There's one important point that didn't quite get the recognition it needed, IMHO. And that is, sensor technology is the primary driver for both digital photography and videography. Until I retired at the end of 2009 from a major chip design automation software company, I was in the business of marketing lithography and DFM (design for manufacturing) software to chip companies, including the sensor guys (Sony, Samsung, Sharp). Even at that time, Sony was massively investing into image sensor technologies, far more than the others. In fact, Sony really went for broke, likely driven by Kazuo Hirai.
Today, Sony is 3+ years ahead of everyone else in the field with its backlit stacked CMOS sensors, and maybe even 5+ years when you consider some of the cutting edge work Sony is doing, like the Trichromatic and curved sensors. The field includes Canon, Panasonic and Samsung, the only other camera companies that makes their own image sensors, and a handful of independent sensor makers (CMOSIS, Jazz, a bunch of Chinese companies at the very low end). Sigma was a pioneer with the Foveon, but could not make it mainstream.
Canon makes a range of sensors from crop size to full frame, but only for their own use - they are not in the sensor business. If they were, they would not be competitive. A lot of their dynamic range and noise is fudged in firmware.
Samsung fell behind and got out of the photography business, but makes sensors for its phones and tablets (and likely sells sensors to other phone makers). APS-C was as big a size as Samsung ever got to, but AFAIK, they're now down to only phone / tablet-sized sensors.
Panasonic makes its own sensors, but is limited to the MFT size max. I don’t think they have any APS-C sensors, and they definitely don't have any full-frame. I don’t know what other companies orbit around Panasonic, other than Leica, which gets complete cameras designed by Panasonic (e.g., the D-LUX, V-LUX, etc.), not just the sensors.
Olympus got into a serious scandal a few years ago that resulted in many senior execs being fired, big time fines, legal expenses and other financial losses. They have never recovered from that, in spite of making some superb lenses for the MFT, and being a pioneer in mirrorless way back in 2010, along with Sony. They have stabilized themselves by building a close partnership with Sony that gets them a solid source for sensors, and also revenues from cross-licensing (e.g., the 5-axis image stabilization Sony uses came from Olympus).
Pretty much all other serious camera makers are Sony vassals - they get their sensors from Sony, and they are pretty deeply married to the Sony sensor technology roadmap, which pretty much puts them at the cutting edge of innovation (e.g., the new 150 MP and the 100 MP Trichromatic and Achromatic medium format sensors).
So there's a pretty impressive list of camera makers that have made deep, long-term commitments to being a part of the Sony eco-system. An image sensor is not like a commodity chip like a DRAM that you can easily switch vendors with; a camera maker has to make much deeper commitments in terms of capture firmware, noise processing, autofocusing, buffering, sensor read out to LCD display, and even lens design, which are all increasingly integrated with the sensor these days.
The Sony universe now includes everything from the ultra-high-end (Phase One, Hasselblad, Fujifilm, Pentax, all medium format), to 35mm full-frame (Nikon, Pentax, Sony), APS-C (Nikon, Sony, and probably, Fuji), 1" (Sony), 1/2.3" (Sony, possibly others), and phones (Apple). Sony doesn't have any MFT camera offerings, but it wouldn't surprise me if Sony were the supplier to Olympus.
I don't know who Ricoh got sensors from for its cameras.
Sorry for the long preamble, but with that long preamble, here's the way I look at the various camera makers:
- Sony: in the driver's seat, sets the pace. New cameras have been coming out at an amazing pace, and that is likely to continue. It's incredible that there are SEVEN versions of the RX100 all still shipping today: the original RX100, the Mark II, III, IV, V, VA and VI.
- Phase One, Hasselblad, FujiFilm, Nikon, Pentax, Olympus, and potentially now Zeiss, are all value-added resellers of Sony sensors. Their value added is their camera design, proprietary CPUs used inside, other hardware included in the camera (DRAM buffers, flash, other things like GPS, vibration sensors, flash transmitters, etc.), firmware, LCD, camera functions, features and controls, and of course, lenses, flashes and other accessories to build out a system, performance/price, range of offerings, and last but not least, service. There is an awful lot of scope to add proprietary value. Nikon is showing it now with its Z cameras, even if the very first efforts might be lacking a bit.
All these companies can march at their own pace, gated only by two items: Sony's sensor wavefront and Moore's Law. The rest of it is up to their own investments into R&D, marketing, distribution and service. Not to mention vision, management competence, and operational execution. (Pentax, did you hear that?)
- Canon: Ditto as above, but gated by its own sensor technology. At some point, Canon will have to bite the bullet, decide to invest hundreds of millions / billions of dollars into image sensor technology to at least somewhat catch up with Sony. My guess is, this is happening quietly in the background. There is too much at stake. It is unlikely Canon leapfrogs Sony in two years - Sony has a decade of learning curve under its belt. But if Canon can be somewhat competitive, that should at least buy them time.
- Panasonic: Probably continue to thrive in the MFT space, with a video-centric vision. They have a solid installed base, and they can continue to do well in it, while supplying Lumix designs to Leica. But I don't expect anything beyond that (APS-C, FF, etc.)
- Ricoh: No idea. Perhaps join the Sony ecosystem? It's probably too little, too late already.
- Sigma: Doomed to be a niche player, but even so, there are some creative things they could do, instead of endlessly limping along as they have been doing for nearly a decade now. For instance, I have no idea why Sigma can't put together a simple Foveon Landscape camera bundle which does only one thing: take fantastic landscape photos on a tripod. Make a camera with a full-frame Foveon processor, a superb fixed 16-35mm f/4 lens, high res EVF and touch-sensitive LCD with superb live view, sensor and lens stabilization, and ISO frozen to a 50-400 range. It does only one thing, take landscape photos, but it does it brilliantly.
- Notably missing "L" Word, Leica. Doomed to source lower resolution sensors from second-tier sources like CMOSIS. Their low volume assures high costs for sensor technology that is years behind the leader, continuing the trend of introducing cameras that are obsolete on day one in the eyes of anyone not an ardent Leicaphile. Panasonic could become a more formidable competitor in photography by acquiring Leica, but they would need to make serious investments into proprietary sensor technology, which I suspect is what has kept them from picking up Leica.
....
When I got into mirrorless back in early 2011, I bought an NEX-5 and sold my M9. Even the APS-C NEX-5 was already such a nicer camera to use than the M9, in spite of not being able to use the M lenses to their full capability. I had a choice between the Sony and the Olympus EM-1. The guy at the local camera store was heavily pushing the Olympus, because it was a more complete system at that time, but for me it was a no brainer: go with Sony! Just based on the sensor technology and roadmap. It turned out to be the right decision.
