DSLRBodies

At this point you can find the basic claim I first made back in 2007/8 all over the place in the media: smartphones are cannibalizing the camera market. Or have cannibalized the dedicated camera market. Or continue to cannibalize the camera market. Basically, smartphones are cannibals, apparently.

Of course, when I first began writing that, the data was slim to support my contention, and I was actually putting out a straw man thesis partly based upon personal experience I had in Africa in 2007 with an iPhone and what I knew was going on in Silicon Valley at the time (e.g. future mobile products, both hardware and software).

Today, the numbers don't lie. Only, the numbers are correlational, not necessarily causal. All the statisticians among you know what I mean by that. Just because the smartphone market has climbed to the billions of units while the camera market has contracted to just less than 20 million units a year doesn't mean that smartphones are the cause of the camera contraction.

Aside: anyone looking for a PhD thesis should consider looking at household penetration rates for cameras versus new camera sales, and how those have completely changed in the last 20 years. It's a data point about new household penetration rates that provoked today's new straw man article.

In any consumer business, you have to pay very close attention to trend lines. In fashion, for instance, things like styles, materials, and colors change all the time. If you don't correctly anticipate what mood the consumer is in and how their predilections are changing, you'll end up with a lot of inventory you have to fire sale to get rid of (and that's the best case scenario ;~).

In the auto industry, the contradiction between gas prices and convenience has been making it tough for the auto makers to rationalize consumer trends perfectly. Here in the US, everyone wants big SUVs and trucks, but when gas prices skyrocket, the demand for more compact vehicles goes back up (another interesting thesis project). 

So what are the full set of trends that have been impacting dedicated cameras? Actually, quite a few:

