It’s getting quite close now to all the at-Photokina announcements, and I expect a number of new mirrorless camera entries. So wrap your mind around this thought:
The very best mirrorless camera we currently foresee will be no better
than the very best DSLR camera we currently foresee.
Put another way, mirrorless is just a different way of designing the same thing.
As I’ve noted many times, mirrorless is an inevitable shift from DSLRs and their mirrors for the following reason: fewer parts and fewer alignment steps in manufacturing. Simple as that.
We can argue all day about whether an optical view or an EVF view is better, but that point is actually totally moot, as Fujifilm has proven with the X-Pro1/2 models, which have both ;~). Those that say you can’t get real time histograms in the optical DSLR viewfinder also are barking up the wrong tree. We already have real-time horizon indicators in DSLR viewfinders; there’s absolutely no reason—other than the Japanese not thinking of it, or perhaps not thinking it necessary—that we can’t have a functional histogram in a DSLR viewfinder.
The whole notion that first there were SLRs, then DSLRs, and eventually mirrorless cameras is a false belief. We had analog (film) cameras, we now have digital cameras, and that was revolutionary because it changed from chemicals to bits. But another revolution via mirrorless taking over from DSLR? Not really.
The Nikon D1 in 1999 marked a real change. Film cameras were already on the decline, and the D1 and competitive rush to digital it triggered was a revolution. Things changed. Dramatically. We’re now far better off than we were in film, in my opinion. Many people don’t understand why DSLRs took off so fast and soon produced sales numbers that far exceeded film SLR’s peak. It’s simple really: feedback.
With film you had no idea whether what you tried worked or not until you got the film back from the lab. If you were on vacation, that could be two weeks or more later. By then you’d forgotten the details of what you were trying, and the do/fail/retry loop was so slow that most simply gave up.
Digital solved that feedback loop problem. Take a picture. Pop it up on the LCD. Did it work? No? Then try something else. Rinse, Lather, Repeat (or rather Shoot, Adjust, Repeat). As crude as the LCDs were in the first days of DSLRs, they were enough to solve the basic problem for most. You could quickly evaluate exposure, focus, timing, all the parameters necessary to make the shot you wanted. Suddenly photography was fun again. Wait, what, a better camera came out? Great, I’ll buy it.
This “digital is great, digital is fun” thing continued right up until the iPhone. Technically, the iPhone wasn’t first. But Apple’s marketing machine coupled with an engineering team doing their usual “think like a user” design, started the ball rolling. When Apple quickly figured out that they needed to let others produce apps, that opened the way for Facebook, Twitter, et.al. to do the thing that suddenly made Japanese digital cameras not so fun any more: instant sharing of images. Here it is eight years later and I can’t point to a single Japanese camera company that’s got the sharing thing right yet. Some get it just convolutedly wrong.
Smartphones had all the attributes of digital cameras (e.g. instant feedback), but added something that the Japanese camera makers simply didn’t match. That’s surprising, because the first cameras in cell phones were made in Japan in the 90’s, so it’s not as if they didn’t have the hardware ingredients figured out. What tripped them up was software.
The current decline in digital camera sales has people constantly asking me what the next revolution will be (some suggesting mirrorless, thus my point, above).
Back in 2008 I suggested what it would be: not a change in analog/digital or a change in camera specifications per se, but a change in workflow. Simply put, digital camera workflow sucks in the modern world. Sucks badly. That instant sharing of the smartphones simply did the same thing to DSLRs that DSLRs did to SLRs: made them dinosaurs that looked cumbersome.
The next revolution should be as I described it in 2008: communicating, programmable, and modular. Each of those three things has the ability to not only transform workflow, but to bring it as far into the modern world as far as is possible.
I’m going to start backwards this time: modular.
Technically, modularity is the least needed of the three changes, yet it has potential for a profound difference in how we work.
One of the most common emails I get from readers goes something like this: “I really wanted a D5 sensor in a D800-type body, so what should I buy?” Well, modularity would have solved that problem. Two pro bodies, one big with the dual grip, one small and compact, both modular. All Nikon’s sensors available for either. Were that the case, almost all of us pros would have bought two bodies and three or four or more of the following sensors: 20mp high-ISO, 24mp general purpose, 36mp high resolution, and all in color, monochrome, and UVIR versions. We’d be buying new sensors as they get better, and we’d still buy new bodies when things like the D5’s autofocus system come along and improve the basic body experience. Somehow I don’t think Nikon can prove that they sold more D4/D4s, D800/D800E/D810 models than a modular combo would have. Indeed, I’ll bet that the dollar output from pros and serious enthusiasts would have been higher.
Why? Because it solves so many problems in our workflow.
One day I’m shooting sports at night and need a sensor that’s tops at low light. The following week I’m out shooting landscapes and need all the resolution I can get. Next month I get an RFQ for a monochrome shoot. Yeah, got it all covered (and more) with a modular system.
I also originally outlined the communications as needing to be modular, too. If I’m shooting for a newspaper, they want JPEG delivery Right Away and cellular probably is the answer. At events, I want a fast Wi-Fi system that can talk to the server I set up at the reception (slide shows, etc.).
