A better example of what Nikon is likely up to.
All the hubbub over the much proliferated, rumored Nikon Road Map made me realize that a lot of folk still haven't quite caught onto Nikon's methodology, let alone how specific part and lens decisions get made at the company. Nikon, if anything, has proven to be relatively predictable for several decades now, with only some of the fine engineering details being surprises. Short answer: Nikon likes to reuse parts across bodies, and Nikon has a fixed set of design teams that rotate to "the next design need."
With such predictability comes great risk, as Nikon is vulnerable to a lot of different risks: (a) disruptive designs or technology will likely not come from Nikon, so they'll be a follower, not a leader. We can see this already in the G1, G1h, and E-P1 designs. Those cameras have exposed a large weakness in Nikon's derive-new-cameras-from-old design philosophy. More on that in a bit. (b) competitors can basically predict what Nikon will do and use that to attack where they see vulnerabilities. The original 5D and the new 5DII and the Canon f/4 IS lenses are such examples, as Canon clearly saw that Nikon was not in a hurry to push FX, let alone push it down into the prosumer class. (c) customers are getting "upgraded out" as the proliferation of minor changes and additions to the same basic product is starting to wear them out economically. Coupled with the Yen appreciation, which caused price increases, and the recession, which caused demand decreases, this puts an increased tension on the 18-24 month schedule of "add some features, improve IQ slightly, but basically design the same camera over and over."
As I've written before, new generations of pro Nikon bodies, where new technologies tend to get introduced, come at four year intervals. Consider:
- 6/99: D1 intro, starts D1 generation (D1, D1h, D1x)
- 7/03: D2h intro, starts D2 generation (D2h, D2hs, D2x, D2xs)
- 8/07: D3 intro, starts D3 generation (D3, D3x, and so on)
- est. 7/11: D4 intro, starts D4 generation (D4, D4x, and so on)
That's a pretty safe prediction, actually: the D4 gets introduced late summer or early fall of 2011. Indeed, any Nikon "road map" almost certainly has that last entry. We'll get to what that means in a bit.
For the prosumer bodies, Nikon has hiccupped a bit:
- 2/02: D100
- 12/05: D200
- 8/07: D300
- 7/09: D300s
Even with those hiccups, it should be obvious that Nikon is iterating the prosumer DX body at about two year intervals (two years is their likely goal, but they haven't always hit it perfectly). The D200 appears to have taken longer to get to, but we're now back on some sort of reasonable two-year cycle (though with less changing between cycles). Thus, a D400, if there is one, is likely scheduled for 2011, alongside the D4 intro, if possible. Where the D700/D700x fits into this is still a little unclear, but because these two bodies are co-dependent (FX sensor changes and DX prosumer body changes), I'd suspect them to lag the above two schedules, which is exactly what has happened so far. Thus, a D700s is possible in 2010. A D700xs would be a year behind that, should it appear.
The "hobbyist camera" has been on a two-year cycle as well (the D70s as an interim very minor upgrade):
- 1/04: D70
- 4/05: D70s
- 9/06: D80
- 8/08: D90
Thus, one wouldn't expect a D7000 (my guess at the D90 replacement name) until mid-2010. Not predicting such a camera would mean that the predictor isn't paying any attention to Nikon's past practices: we've had a hobbyist camera since the N70, perhaps even further back (N6006).
The lower-end consumer bodies are a whole different story. We've had a new consumer DX body basically once a year at the low end, though there's been some jump back and forth between whether that's the lowest end model or a more mid-range model:
- 4/05: D50
- 12/06: D40
- 6/07: D40x
- 1/08: D60
- 3/09: D5000
- 7/09: D3000
The easy changes take less than a year (D40x from D40), the more difficult changes take a full year (D5000 from D90/D60).
Now, as to why these schedules are relatively predictable: Nikon uses the same engineering pattern and teams and rotates them between projects. When a team comes off one design, they get re-assigned into something else on the schedule. Nikon isn't really adding new design teams (though they have added people to the teams), thus there is a fixed capacity for designing new cameras, and it has a predictable pattern to it. Let's use that pattern to predict forward:
- fall 2009: D700x
- early 2010: D2000 or D4000
- early 2010: D3s
- mid 2010: D7000 (D90 replacement)
- late 2010: D3xs
- early 2011: D5500 or D6000
- mid 2011: D4 (D3 replacement)
- mid 2011: D400 (D300s replacement)
One underlying assumption in that list is that Nikon continues the same practices as before: technology introductions in the pro generation that wind their way down to the consumer bodies, heavy parts and sensor sharing to reduce costs, and minor enhancements being done with each camera and then passed to subsequent camera intros (easiest to see: the additions to the RETOUCH menu that seem to come with each new camera). Nothing indicates to me that Nikon has changed this basic pattern at all.
The only thing left to do is predict what each camera is likely to be:
- D2000--very low end 12mp DX camera with almost all automated features and only basic manual ones.
- D4000--split the D3000 and D5000 features (e.g. add video but not flip screen).
