What Will Happen in 2003?

Digital photography is here to stay, and the pace of change is relentless...

[December 2003: I've made a few annotations and corrections to the article. They appear in square brackets and in red, just like this note. One of the announcements made in the conferences prior to PMA was that this year digital cameras will outsell film cameras worldwide. No surprise there.]

Okay, I'm a little late with my 2003 predictions. I'm almost certain this is going to be a watershed year for digital photography, so let me tell you what I think will happen:

• Market Consolidation Rears its Head. The big will get bigger, the small will merge and disappear. I really don't think we'll have 30+ brands duking it out for the consumer digital camera market for Christmas 2004. Let me let you in on a little secret: those 30+ brands aren't really 30+ digital camera designers/manufacturers. One of the reason we have such a proliferation of cameras at the consumer end is that you can go to one of a few major players (Sanyo, Kyocera), license most of the guts of a camera, tweak the industrial design and software a bit, and have those same companies manufacturer it for you under contract. The competitive price drops we've seen recently have made that strategy one that's tougher to make money at than building, say, an IBM compatible PC. Unless you have unique, distinguished products, you will fail in the Christmas 2003 market, and you will be an unprofitable also-ran in 2004. Moreover, you need some vertical integration to compete. Sony holds the largest market share in consumer digital cameras and they make the vast majority of CCDs. How do you compete with that? Something about your camera had better be distinguished, and having vertical integration with at least some of the parts (lens, DSP, memory, etc.) is also necessary in order to drive out third-party markups and keep your costs down. So who rocks on and who merely resembles an inert rock? Sony and Canon are sure to continue in the consumer winners circles, though Canon is going to have to show that they can do so with their own CCDs (or CMOS sensors) at some point. Nikon, Olympus, and Fuji live on mostly through sheer volume and imaginative designs. (Since I'm a Nikon expert and advocate, what does Nikon bring to the party? Lens design, lens design, lens design, plus chip integration ability.) Kodak, HP, and Minolta seem like clear losers long-term, unless they have something up their sleeve no one yet knows about. The Ricohs, Casios, and other marginal players will simply become even more marginal. [The onslaught of new digital camera introductions didn't stop at PMA. Dozens of new consumer cameras appeared, most undistinguished. And did I count right? Did market leader Sony introduce 10 new cameras, or were a few of those already announced? It really doesn't make any difference: the market leader has very strong products and pricing control. Exactly how do you overcome that? Most won't. Unfortunately, the market growth is so big still that everyone is producing at capacity, so market consolidation hasn't yet occurred. Still, a few marginal players have changed course or dropped cameras, such as the Contax full frame DSLR.] 0-1
• Foveon is successful, bankrupt, or bought. Successful is unlikely. To achieve that Foveon needs to have multiple, major licensees who will ship a unit volume of millions in 2003, double or triple that in 2004. PMA in early March will tell us if that has any chance of happening. [No significant Foveon announcements at PMA.] And if it doesn't happen, Foveon as it exists now is history. We're talking about trying to pay back tens of millions of dollars in venture capital by selling chips that will have to cost tens of dollars or less by the end of the year to be competitive. (As it is, at the moment I consider Foveon what VC's call "walking dead." Walking dead generate just enough cash to continue operations, but don't have an identifiable route to profitability AND investor pay back) The most likely scenario is that another chip player will purchase their assets to make their own technology more likely to succeed. Kodak seems the most likely possibility, but there are other players, some mostly hidden from consumer view (TI, for example). Sigma could be a long-shot purchaser to preserve the SD-9. Which brings us to the next most likely scenario: Foveon doesn't hit critical mass, can't find a seller, and thus is on the death watch at the end of the year. But by the end of 2003, one way or the other, we'll be able to make a near certain prediction about Foveon's future. [It's looking less promising for Foveon. It's clear that the consumer chip they announced last year hasn't been picked up by anyone yet. That's not good news because the volume is in the