Let’s be reasonable. From day one of DSLRs until today what have been the things that have really changed? Sensors, communications, and card sizes/speeds. Those are the things that should have been modular. Viewfinders, autofocus systems, controls, LCDs, etc.—those were pretty good on the D1h and frankly a new body with improvements every four years would have been perfect.
Let’s move on.
A lot of assignments (self imposed or real) end up with shot lists. Exactly where do I store that on my camera? Oh, right, I look at a list on my smartphone, which is becoming a better camera every day. Oops. Why can’t I send a list to my camera and tick it off as I shoot? Heck, if I’m ticking with shot IDs, why wouldn’t I also be tagging those same images?
Funny thing is, this idea came to me back in 1977. Say what? Well, that’s when I was starting to shoot video news using some of the early ENG cameras. Hey, I shot three takes in the camera, why can’t I mark each one so that when I get back to the editing station I don’t have to blindly search for what I’m going to stick into the edit? The fact that we really don’t have a good solution for this even today tells me that the engineers really aren’t looking over the shoulders of users.
Even some of the products that get it partly right don’t get it completely right. The Blackmagic Design cameras, for instance, allow you to enter take information into the metadata that follows the clip. But the notion of “I’ve got four takes and take three is the one to use” doesn’t have a component in the software. Nor can I pre-enter my scene information.
The reason why I write that “programmability” is a necessary function for the next camera revolution feeds from that last paragraph. The people that know what we want in our workflow best is, well, us.
Let’s expand all this list/tag thing to a wedding photographer. The shot list that sits in the camera came from the pre-wedding meetings with the bride and her mother. As I’m shooting, I want to be able to pull up that list, and after I’ve shot something I want to be able to tag it as to where it should go (e.g. live slide show at the reception, best take for post processing, save but don’t prioritize, etc.).
I’d really love it if my list also provided key IPTC information, such as subject names. Let’s move to the football game and see how that works. I want the event information, location, and maybe even the quarter to be in the IPTC information already. I’d love to be able to chimp between shots and tag player numbers and get the player team and name into the IPTC. Why? Because I have to supply that to the news services at some point. I don’t want to be spending hours after the game typing things in manually.
I should point out that Nikon did add IPTC information (D4, D4s, D5, D500) after I pointed out what sports shooters were really doing. They even gave us the ability to enter it with a touchscreen (D5, D500). But it’s still not a perfect solution, and I have no way to tag players using IPTC. My photo editor wants a caption with my photo when it receives it on deadline. He hasn’t the faintest idea who the players in the shot are or what they’re doing. One solution we’ve been using sometimes on the D3/D4/D5 cameras is to record an audio file with an image, but that’s adding a workflow step down the line, and you’d better keep the JPEGs and audio files aligned, which I’ve found that photo editors don’t seem to do well ;~).
Of course, in both those examples I just gave, there’s an element of communication suggested, which is my last point. My photos should automatically go to all the places where they need to be. Not just one place, but all the places. All my images should go into my cloud storage. My best images (tagged during shooting) should be ready in a folder for post processing on my desktop. Particular images I picked out while shooting might go to Twitter, Facebook, Instagram, a wire service, to a client, or somewhere else as I’m shooting. All images from a shoot should be in a folder for the shoot in my images RAID.
This is another reason why the camera needs to be programmable: all the services I use and options I need change constantly; photographers and imaging companies should be able to write or commission an app for our cameras. Certainly the camera companies won’t be able to do it, let alone keep up with the constantly changing environment. Even after presenting my needs directly to Nikon, I’m not 100% sure if they fully understood everything I presented them. Moreover, some of my needs are unique to me, and Nikon isn’t going to put their “crack” team of programmers on solving something only for me.
Ironically, we’re seeing more companies try to tackle some of what I write above in the smaller video camera industry than in the bigger still camera side. Indeed, companies like RED went modular from the get go, companies like Blackmagic Design have been seriously looking at the programming side. As I learned at NAB back in April, everyone is looking at VOIP as one critical way to move video data over virtually any communication channel.
The still camera companies? Ricoh abandoned their modular camera (GX). “Programmability” seems to mean “enter some meta data” to most of camera companies, though it doesn’t always go the right place ;~). And communications? Gee, Wi-Fi is proving to be difficult for them. It took them forever to solve the quick connect problem, but no one yet seems to want to tackle both AdHoc and Infrastructure use.
Unfortunately for us still photographers, the Japanese camera companies are still taking baby steps. They’re fully locked into their mild iteration each generation, major iteration every few generations mantra. And yet the “major” isn’t all that major now that smartphones have revealed just how much they have left to do.
All this is why I don’t think “mirrorless” is the answer to all our problems. It’s just an iteration step. I’m sure both Canon and Nikon have a big planning document that shows just how far and fast that iteration goes for each of them, and how their DSLR lines slowly morph into mirrorless.
But I’m here to tell them that this doesn’t solve their bigger problem, no matter how well you design a mirrorless camera.
I’ve written this before, but each significant new surge in still camera sales has been because a big user problem has been solved: exposure meters, autofocus, and digital. Those were all big leaps forward that solved a problem that those that were interested couldn’t easily master. Smartphones just solved an even bigger problem: workflow to share an image. It will take something bigger than that to grow the camera industry rapidly again. Maybe to grow it at all.