where micro-four-thirds is actually going
the journal of Michael Werneburg
twenty-seven years and one million words
I'm a daily-carry hobbyist photographer. I use the Olympus micro-four-thirds camera system. I do so because the results are great (for what I do) and the gear is fun and the prices are good. But I keep having strange videos suggested to me on Youtube in which some presenter screams the manufacturer release this or that fantasy product. Invariably, these mirror previous failed models. I want to explain why I think almost so many of these Youtubers are completely wrong and where micro-four-thirds is actually going.
I've been living in Tokyo for the past five years and frequent camera stores. I have never seen someone using any of the models that the Youtubers clamor for. The fabled Pen F, for instance. There are current all of eight used bodies of that 2016 model for sale right now at Map Camera in Shinjuku. That compares for instance with twenty-two used Nikon Z5 bodies -- a full-frame photo-centric (not video-centric) street-shooter's model from 2019. The Olympus Pen F was a failure in Japan, and I suspect that if it wasn't selling here in Japan, it wasn't selling anywhere. The current owner of the product line, a company called OMDS, doesn't have room for more of the market flops that drove Olympus (the corporation) to sell the product line. What I propose today is the opposite approach to repeating the mistakes of the past.
My basic premise is that OMDS today lacks a digital market strategy. I believe they have been relying on hardware sales, believing that "digital work flows" that suit professional photographers are good enough to sustain their business. But pros are different. They will invest their time in those work flows because it's how they earn a living. And serving pros can be a very difficult group from whom to make money. Yes, they will make the investment in a lot of gear, but they are relatively few in number and require a lot of support in terms of technical support, the production of rare lenses, and the need to be seen to be using a credible platform (a top name on the camera). Anyone selling electronics to pro users needs to supplement this with consumer-grade products that will pay the bills. And today, consumer-grade cameras have to compete with smart phones.
I believe that OMDS need to realize that smart phone cameras succeed because they are inherently a part of larger ecosystems that allow no-cost sharing and publishing. And that those ecosystems belong to people who don't care if electronics makers live or die. I would suggest that OMDS (and all camera makers) stop conflating the pro clients with the retail clients and employ separate strategies for both.
some trends
equipment
As much as I love using this photo gear, there are persistent problems.- The menu interface on traditional cameras remains terrible to a point of disbelief that anyone could be this slow to evolve. The menus are so bad I have second-hand embarrassment.
- Getting photos onto popular sharing platforms is still a mess.
- Lenses just keeping getting bigger and vastly more expensive.
- The manufacturers are changing the memory card format to suit videographers.
competition
Smart phones represent a generational leap beyond the offering from traditional camera manufacturers. They have ever more computational photography capabilities, allowing photos to look good under just about any circumstances. Great colors, good sharpness, excellent noise control, good low-light performance... Phones today offer exposure compensation, color correction, cropping, slow-motion, stop-motion, de-noising, software filters, semi-credible high-dynamic-range, a variety of annoying AI gibberish and the ability to retrieve photos where you actually miss the critical moment. I have a Pixel 7A phone from Google, the photos it produces are excellent and the colors in particular are great. But size is the limiting factor. The sensor can't get larger without making the lens in front of it larger, and the whole thing becomes unwieldy as a phone with a large camera block jutting from the body. Beyond the camera capabilities, smart phones offer total integration with every conceivable sharing and publishing venue. There is no point in attempting to enumerate these, they have permeated all aspects of private life, work, and everything else. For fifteen years, these super-cameras, which are effectively free with the phone, wiped out something like 95% of camera sales for the twelve years following 2010.As if the pressure from smart phones weren't enough, the last few years have seen a resurgence of interest in point-and-shoot digital cameras that sell well under ¥30,000, and in some cases under ¥20,000. I suspect that this is part of a move away from the all-digital experience of smart phones and the years-long resurgence in interest in low-res photography. I don't know how long it will last, but it certainly means that there is a floor to what any camera manufacturer can sell before they're competing with whoever is producing the all-but-disposable point-and-shoots from "Kodak".
Chinese manufacturers have produced all the new m43 lenses released in the past few years. They have even produced their first m43 camera models. If the story of Huawei coming back from the US "Chips Act" inside a year with credible products of their own is any indication, the Chinese lens/camera manufacturers will be capable of astonishing innovation. I see this as a second-generation challenge of equal threat to that posed by the rise of the smart phone in photography fifteen years ago. I predict that Chinese market entrants will rapidly produce camera models of their own and put incredible cost pressure on the Japanese manufacturers -- as they have with smartphones, laptops, and anything else that employs circuit boards.
And finally: In the long term, Metalenses also threaten the lens sales revenues of the manufacturers. With a single metalens, a flat piece of glass with a special construction sits in front of the sensor and the sensor's software is used to extract different types of images. It throws out everything we've been doing with optics for the past two-to-four hundred years and represents the advent of software-defined optics.
Succeeding in cameras is tough, maybe the toughest it's been since the Second World War.
it's all about software
I believe that if OMDS is going to survive this, they'll have to become a software company. This is a huge break with Japanese camera manufacturer tradition, but here's what I think it will accomplish:- To divert smart phone users to an improved, affordable platform.
- To differentiate between their system and the much more expensive systems with larger full-frame sensors.
- To differentiate between themselves and the Chinese manufacturers.
- To offer a credible cloud editing/storage/sharing alternative.
computational wizardry
In adopting something like Android or perhaps FreeBSD (as have some game consoles), they could produce highly engaging "just works" software that is fun to use and allows people to actually learn photography and how to make the most of the hardware. Smart phones do an okay job of this, but the sensors are pathetically limited and all the software accomplishment has in part been to work around those limitations. Today, OMDS has OI.Share, a smartphone app that can connect to your phone to take pictures and 'download' the photos from the camera for editing and sharing. It's a good start, and as far as these things go it's about as good as I've used. But it still feels tacked-on and very first-generation. And ultimately, it's not native to the camera.