• Size/weight — A whole bunch of sub-factors enter into this basic trend. First, we have the aging of the interchangeable lens camera owner. As they get older and older, the idea of carrying a five pound necklace around all day starts to become a big factor in decisions about new gear. Even two pounds is pushing it now for the 60+ crowd. Yet those are the ones with the time and disposable income to buy a top end camera and use it regularly. The airlines haven't helped things for the camera makers. There was a time when my carry-on camera backpack tended to weigh 40 pounds on most trips. The basic International weight limit for a carry-on these days has dropped to 15-22 pounds (7-10kg). That's almost the weight of my D5 and 400mm f/2.8G when you throw in an extra battery and a couple other goodies. Then there's what the camera makers have done to themselves: why should I carry a heavy 24-70mm lens around for my DSLR when a Sony RX100 that fits in my shirt pocket can hold its own much of the time and for many photographic purposes? 
• How images are shared — This trend was clearly evident before the turn of the century. Amazingly, it was visible first in Japan, where the initial camera-enabled phones appeared and the carriers had to figure out how to deal with data and where users put their images. The camera companies didn't catch onto this trend then, they didn't catch onto it when the iPhone appeared, they started to think about it when all the Silicon Valley companies began to roll out cloud image storage and social networking took off. The Japanese camera companies eventually and begrudgingly put older, slow communications chips into their cameras to give lip service to the idea that an image might somehow be communicated rather than printed. Realistically, the camera companies are well over a decade behind the trend.
• Reliance on displays — Related to how images are shared is how they're displayed. Today, images are increasingly being displayed on electronic displays, not on paper. That's becoming true for even large forms, such as billboards. Those displays are not 4:3 (sorry m4/3 users), nor are they 3:2 (sorry 35mm SLR and DSLR users). They're 16:9 in the most common form, and I'm starting to see even wider ones now. It amazes me how few cameras have the ability to shoot stills directly into the 16:9 aspect ratio. It really should be the default, not the option. Then there's the issue of how your image gets from the camera to the display itself. Yes, I know I can now hook an HDMI cable between my camera and display. But cables? That's so 20th Century. Has any camera maker even considered something like Apple AirPlay? Apple certainly has considered it: with AirPlay 2 we now have four major television set manufacturers (LG, Samsung, Sony, and Vizio) who allow Apple devices to wirelessly connect for display, which, of course, includes iPhones. 
• Sontag's thesis — If you haven't read Susan Sontag's On Photography, you should. Whether you agree with her ideas or not, they'll provoke a great deal of thought on your part (or should). At the risk of greatly over-simplifying and slightly changing one of her message: we all stand on the shoulders of what came before us in photography. It's no longer good enough to just take a picture of Half Dome in Yosemite from the valley floor as millions have done. It's important for us to build on that and show our own experience/feeling/thoughts. One of the ways we can do that is if our photographic gear evolves to allow that (my thought, not Susan's). One interesting thing that emerged in the last decade is the selfie: an image of you in the place rather than just the place, with you becoming more important than the place. The camera makers have only partly acquiesced to that (e.g. selfie-capable rear LCD viewing), but one wonders how many other things that are photographic trends exist that the camera makers are slow to pick up on. Curiously, both Apple and Nikon have tried to do the opposite, and give us something new, as with that Harry Potter-esque still and motion together idea, but we didn't embrace it. But we need to encourage that sort of experimentation and insist that the camera companies catch on to the things we actually do with our cameras faster, too.
• Moore's Law — The number of transistors doubles every two years. The things that are affected by that are: pricing, memory capacity, speed, and more. While the basic unit of the image sensor—the photo diode—doesn't really gain from semiconductor size decrease, the supporting electronics certainly do. The problem for camera makers is that in order to keep up with Moore's Law, you need market size. Otherwise the increases in R&D, test, and manufacturing costs go up in ways you can't recover. Thus, if you look at the current A12 Bionic chip Apple uses in the iPhones you find virtually every advancement you'd expect from Moore's Law. It's a fast 64-bit 6-core CPU, with an 8-core neural network, with a 4-core GPU, while being very low in power use. There's not a dedicated camera that comes close to that with their SOC (System on Chip). As we move further into computational photography, the chip at the heart of the device becomes more and more important to be state-of-the-art. And it's not just at the core of the system that the camera makers have been lagging. They're still using older and slower interface systems. There's not a camera out there that can write to a card as fast as the fastest card can handle, for instance. And don't get me started about Wi-Fi and USB speeds.
• Longevity of product — Cameras have always been relatively rugged and reliable. There's long been the notion of the "serious camera in the closet." In other words, you bought one once, and it comes out for some special occasions, events, vacations, but otherwise it sits somewhere not actually in use. You keep it, because it's good enough and reliable enough, and probably all you'll have to do is recharge the batteries and remember how to use it (;~). For that crowd to want a new camera they have to be convinced that there's something they need photographically that they can't do with the camera-in-the-closet. What the camera makers never figured out is how to sell you anything other than an additional lens for that camera in the closet. Even something as simple as a real, tangible, paid firmware update seems to escape their ability. Heaven help them if they had considered more modularity than a lens mount.
• Need to customize — We're all programmers today. No, not nerd coders that are running around deep in some arcane programming language, but rather we are customizing all our products to our specific needs more today than ever before. Sometimes it's just vanity and pride, such as putting a different background image on your screen. But more and more often, it's pruning out things you don't use, and elevating the things that you do use to primary position. Quick, which apps are on your mobile device's home page, and why? Camera makers have enabled a fair amount of customization over the years, but it's not necessarily the customization the user wants. Why is it, for instance, that bracketing is a dedicated thing specific to only a few items and choices? Why can't I bracket noise reduction, for instance, and do so at a near continuous set of levels? Or bracket in-camera lens corrections? The reason is that the Japanese companies all design paternally: they know best what you need. Or so they think. Nikon owners these days are asking why the D850 allows AF-Area mode to be changed automatically with a button press (AF-On, thumb stick, etc.) while the Z7 doesn't. That's paternalism at work. Z7 owners don't need that, says Nikon. Unfortunately, that's the opposite of the trend. The trend is that we customers want more customization and control, not carefully selected customization and control determined by salarymen sitting in offices in Tokyo.
• Simplification of building — Tech products inevitably want to get rid of mechanical parts and reduce the overall parts count via semiconductor consolidation. They also want to get rid of alignment procedures and screws. Most DSLRs had on the order of 2500 parts, many of them screws, and multiple complex alignment procedures. It was inevitable that the camera companies would want to get rid of parts and alignments simply from a cost standpoint. That's exactly what the shift from DSLR to mirrorless is all about, actually. You could build a DSLR that does everything the best mirrorless camera does today, but it would actually take more parts than are in the current DSLRs to do it. You wouldn't get rid of manufacturing complexities, you'd add to them. Not the trend that the Japanese would want to follow. In the best of all worlds, cameras would get down to one display, one sensor, and one SoC (System on Chip). Of course, we won't get that far, but that's the trend line the companies want to pursue with camera designs: reduce complexity and parts.