consumer cameras (on the order of 20:1 or more). That leaves the digital SLR market, and that doesn't look good either: Canon does their own CMOS; Nikon and Sony have a close relationship due to Nikon's seminconductor equipment division and now Nikon's got their own chip designs, too; Kodak and Fuji have their own capabilities (though the Kodak DCS Pro 14n doesn't use a Kodak chip for some strange reason); Pentax's new *ist (what kind of name is that?) uses a Sony chip; of the possible significant players that leaves only Minolta as possible licensee for DSLRs, and if Minolta finally rolls around to that market, they're coming too late with too little. In short, Foveon doesn't have enough traction to hold on with only Sigma as a licensee. It's just a matter of time now.] 0-2
• The first 20+ megapixel announcement occurs by the end of the year. It may simply be a chip announcement for future delivery, but it will happen this year. Deliveries of bodies with 20mp won't occur until Spring 2004 at the earliest, though. (For those that are curious, chip resolution has been doubling on about a 16-18 month cycle.) [Well, this one came true with Fujifilm's medium format digital back, at 21mp. Delivery will be a bit later than expected, and we haven't seen an SLR announcement with this much resolution yet, though.] 1-2
• Canon finally begins to out-momentum Nikon. The Nikon D1 absolutely caught Canon flat-footed. It was earlier than Canon expected, and it was far better than Canon expected. The early success of the D1 also generated another side effect: Nikon got to full-scale manufacturing with their digital SLR first. With the quick follow-on of the D1h and D1x, Nikon stole Canon's thunder a second time--the Canon 1D was too late and missed the big switch of photojournalists to digital. Nikon continued to keep the strong position in 2002 when they essentially matched Canon's D60 in performance and price with the D100, again catching Canon off guard. But with the 1Ds and upcoming [10D], Canon is finally beginning to nudge perceptually ahead on the technology side. For pros and serious amateurs, Nikon owned digital SLRs in 2001. Canon got into the game and basically matched Nikon in 2002. And on a feature and performance basis, Canon might move a bit ahead of Nikon in 2003. But I'm assuming that Canon has worked on the kinks on the manufacturing side (or else they'll lose the momentum they're gaining) and that they can serious drop their prices. [Certainly Canon has gained momentum, though it is unclear if they've managed to dislodge Nikon from the #1 DSLR position. With the D2h announcement, Nikon for the first time talked about actual DSLR shipments. In 2002, Nikon claimed 60% of the DSLR market. From there, it gets murkier. Canon is claiming to build and ship 30k 10Ds a month, and 2k 1Ds's (no word on the 1D, and the Digital Rebel number Canon gave of 70k a month is clearly not being met yet). So, spread out over 2003 that would be 350k units plus however many Digital Rebels managed to get shipped by year-end (100k?). But Nikon is estimating they'll ship 600k+ DSLRs in 2003. At those numbers, my prediction would not come true in DSLR unit sales. Still, I think it clear from market buzz that the Canon 10D definitely put Nikon on notice, and the Digital Rebel now is the DSLR body volume leader. In the compact digital camera market (Coolpix, G5), Canon clearly has the momentum, with ~8m units to Nikon's ~5m.] 2-2
• Nikon will introduce a US$1200 digital SLR (I'll call it the D65 for lack of a better name). I have no inside information on this, I'm merely basing this guess on what I'd do if I were Nikon: you take the current generation of CCD and reduce all the other costs around it to produce a lower cost product. That means an N65 body, even more LSI integration, and the existing D100 CCD. Can you get US$800 of retail price out of that? I think so. We've got maybe US$100 in body cost difference, I'll assume we can get another US$100 out of electronics costs (remember, these are retail numbers--the actual amount of cost reduction in manufacturing is about 1/3 this amount), which means we have to get Sony to drop the CCD costs enough so that we can pick up the other US$600 (implies a US$200 cost reduction to Nikon). Well, we'd be easily quadrupling the volume, so that should be well within the realm of possibilities. If we're careful with the other pieces (use existing battery and charger technologies, leave off the neck strap and video cable, etc.), the US$1200 price point is obtainable right now. [With Canon's introduction of the 10D at US$1500, that pushed Nikon into reducing the price of the D100 to match, so we're only $300 off from that target with the D100 body! Unfortunately, Canon beat Nikon to the punch, introduced the Digital Rebel for $899 before Nikon could introduce the D75. Since Nikon is production constrained, and it is so close to Christmas, it looks like the D75 announcement will wait until early 2004 (Japanese business press is saying "first half of 2004.") And it'll be cheaper than I predicted.] 2-3