My advice for OMDS is to bring all the same computational photography to their much-improved platform. To consumers of my kids' generation, all the wizardry and ease of sharing is photography. The sensor in my Pixel 7A is about 40 square millimeters in size. The m43 sensor is 220 square millimeters. That is a lot of real estate with which to compete with the smart phones at their own game.
I believe that this new improved software must include a native cloud sharing/publishing service (see below) and seamless integration with popular cloud/sharing systems. This would be unlike any of the proprietary systems that the camera manufacturers have always preferred. They have to stop pretending that someone wants to use something that the camera manufacturer wants them to use, and recognize that this battle was won twenty years ago: people want to use the ever-changing "big tech" platforms of the day and will never use some clumsy proprietary thing. Unless it's made by Apple. Which OMDS is not.
Camera sales will never recover the ground lost in the past twenty years. It will always be tough, and the advent of Chinese manufacturers means that the camera bodies are heading for commodification. There is no more room for market failures. And because the Japanese manufacturers are competing with a country with low salaries, they'll need to figure out how to contain costs. This is where accelerating the transition to software competition really helps. If your hardware lasts five to ten years, and your product evolution is through software releases to existing hardware, you need to make your hardware releases "future ready" with the software to come.
interface
What I envision is that they retain the decent build quality and the current approach to well-built, minimalist dial-based interface. But then they throw out everything they think they know about the digital interface and go with a phone-like interface. You are not going to attract new users with horrifying menus, it's just that simple. They need to support an app ecosystem, and they need to throw in the hardware to make this happen. This may sound ambitious but the cloud is too good an opportunity to pass up.the cloud
That's right, time to talk trends again for a moment. Olympus has a cloud service for registered pros to upload their stuff. They're 90% of the way there. But I think they need to make something friendly for all users.I think the days of free cloud services are numbered, and that a lot of photographers are going to need a home. Google is in trouble with regulators, and could be heading for a break-up. It would not surprise me if they are forced into some actions that will for instance limit their interest in open-ended gmail storage. Google has proven itself largely hopeless at extending beyond its advertising base and also had not created a convincing cloud strategy around the smart phones they sell.
Apple is a wildly successful company in almost everything they do, but I think that the world has learned from their example and that the days of their being able to charge exorbitantly for cloud storage and hardware are waning. My kids want me to pay for their Apple cloud services but I have refused on the basis that it is simply far too expensive compared to the cost of a PC or network-attached storage. I've been tinkering with computers for more than forty years and am likely an exception in that regard, but in one way I am the norm: I don't have excess "disposable income".
Which brings me to Adobe. Their cloud offering is very expensive but also quite limited. I have 20x as much photo/video storage on hand than I can store on their irritatingly expensive cloud, and the full cost of storing what I actually have is obscene.
Here's where manufacturers like OMDS stand to gain if they do what I am proposing here. If they adopt a ~5 year time frame for their hardware, guaranteeing software upgrades for that duration, they will need real-time feedback on what works, where problems are, etc. They will also need a massive amount of data to train their AI software. Where can they get the data for these efforts? By hosting a cloud service and putting something in the terms about using the data to improve their software. If that sounds familiar, it's because all software everywhere already has those terms.
in the short term
All of this will take time, but not a lot of time. Software development is now a better-understood discipline and is downright rapid (after some decades of struggling). In the meantime, the company would have to transition.fewer products
OMDS can continue to sell on its reputation for quality products. This will likely differentiate it from Chinese manufacturers for a time, but not forever. It would be nice to think that OMDS could bring out a killer street shooter with great auto-focus, good weather sealing, and a decent fixed lens. An improved Ricoh GR-series or improved Fuji X100-series. But I think such a model is technically challenging, will not be pocketable, and in any case has only a limited market. And because OMDS can't afford market flops, these things should be avoided.The continued near-absence of OM System cameras from the top sales lists here in Japan reflects uncertainty about the new owner and the exiting product line. It's the latter I want to touch on with just one thought: I believe that until the product changes significantly, the E-M10 entry-level product is not significantly different to prior models and that it will continue to under-perform the purpose of entry-level models: to drive brand interest and sales volume. The simpler E-P7 (an E-M10 with no viewfinder) is occasionally in the top sales ranking. Somehow, both products are still selling under the Olympus name. When that has to stop, I think the E-M10 should go away.
multiple platforms
I believe that Panasonic, the other major four-third player, is already exiting micro-four-thirds. They announced a new full-frame/APS-C lens "L mount" in 2018 and it seems to be taking off. The only recent camera bodies from Panasonic in the m43 space are the same as the full frame ones in size and shape. It's the same body but with the tiny m43 sensor inside.I believe it's time for OMDS to also join the Panasonic L-mount platform and sell more expensive bodies and lenses for that mount to generate revenues. Earlier, I mentioned the Nikon Z5. It's a compact full-frame body and is photo-centric, as Olympus products have always been. According to Map Camera, it's in the top-ten of all camera bodies they sell used. Compared with my Olympus EM-5.3 from several years ago, it lags in a few ways that suggest that Nikon purposefully limited its capabilities to protect their other products. The lenses for the Z5 are obscenely expensive, so much so that I cannot justify buying a used Z5 even though it is roughly the same price as a replacement for my current camera (which I bought used). I believe Olympus has an opportunity to compete with Nikon in this space by adopting the L mount. While Panasonic pursues video-centric options, OMDS can have the photo-centric.