- D5500 or D6000--splits the D4000 and D7000 feature set, uses the D7000 sensor. Looks like a minor update to the D5000 to me.
- D7000--update the D90. It's probable that a new DX sensor is ready in this time frame, and thus I'd suspect a D7000 to initiate Nikon's use of it. Whether that's 14mp or 15mp (or even a new Nikon 12mp) is actually a rather uninteresting question, as none of those choices really up the ante in terms of image quality. More likely changes would be 1080i video support, a tilt-screen, and changes to Live View and video autofocus.
- D400--New technologies spawned by the D4 design group passed to DX (see D4, below).
- D3s--add the extended buffer, add sensor cleaning, no major changes otherwise.
- D3xs--add sensor cleaning, no major changes otherwise.
- D4--New technologies unveiled. The question is what those new technologies are. As good as the AF system is today, I think that it's highly likely that we'll get another iteration. It strikes me that even more integration with the viewfinder sensor is possible (e.g. if you've got a 50mm focal length lens on and focus on and detect a 50 pixel patch of skin tone and that skin tone moves off the AF sensors but becomes 100 pixels in size, you know that the focus needs to be moved forward, even though you can't exactly detect how much; sophisticated calculations could get you very close to correct, though, so I think we're going to get a focus ASIC added to the cameras). Of course, this is also likely a sensor transition time for Nikon, as well. Again, I don't expect large gains. I think everyone would be happy with a 14mp FX sensor that performs like the current 12mp one. 16mp is a stretch, I think, as I can't see us getting to that pixel count with the same high ISO performance, and I don't think Nikon can afford to retreat from the D3 level with this camera. There's very little chance that the D4 unveils a three-color sensor--I deem that to be a very long shot. We'd hear about Nikon taking over fab production somewhere before such a sensor appears, as to create Nikon's design it will have to be done in the die and fab. This isn't something that you could hire an outside fab to do for you.
Okay, that's the likely Nikon roadmap based upon past performance. I'll be honest: I don't like it. If Nikon executes that "safe" set of cameras, I think they'll find that their market share erodes and that growth in camera sales is zero or negative for them going forward. If I were in charge, I'd only be executing the mid-range the same as before (D6000 and D7000) while doing the expected iterations (D700x, D3s, D3xs). There's little to no risk in those decisions, and if I'm wrong about my risk taking, it still leaves Nikon with a base class of DSLRs that'll get them through.
Where the problems exist are in the low-end and high-end. Let's start with the low.
The Panasonic G1 and G1h, and the Olympus E-P1 have shown that the gap between compact cameras and DSLRs is real and can be exploited. Those products need to be shut down before they establish a firm grip on that gap. Panasonic, in particular, has a very strong LX3 and G1 that positions them into the serious-but-not-too-expensive realm. Indeed, those products established Panasonic as a real player, as they can now point to a solid compact line that extends very seamlessly up to the G1h. Panasonic isn't going to be challenging the D3 and D4 any time soon, but they have effectively made you wonder about how successful a D3000 will be. Likewise, the Olympus E-P1 also fits in this gap and has attracted a lot of attention (I predicted the initial production run would sell out, and it looks like I'm correct). New low-end, traditional DSLR designs are going to be met with yawns from customers. Thus, if I were in charge of Nikon's development, I'd have fast tracked something, anything, that fits into this gap segment. A P6100 and a D3000 is not going to hold serve. Indeed, the risk is that those two cameras will define Nikon as "not getting it." The only saving grace is that Canon and Sony don't get it, either. But seriously, Nikon, what do you think you will sell to the next generation at Best Buy when most of the local camera stores are gone?
The good news is that neither the G1 nor the E-P1 got everything right. These initial gap products both have liabilities that can be exploited. The Panasonic G1/G1h really messed up on the ergonomics, in my opinion, which is surprising, because they've been very photographer-centric and had good ergonomics in their previous designs. The E-P1 has focus performance issues and the messiest menu system of any serious camera. But by the time Nikon could put anything into this space, we'll have a G2 and an E-P2, so there's no room for mistakes.
Ultimately, I think Nikon has to put an APS compact and an mini-APS interchangeable lens camera into this space. The former to "solve" the Coolpix high-end weakness, the latter to protect the low-end DSLR realm. My personal opinion is that Nikon should target the high-end of this low-end. That means that you tackle the "poor man's Leica" with the interchangeable lens camera: strong user control, optical or EVF viewfinder, a set of small primes (28, 35, 50, 85mm), and clear image quality improvements over the m4/3 G1 and E-P1. Likewise, the APS compact should be targeted at the DSLR user who wants a compact camera to carry around all the time: strong user control, collapsing 28-70mm lens minimum, clear image quality improvements over every compact, still at a compact (shirt pocket) size.
