There have been some numbers floating around online (from an article in the Japanese business publication Nikkei.com), showing shipment rankings for 2019 among the leading camera companies. Nikkei numbers show shipments (not necessarily sales – they include cameras sitting on store shelves), and they are worldwide. They are a lagging indicator – these are from 2019 – but relatively accurate.
We also often see BCN numbers, which come in much faster, but have serious flaws. BCN looks only at the Japanese domestic market, which is not proportional to the world market (mirrorless market share is much higher in Japan, and tiny cameras sell especially well). It also captures only about 40% of the Japanese market, and my understanding is that it is skewed towards chain consumer electronics stores (as if someone sampled the US camera market by looking at Best Buy but excluding B&H).
Nikkei shows two rankings – one for all interchangeable lens cameras (DSLR plus mirrorless, possibly plus a tiny number of Leica M’s which aren’t DSLRs, but aren’t conventionally mirrorless either), and the other for mirrorless alone. I am not clear whether these are just Japanese companies, or whether other companies are included. Most camera companies are Japanese, so the practical difference is whether Leica is included or not – Hasselblad and Phase One are so small that they’d barely register.
The overall Nikkei interchangeable lens camera numbers show a familiar picture:
Canon 4.16 million Nikon 1.73 million Sony 1.66 million Fujifilm 500,000 Olympus 330,000 Others 280,000.
Canon is a behemoth, Nikon and Sony are more or less tied, with Nikon slightly ahead of Sony, Fujifilm is about 1/3 the size of Nikon or Sony, and Panasonic, Pentax and maybe Leica combine to ship about 1⁄2 of what Fujifilm does… No surprises there (except, perhaps for just HOW big Canon is).
The numbers that have generated far more comment are mirrorless alone:
Sony: 1.65 million Canon: 940,000 Fujifilm: 500,000 Olympus: 330,000 Nikon: 280,000 Others: 240,000
At first glance, it looks like Sony is doing REALLY well in the mirrorless business, and Nikon’s Z series is a flop. That is a misinterpretation of the numbers for two reasons – Sony IS doing well, but Nikon is not struggling nearly as badly as those numbers suggest.
First, Sony (and Fujifilm, Olympus and Panasonic) sell essentially nothing BUT mirrorless cameras (Sony sells about 10,000 A-mount bodies per year, but that’s just selling through old stock). Canon and Nikon also sell large volumes of DSLRs, including entry-level models that account for the largest share of their total cameras shipped. Every camera company, including Sony, sells more inexpensive cameras than expensive ones. Sony’s inexpensive cameras are all mirrorless, while none of Nikon’s were in 2019, so the mirrorless comparison is comparing Sony’s volume sellers (plus their expensive models) against some of Nikon’s most expensive models. Nikon has the same kind of volume in D3500 and D5600 DSLRs that Sony has in A6000s and the like.
Even looking at B&H’s list of best-selling Sony mirrorless cameras, half of the first page (top 24) are APS-C. B&H is a professional camera store that hugely oversamples full-frame buyers. Looking at Best Buy (which is more representative of the consumer market) four of the five top selling Sony models are APS-C, and the top two sellers are both kits with the six year old A6000. The lone full-frame model to crack the top five at Best Buy is the A7 II, introduced in 2014. At Amazon, the top seller is the A7S III (probably due to preorder volume), but the next four are all APS-C. There’s no way to tell how many of those Sonys are full-frame, how many are upper-end APS-C and how many are A6000s and A5100s sold at fire sale prices.
Nikon was selling mirrorless cameras only in a very specific market segment in 2019. Just about every mirrorless camera they sold was priced above $2000 with a lens (and pretty much every one was sold with a lens, because the line was new and nobody had Z lenses).
Apart from a few Z50s that appeared at the end of the year, every one of those 280,000 cameras was full-frame, either a Z6 or a Z7. Comparing sales of the Z6 and Z7 alone to sales of everything from the A5100 to the A7R IV and A7S III is not really relevant – the bulk of any manufacturer’s unit sales is going to be the inexpensive “Best Buy” models, even if they represent a smaller share of value and an even smaller share of profit.