• Nikon will introduce replacements for the D1h and D1x. I'd predict that this will be one body, but who knows, they may surprise us again and keep separate development going. After all, 40 frame buffers at 5 fps of 11-14mp images is borderline affordable at this point in time--to keep the cost down and pick at Canon's weak point on the 1Ds, Nikon might have a lower resolution D2h and a higher resolution but slower/smaller buffer D2x in mind. I made a prediction in my D1 Report back before PMA that these bodies would NOT be full frame, and I think that's still the most likely scenario. I'm not convinced that full-frame CCDs have high enough yields at low enough cost to be in high volume products yet. [My birthday in July brought the first of the replacements, the D2h. It's unclear yet whether we get one or two replacements this generation. The big surprise was the step backward in resolution and the step forward in Nikon's own sensor design.] 3-3
• The long-rumored Nikon F6 becomes at least a partial reality. Nikon's relied upon the same autofocus and metering technologies for quite some time now. The F6 project was to develop and showcase the next generation of Nikon camera technologies. So one of two things will happen soon: either the F6 will actually show up as a surprise (and once again catch Canon off guard, this time in the film realm!), or the technologies will pop up in a new digital body, most likely the D1h/D1x replacements. [The F6's autofocus system is in the D2h, so it looks like we get the partial realization for sure. Now, will Nikon surprise us with a full F6?] 4-3
• The D2 disappoints some, excites others. Nikon will likely announce a new high-end digital body replacement for the D1's at PMA (see above). I'd have to guess that it will have the 1.5x field of view change, but significantly more photosites (best guess: 12mp, achieved with the D100 chip ala the D1x modification). IF Nikon fixes the few D1 flaws (battery, primarily), keeps the cost under the US$5000 mark, AND delivers new lenses (all of these are expected), plenty of Nikon aficionados will proclaim it "enough" and keep Nikon competitive in the digital SLR game. [It appears that I nailed that one. Nikon delivered what we expected, except for resolution, across the board and we got online forums filled with the disappointed/excited syndrome.] But the full-frame Mafia will pooh-pooh the offering and insist you should spend more for a Canon, Contax, or the same for a less-well specified Kodak. Personally, I'll be looking at the image quality. Give me double the resolution and lenses that deliver edge-to-edge performance, and it doesn't matter to me what the sensor size is. 5-3
• The action moves to software. It really isn't that hard to demosaic raw data. Heck, even Adobe, who continues to not understand digital photography very well, has managed to do it. To bad they didn't enable more 16-bit functions in Photoshop to take advantage of that RAW converter, eh? It isn't that difficult to build an image browser. Image manipulation libraries and source code are easy to come by. Yet not a single product (or product suite, which seems to be where Adobe is headed) I know of does all these things well (many don't even do one thing well). I'd swear that Adobe doesn't have any digital photographers working with them on Photoshop. That Canto can't see beyond the word database. That every software vendor first and foremost thinks of digital camera manufacturers as a source of OEM licensing revenue and doesn't realize that the users are hungry for well-designed, high performance software. The list goes on. Meanwhile, both Canon and Nikon have limped along with Macintosh software development, barely getting to the realm of usability. The camera manufacturers simply don't get software. But there's been a couple of teams of well-known imaging experts working on the problems inherent in digital imaging for awhile now (if I wasn't so swamped with my own nascent business, I'd probably be leading a third group). In 2003 I expect to see the first product that turns heads. And it will have the entire industry going "why didn't I think of that?" Save your lunch money, folks, the good stuff doesn't come cheap...[Getting closer, but still no cigar. Kodak's Photo Desk "looks" are interesting, Capture's linear distortion fix for the 10.5mm fisheye is interesting, but no big software revelations yet. Photoshop CS raised the ante on workflow, though it still doesn't quite match what photographers need or want.] 5-3?
