct them on the spot. Focus was more carefully evaluated and tweaked, and depth of field suddenly became easier to see and control (SLRs stopped down lenses to show you DOF preview, but the view was very dim on a not-so-great screen).
In the last few years I’ve seen the opposite: lack of progress in students, both amateur and professional. Why? Because the camera controls and options and interdependencies are getting more intricate and nuanced, and nothing’s improved to help them master those things. Menus are deep, broad, and dense with lots of poorly worded options. Focus modes have proliferated into so many choices, most of which are unsatisfactorily undocumented. Heaven help you if you want to beam your images directly from the camera to Twitter (the Nikon D5 even has a full computer server in it to handle an Ethernet connection, but you’d better be an IT pro if you want to even begin to configure it for use).
So, let’s get back to that workflow thing that’s behind a lot of folk preferring mirrorless to DSLRs: mirrorless closes the feedback loop further. In particular, I get a lot of answers that talk about the “real time histogram” (more on that in a moment), and the more what-you-see-is-what-you’re-getting viewfinder. In theory, no longer do you have to take a shot, evaluate it on the LCD, change a setting and reshoot. No, good mirrorless cameras let you see your “results” as you’re shooting. Sort of.
Which brings me back to the “real time histogram.” It’s not accurate. Why? Because it’s not made from the data pixels of your shot, it’s made from the data stream to the EVF, which is not only a subset of the data, but is manipulated in various ways. Olympus OM-D E-M1 Mark II owners are starting to discover this, as the various EVF modes you can set will change the in-viewfinder histogram!
Still, I’ve heard this statement from enough mirrorless users now that I have to accept it as something that is compelling them: the instant feedback of a well-designed EVF is better to them than the shoot-then-evaluate feedback of the optical viewfinder DSLRs.
Of course, there’s no reason why Canon and Nikon can’t put histograms in DSLR viewfinders (both companies have overlay technologies that could do this). They simply don’t understand their customer base enough to understand that this is an unmet demand. But then again, they’ve never even understood that there’s been at least a 17-year demand for raw histograms, among other things ;~).
Circling back to where I began: many (if not most) of those that assert that mirrorless is the future are pointing to features of the mirrorless cameras that the DSLRs aren’t duplicating (or can’t duplicate in a few cases). And most of those things center on user problems. “What’s my picture going to look like?” Most good mirrorless cameras are showing you that even before you shoot. “Is my exposure correct?” Most good mirrorless cameras will give you a lie that you accept (or have at learned to judge well) in the viewfinder ;~). DSLRs can’t do those things for those users.
Coupled with smaller and lighter, these things all start to add up well enough that there’s absolutely a subset of photo enthusiasts that would pick mirrorless over DSLR. As other features of mirrorless—focus speed and accuracy in subject tracking, in particular—get closer to DSLR levels, the mirrorless cameras look more and more interesting to long-term SLR/DSLR users.