Although Nikon’s 2019 mirrorless sales numbers look similar to Olympus’, their markets are vastly different. Many of Olympus’ sales were sub-$500 PEN and E-M10 cameras. It is impossible to know what percentage were composed of these low-end products, especially because many of the PEN models are built for Asian markets and barely sell at all in North America, making Amazon and B&H US sales trackers next to useless for those models. Multiple kit variations of all four current Z models (as well as a large number of Canon, Sony and Fujifilm cameras) are presently outselling any Olympus camera at B&H.
Nikon sold 280,000 mirrorless cameras in 2019, at an average price well over $2000. Depending on the mix of Z6 versus Z7, and on how many buyers bought an additional lens or an FTZ adapter (which wasn’t free for most of 2019), the average selling price per kit may have approached $3000 – territory normally occupied only by Leica. Olympus sold 330,000 cameras in 2019, but I wonder how far over $500 their average selling price per kit was.
Of course, Nikon’s average selling price for all cameras (including DSLRs) is much closer to the industry average, which is around $500 per body, with included lenses pushing that figure up somewhat – but that’s on 1.7 million cameras, not 280,000. Nikon does need to increase their mirrorless volume, and the way that’s possible is by convincing midrange Nikon DSLR users to switch to Nikon mirrorless instead of switching to Sony. If they succeed, their mirrorless volume will go up substantially, but their average selling price for mirrorless will go down – those switchers from a D5300, D7100 or D610 will be buying Z5s and Z50s, not Z7s with multiple lenses.
From an online serial number tracker not run by Nikon, it is possible to get a minimum estimate of Nikon sales by model (minimum because the highest serial number for any given model and region is unlikely to be reported to the tracker). This information doesn’t seem to be available for Sony or Canon. Looking at serial numbers, the D850 sells at least 83,000 units/year, with the D750 and D610 adding another 110,000 or so each. With much smaller numbers of Dfs and D5s (the D6 was introduced in 2020), Nikon sold something like 350,000 full-frame DSLRs and another 40,000 or so high-end APS-C D500s in 2019. Add those to the nearly 280,000 Z6s and Z7s for somewhere between 600,000 and 700,000 non entry-level cameras in total for the year – somewhat over 1/3 of Nikon’s total sales .
There is a commonly seen number that full-frame is only about 10% of the total interchangeable-lens camera market. At least for Nikon, that’s not true. Full-frame mirrorless alone, broken out in the Nikkei numbers since essentially all Nikon mirrorless in 2019 was full- frame, is around 15% of Nikon’s total sales volume, and estimates of full-frame DSLR sales from serial number data indicate that FF DSLR sales are almost certainly equal to or greater than Z series sales.
Any similar number for Canon or Sony would involve much more guesswork, because similar serial number trackers don’t exist. I wouldn’t be surprised if the model mixes were somewhere in the same ballpark, especially for Sony – Canon may be more skewed (as a percentage) towards the wildly popular Rebels. Canon’s much higher volume almost has to come with a lower full-frame percentage, but there is no good way of estimating that percentage. If I’m reading my very new A7r IV’s serial number correctly, it’s about the 100,000th unit produced (and that model has been out a year or so). If there are about 100,000 A7r IVs produced per year, 600,000 or 700,000 full-frame bodies passes the “sniff test”.
From an initial read of “Sony mirrorless outsells Nikon mirrorless by something like five to one”, a closer examination of the available data shows that they are much closer than that. Nikon sells a bunch of expensive mirrorless cameras, quite a few higher-end DSLRs and a ton of D3500s. Sony sells a lot of cheap mirrorless cameras along with quite a few expensive ones. The real story is that they each sold just about 1.7 million cameras in 2019, and something like a third of those were full-frame. Most of the rest are entry-level cameras (the D500 and D7500 are relatively slow sellers, for example).
In a broad sense, Nikon’s and Sony’s camera businesses are relatively similar. Sony’s is probably healthier, but incrementally so, not by the 5:1 margin that is suggested by the top line in mirrorless sales. Sony’s big advantage is that essentially their entire customer base is already converted to mirrorless, while Nikon (and Canon) are still working on theirs, and are in the relatively early stages. That advantage could also be a disadvantage for Sony – there are a lot of Nikon and Canon DSLR users out there, and Nikon and Canon’s job is to convert them to a system where they can bring their lenses and already know how the cameras work, while Sony’s is to convince them to switch systems.