• Disposal digital cameras get announced, if not shipped. Well, "disposable" is probably the wrong description. Even the film cameras [usually] call themselves "single use" and avoid the connotations of the word disposable. The point is this: it is entirely possible to build a box that houses a digital sensor of modest resolution (certainly 640x480, but possibly higher), some memory, a "just good enough" battery and the usual plastic lens/viewfinder combo at a cost that would allow it to be priced at, say $29.95, including prints, $39.95 including CD and prints. The vendor would be reconditioning these boxes and reselling them, as, unlike film, there isn't anything particularly disposable in them (battery, perhaps). Kodak and Fuji will hate this idea and avoid it for as long as possible (to their detriment), as it doesn't drive their film and paper businesses. [Kodak introduced a "disposable digital" product at PMA: a film-based single-use camera that gets you a CD with your prints when you turn it into a developer. I told you they wouldn't like wouldn't like the all-digital idea, as it doesn't drive film sales.] But a big-box vendor, such as Wal-mart, could have a field day with such a product. [Well, we'll see about that last statement, because Ritz Camera has announced the $11 Dakota digital disposable (as has Walgreens). Yep, $11 and fully digital (2mp). While Ritz and Walgreens might not quite be a "big-box," they're mainstream enough through the US to see if I'm right.] 6-3
• It will become increasingly clear that Kodak's digital future is poor, thus their entire future outlook is poor. The signs are everywhere, actually. The overhyped 14n doesn't even use a Kodak-developed sensor nor a Kodak-developed body. If you're going to become a parts consolidator rather than an innovator (ala Dell in PCs), you'd better have some trick up your sleeve to lead the way. Kodak appears trickless. I'll temper that one small bit: Kodak's Photo Desk software is better than any other program coming with a DSLR these days, and facilitates a useful, clean workflow. The big splash Kodak made with the 14n at Photokina has now been wasted with non-delivery and by the sudden quietness of the product marketing arm. [At PMA Kodak announced that the camera has "shipped." Hmm. I think this was marketing talk for demo units were shipped to dealers. Still, it is nice to see the Kodak marketeers talking again. Moreover, their engineers really do seem to be making the Pro 14n perform better with each new firmware iteration. Is that enough? Probably not, but see my review for more thoughts on their marketing of this product.] Meanwhile, we still don't have Macintosh support for their latest dye sub printer, which shows that Kodak doesn't (fully) understand the installed base of serious digital camera users. Everywhere you look at Kodak's offerings and strategies, you see gaping holes, misguided decisionmaking, and lack of execution [That "disposable digital," for example. Kodak's version can't be developed at most one-hour photo locations; you get prints and a CD in two days. Compare that to Walgreen's disposable digital, which gives prints and CD back in 15 minutes at those locations.]. Kodak will not survive the digital transition, becoming just another Northeastern manufacturer that didn't understand the business it was in, didn't transition with its users, and ultimately falls into negative sales growth and unprofitability. No, the end game won't happen in 2003, but it will become clearer this year that the end game is approaching. [And as the end of the year comes to a close, Carl Icahn has moved in and purchased 7% of Kodak's stock. The buzz is that he believes that value of selling off Kodak's many businesses is worth more than the current stock price, and he's probably right. But splitting Kodak up is an almost sure way to put it out of business long-term, as it simply won't have the capitalization to compete. Meanwhile, Kodak's president has outlined a "digital future" for Kodak, cut the dividend to help pay for it (thus raising the ire of many of the shareholders and triggering Icahn's investment), and moved the digital team to Japan. Huh? This makes Rochester look even more vulnerable. Why would a digital photography company that puts its main R&D in Japan have headquarters in Rochester? I see most of the Kodak moves as being too little, too late. Their film sales are eroding rapidly, and their digital growth isn't fast enough to make up the difference. Compare Kodak's position to Nikon, whose Imaging division is already 80% digital in unit volume and even higher in sales volume.] 7-3?

Well, there you have it. I rate my accuracy ~70% (7-3 with one no-call). Yes, I gave myself full credit for the Kodak decline and D2h/D2x announcement when it could easily be argued that I should only get half for each (and thus be 6-4). But I also didn't give myself credit for Foveon's failure to sign up any new licensees (and thus well on the road to the bankrupt or bought scenario). And besides, you've already formed an opinion on how well I did and nothing I write is going to change that, right? The important thing is how am I going to use that 70% value? Well, it's too good--it means I didn't stick my neck out far enough. So for my 2004 predictions, I'm going to add several stick-my-neck out guesses to the easier ones. I'd like to push the margin and only be 50% in my public predictions in the future. We'll see if that prediction is right ;~).