So what if we put all these trends into one? Hmm, we'd have the Communicating, Programmable, Modular camera I outlined back in 2007. 

But to be specific, we need:

1. Something smaller, lighter, and more convenient to carry.
2. Something that ties directly into our social networking, email, and home computers, and offers a more convenient and automated workflow.
3. Something whose on-camera display can be completely mimicked on any display we wish to plug into or wirelessly communicate with.
4. Something that lasts a long time, but can be "updated" either by firmware and/or modular components (beyond lenses). 
5. Something that we can customize to our use and workflow, both in shooting and sharing.
6. Something that's simple to manufacture and thus reliable.

Of those, note that smartphones score big on #1, #2, and #6. #4 and #5 are part scores for smartphones. Cameras?

1. We're seeing some reduction in weight and size for the equivalent camera. Lenses, however, are turning out to be a problem (other than things like the PF lenses).
2. The camera makers have all now basically got a camera-to-smartphone connection that works. Slowly. Pulling the phone off the Wi-Fi connection it had. Without much in the way of helpful automation for sharing.
3. HDMI cable is all you get for the most part.
4. This one is difficult to judge at the moment. The transition to mirrorless has everyone scrambling to get their firmware up to snuff, so we do get substantive firmware updates from some. But I fear that's a temporary thing that's mostly just the software side catching up with the already released hardware. And, of course, no one makes money from that, so there's no incentive to do it. Catch-22.
5. Just being able to assign any function to any button was a start (and still not met by some companies), but it's only a start, unfortunately. There's almost nothing to help us with workflow, and the attempts to let us simplify (e.g. MyMenu or Quick Menus) are actually additions that add some complexity, not remove it. True customization would mean that our cameras do just what we want them to, the way we want them to, and everything else that they could do disappears under the covers (we still want to get down there and do things under the covers sometimes, but we don't want 100+ items clogging our menus all the time).
6. Bingo. The camera makers actually are on this trend like a chicken bone. That's because there are cost savings to be had in doing it, not because we customers wanted it.

I've written for over a decade that ILC volume will continue downward and won't go back up into growth territory until the camera makers embrace something, anything, and preferably everything that would make most of the closet-camera owners believe that they needed to replace their existing gear. Mirrorless isn't quite that thing, though the what-you-see-is-what-you-get EVF is enough for some, so we're getting a little blip there. 

It still seems clear to me, now 12 years later, that that thing that causes people to buy a new dedicated camera is how the image gets from the scene in front of you to the display in front of your audience. I don't know how to express that thought any simpler. I just had someone yesterday complain to me about how terrible it is to process a bunch of NEFs in Lightroom. That despite the fact that Lightroom has some of the best batch processing capabilities around (process one image, apply to all from a session, tidy up the few that still aren't right).  

The ball is still in the camera maker's court (and has been for over a decade). I have to assume that they are fine with market contraction, because I really don't see them pursuing the things that would reverse the market.