There are other interesting numbers here – one is just how dominant Canon is. This suggests that Canon interchangeable lens cameras outsold Sony and Nikon combined in 2019. This has been true for a few years if you look at all cameras, but I wasn’t aware it was true for interchangeable-lens cameras. Canon has a substantial compact camera business ranging from $129 specials to $1000 models, Sony sells some higher-end compacts, including RX1 models over $3000 and the popular RX100 series, and a few lower-end models, and Nikon has more or less given up on compact cameras except for a few ultrazoom and waterproof models. According to these Nikkei figures, Canon outsells Nikon and Sony combined even when you exclude compact cameras.
The 940,000 mirrorless cameras Canon sold in 2019 include everything from $400 clearance EOS-M models on up to the most expensive EOS-R kits. I haven’t seen any numbers on the breakdown of EOS-M versus EOS-R. This is further complicated because the EOS-M line may well be much more popular in Asia than in Europe or North America, making North American trackers misleading. Very small, viewfinderless mirrorless cameras have struggled to find a market in the West, while racking up excellent sales in Asia. The one piece of evidence I’ve seen that makes me suspect that Canon’s EOS-M sales are regional is that Canon’s share of the Japanese mirrorless market (BCN rankings) is always much higher than their share of the global market (Nikkei).
The 3+ million Canon DSLRs sold in 2019 are almost certainly largely Rebels, but there’s no way to tell what the percentages are. Nine of the top ten selling DSLRs of any brand on Amazon are various kits featuring one Rebel or another, and the best-selling non-Rebel Canon is #19. B&H, as usual, has a very different list of top sellers with quite a few 5D IV and 6D II variations mixed in with the Rebels.
Fujifilm has a solid business selling half a million X and GFX digital cameras annually. They break their camera business out in their annual report, and they further break the digital camera business out from Instax and disposable cameras. In most years, the X/GFX digital camera and lens business is profitable, and it’s not just barely profitable, either – they make a decent return on it. That probably won’t be true in COVID-ravaged 2020, but there’s no reason why it won’t be in 2021.
The other interesting number to look at is the “other” category. If it’s Japanese brands only, it includes Panasonic, Pentax and Sigma. If it includes non-Japanese companies, Leica and a small number of Hasselblad and Phase One cameras are in there as well. The easiest cut to make at this data is breaking out Pentax. Conveniently, every interchangeable-lens camera Pentax makes is a DSLR, and the only non-Pentax DSLRs that might be in the catch-all category are ultra high-end body and back combos from Hasselblad and Phase One. There are 280,000 total “other” cameras, of which 240,000 are mirrorless.
Unless the Leica M series is in there and not counted as mirrorless, that leaves 40,000 DSLRs. Almost all of those are Pentaxes, with the possible exception of a few thousand medium format models. Pentax is selling 40,000 cameras/year (possibly fewer if there are Leicas, Hasselblads or Phase Ones mixed in) across three current models and a few older ones that may still linger on shelves. That’s a tiny number (Nikon sells close to 1.5 million DSLRS annually, and Canon sells just over 3 million), and quite a bit smaller than I’d realized.
The entire Pentax line combined sells less than many individual camera models, including several that sell for over $3000. They are selling half as many cameras in total as Nikon sells $3200 D850s alone. Even though their R&D expenses are small because they’re buying sensors and reusing many other parts, can that really be a viable business? I’d previously thought they were selling more like 100,000 cameras/year, and a parts-bin business where the sensors are bought off the shelf and many other components are recycled from previous models might be viable on that scale. On less than half of that??? It’s worth selling through current stock, and it’s probably worth assembling cameras from parts on hand. It may be worth ordering parts to build current models – but it is it worth designing even mild revisions?
Could Pentax’s best play be to work with Sony to introduce a really good first-party mount adapter from Pentax lenses to Sony bodies? That would give new life to the millions of vintage Pentax lenses out there, and give their owners something other than market-lagging parts-bin bodies to use. It would give Sony users access to a huge number of vintage lenses without messing with third-party adapters. Of course they could do the same thing with Canon or Nikon, but both of those companies have their own vintage DSLR lenses, and are less likely to want to help Pentax with theirs.
The large majority of the remaining 240,000 cameras are Panasonic – some are Sigma, and some may be Leica. Olympus didn’t want their camera business at 330,000 units/year in a single mount. How viable can Panasonic’s be at 2/3 of that scale, spread out over two mounts? No matter what the division between Micro 4/3 and L mount is, neither one is doing well at that scale. If it’s overwhelmingly L mount, they have a high-end mirrorless business significantly smaller than Nikon’s, but without the large DSLR userbase to convert or the huge legacy lens line to support via adapters. If it’s mostly Micro 4/3 (as I suspect), L Mount is a curiosity with sales on a Pentax scale, but with much higher R&D costs since they’re trying to introduce new lenses rather than provide a few bodies to fans of their vintage lenses. Their Micro 4/3 cameras are selling at something like 2/3 of Olympus’ rate, which Olympus decided wasn’t viable.
Panasonic’s one ace in the hole is their video capabilities. They have an unusual market that buys GH5s and S1Hs, and, of the “regular” camera makers, only Sony competes there right now (with the A7S line). They are also competing against companies like Blackmagic that still camera buyers rarely think about. Canon is a concern, because the EOS-R5 would be a formidable competitor to the upper end of that business if the overheating issues are worked out. If the EOS-R5 indicates that Canon is prioritizing video in a way that they previously haven’t been, Panasonic could be in real trouble. If Canon starts selling a lot of true hybrids with video capabilities that equal or exceed those of the Panasonic cameras, and Sony has single purpose video cameras (largely the A7S line, but also some of their FS line) that share lenses and other accessories with their still cameras, that’s not good news for Panasonic. The other potential challenge is Fujifilm – the X-T4 is at least as capable a video camera as the GH5, and it’s a much better still camera. Will Fujifilm keep building video cameras that good, and will they use the X- H2 model designation to introduce an even better video camera? Will Panasonic introduce a GH6 that raises the video bar again, far above where competitors can go with better still cameras?
I’d be more optimistic if Panasonic’s still camera division were part of their highly respected pro video division. It’s actually tucked deep in their appliance division, perhaps an unlikely spot… There are some rumors of a restructuring, mostly in the Japanese press – could this involve merging their still and video businesses?
For the most part, these new Nikkei numbers don’t change much about where the industry’s going. We still have a “big three”, although Canon is larger than Sony or Nikon by more than I had previously seen. None of them are in imminent danger of losing their camera
businesses. The interpretations that are apocalyptic for Nikon don’t take all the data into account. Yes, Nikon sold about as many mirrorless cameras as Olympus last year, but they also almost certainly sold them for more than three times as much apiece, since nearly every one was a high-end full-frame model. In addition, Nikon sold close to 1.5 million DSLRs, including several hundred thousand valuable enthusiast and professional models, while Olympus didn’t sell any at all.
Nikon and Canon need to transition their large and loyal DSLR userbases to mirrorless, and Sony will be doing their best to transition them to Sony instead. This will be interesting to watch, but all of Canon, Sony and Nikon will survive COVID, barring something really unforeseen like no vaccine for multiple years. The Nikkei interpretation of Fujifilm is almost exactly in line with what I’ve previously seen – a generally profitable business about 1/3 the size of Nikon or Sony, selling in a loyal niche market. Again, probably doing well enough to survive, and their loyal userbase will serve them well. Fujifilm is smaller than Canon, Nikon or Sony, but they’re also less dependent on low-end sales, where smartphones are wreaking the most havoc.
Beyond the four major players, the Nikkei numbers give much more reason for concern. Pentax’s DSLR business is absolutely tiny – far smaller than I had realized. It is buried deeply in Ricoh’s annual report, so I had never been able to extract any sort of reliable figures about it. My best previous estimate was about twice what it actually turns out to be. At 40,000 cameras per year, my suspicion is that it sells fewer cameras per year than Leica, or than Fujifilm’s medium-format business. Is it possible that Pentax is selling fewer cameras than Hasselblad? Unlike those companies, it doesn’t have high unit prices to keep it solvent. 40,000 cameras at an average selling price of $700 looks very different from 40,000 cameras at an average selling price of $5000. Even as a parts-bin business that buys off the shelf sensors and other components, can they keep it going? Is it time to introduce a really good mount adapter for their great old lenses instead of cameras?
I was also surprised how few cameras Panasonic sold – 240,000 minus whatever portion of that was Sigma and possibly the small European manufacturers. That sets an upper bound on L-mount sales of 240,000 plus Leica if they’re not included, which is wildly optimistic since those 240,000 cameras also include Panasonic’s Micro 4/3 sales. There’s no way to break it down, but L-mount is unlikely to have sold 100,000 cameras in 2019, and that’s probably optimistic. Panasonic’s Micro 4/3 sales are probably under 200,000 – if they aren’t, L-mount sales are really tiny. Is either L-mount or Micro 4/3 viable going forward?
Smaller companies that use minority mounts may be the most innovative players. It would be very difficult, if not impossible, for a smaller manufacturer to license the E, Z, X or RF mounts. EF seems to be available, especially for video applications – Blackmagic uses it, among others. Oddly, Panasonic is among the users of Canon’s EF mount – their professional video gear doesn’t tend to use their own Micro 4/3 or L mounts. For mirrorless mounts, L-mount and Micro 4/3 are available, but none of the big players. The two most unusual camera designs I’ve seen in a while are the L-mount Sigma fp and the not yet released Micro 4/3 Alice camera. The Sigma is a modular camera where the minimum body weight is very low, but it’s missing significant pieces. Add them back in and the weight climbs, but it gives the photographer quite a bit of choice. The Alice uses a smartphone for control and processing – an idea that has been tried before, without much success (but somebody’s going to get it right). Sony had a couple of entries some years back, including the interchangeable lens QX1 . More recently, DxO has tried the DxO One, which is a fixed-lens camera attachment with a decent-sized sensor.
The overall thrust doesn’t change my analysis of the industry at all. All of the big four companies are going to be OK for a while, barring some major shift that hasn’t happened yet (COVID stretching for several years or some major player getting bought by a company that doesn’t want the camera business). A volume of 4 million film SLRs annually supported four or five camera makers in the 1990s (big players Canon and Nikon plus Pentax, Minolta and arguably Olympus). Sony now replaces Minolta, having bought their camera business, and has succeeded in achieving the first-tier status Minolta always sought. Fujifilm is in the role Pentax and Olympus used to play – a smaller, but beloved player who generally makes a profit on cameras. If Fujifilm had been bleeding money in the camera business for years, I’d be very concerned for them, but their digital camera business has been profitable from at least the X-T1 until COVID. A few quarters of losses due to COVID won’t cause them to abandon what they’ve worked so hard to build.
Anything past the big four is in serious danger – the biggest news in these numbers is just how small Panasonic’s, Pentax’s and Sigma’s camera businesses are. At a maximum, Panasonic sells half of the cameras Fujifilm sells in a year, and Pentax is a little over 10% of Fujifilm’s size. Panasonic is splitting that small business across two mounts, which can’t be helping. Sigma is a privately held company, whose growing and successful lens business is much larger than Pentax’s or Panasonic’s camera businesses, and Sigma is exclusively a photography company, while Pentax parent Ricoh is a copier company and Panasonic is a conglomerate. Between its private ownership and the lens business, Sigma has a bit more room, but may be unlikely to keep on with L-mount in the absence of Panasonic. Sigma could just go back to making great lenses.
With limited review resources, Luminous Landscape reviews will continue to concentrate on the “ big four”, plus unusual cameras in other mounts that show interesting trends in design. A Sigma fp is coming in this fall, and I’d love to see an Alice if it emerges. We have very modern, high resolution bodies available in Nikon and Sony mounts, and Fujifilm has always been very good about press loaners. Lenses from Sony, Nikon, Sigma and Zeiss are arriving and departing at a rapid pace for a very glass-filled Autumn. The first comparative lens piece of the fall will look at some of Sigma’s new DN lenses – mirrorless-specific designs with very good ratios of price to quality to weight. I have two here for testing, which have been extremely impressive, with two more coming soon. Canon is the major hole in our current test capabilities – we’re on the list for an R5 body and some glass, but I haven’t heard anything about when.
If there’s anything outside the “big four” that is of especially strong reader interest, I’d love to know what it is. I can try to get ahold of bodies and lenses if I know what’s of interest.