Thing is, the way Nikon engineering is configured, I don't see how they get to either camera without adding another design group or sacrificing one of their already scheduled projects. Thus, if anyone at Nikon is working on anything below a D6000 right now, I say punt on that project and get working on one or both of my suggestions. (I should point out that Nikon does have a special design lab that is dedicated to out-of-the-box ideas and designs, and it's clear to me that this group has worked on a mini-APS interchangeable lens EVF camera design with a new lens mount. But they are not a production group. For their design to get into Nikon's camera lineup, someone at the top has to move that project out of the lab and into the mainstream engineering [message to Nikon management: you should]. This is a bit like concept cars in the auto industry. Put another way, at the lower levels Nikon has a lot of creative, forward-thinking engineers working on things that don't resemble current products. But Nikon's overall management is highly conservative, cautious, and formal in its decision making. It's been somewhat rare recently that any of these special projects make it into the wild as real products. The "multimedia glasses" [available in Japan only at the moment] are the only one I know of in the past couple of years.)
I wrote that I thought Nikon was vulnerable at the top end (D3, D4), too. The vulnerability here is more subtle. It's going to take something really special to get me to want to update from my D3 and D3x. Heck, they're built like bricks, so they'll still be going strong in 2011, I suspect. I'm not alone in this. The way we pros see it is like this: the only way we're really interested in upgrading is if...
- we get significantly more pixels with the same image quality (that means 16mp+ with D3 performance).
- we get some surprising and highly useful feature upgrade (most likely: AF improvements that are tangible).
If you haven't used a D3, I should point out that those are really high bars to get above. As I wrote on the front page of this site recently (now archived with the 2009 news), the D3 is a very nice camera and really hits a sweet spot in terms of performance and ability.
There is, however, something that would turn some heads. Imagine for example Thom's Ultimate D4:
- 14mp FX sensor with D3 performance.
- Modular body, allowing multiple battery/grip options, multiple viewfinder options, upgradeable sensor (this last doesn't have to be a field upgrade, though that, too, would be nice).
- Live View autofocus performance that is close to old phase detection performance.
- TrueRaw (allowing both the ability to turn off any Nikon raw optimizations, like black position, as well as to see true raw data analysis [histograms, etc.], as well as to combine adjacent RGBG data into a small RGB raw file).
That would be enough to get me to upgrade. Any metering, phase detection AF, firmware functions, or additional pixels, etc., that Nikon added to the mix would be icing on my cake. Basically, modularity is one of those "special somethings" that would get pros attention.
I should point out one possibility that has dropped off the radar at the high-end recently: the old MX rumor. The faux larger-than-FX ad campaign that first surfaced late last year certainly pointed to another possibility: that a D4 might use a larger sensor, something akin to the Leica S2 concept. You can build a camera about the size of the D3 with a much larger sensor, which answers the more-pixels-but-same-noise problem without breaking a sweat. A 36x48mm sensor based on the D3 sensor gives you a 24mp low-end MF camera in a DSLR-sized body. Or better yet, a 36x48mm sensor based on the D3x sensor gives you a 48mp mid-range MF camera in a DSLR-sized body. Why would this be interesting? Well, it takes a lot of pressure off sensor development (though yields and price will be an issue with very large sensors), yet gets you to a significantly higher level of pixel count with the same noise tendencies. The primary drawback is that it means you have to design a new set of lenses (though I'd be remiss to point out that if you take the PC part out of the PC-E lenses, you've got existing 24mm, 45mm, and 85mm designs that should work at that size with little change). The plus in this scenario is that it extends the D3s/D3xs lives well into the D4 era. If done with FX and DX optional crops and a mount adapter, an MX even makes a usable replacement for existing FX cameras with existing lenses.
Do I believe that Nikon would make the D4 a MX design? No. The problem of lenses means that you wouldn't get a lot of existing D3/D3x pros updating, because what they need optically just wouldn't be available. That includes most photo journalism, wildlife, macro, and landscape photographers. I can't imagine that Nikon would restrict a new pro generation primarily to studio photographers. Still, MX, like three-color sensors, is a low chance possibility that needs to be considered when trying to guess Nikon's future Road Map.
So that's the cameras, how about lenses?
Lenses are a lot more difficult to predict. This, too, has to do with Nikon engineering culture and organization. In the lens division, there's more leeway for the project leads to pursue things that interest them personally. That's why we get these mini series of lenses every now and then, like the PC-E ones. Tactically, they're not hugely important or big sellers. But someone wanted the design challenge and it did fix a parity issue with Canon's offerings, so it got on the schedule.
One problem with design whim in lenses is glass. Since Nikon produces all its own glass for the non-consumer lenses, pet projects steal glass capacity for short periods of time. Another thing that's happened is that demand has been under-estimated. Since the glass for a 600mm f/4 may take a year from start to finish to produce in the kilns and polish centers, when you underestimate demand for a lens like that up front, you run out of glass and have to dedicate the optical group to produce another batch. That can push other lenses out of the production line temporarily, creating more shortages, and enough of that push and shove can make it hard to get new lenses into production. Eek.
Fortunately, Nikon is nothing if not a self-aware and self-fixing company. While the underestimates are causing a near constant out-of-stock situation with many popular lenses, it does appear that Nikon has been reevaluating their lens production and making improvements that increase both their capacity and their responsiveness. At the consumer end (lenses produced in Thailand and China), Nikon now has things reasonably under control in terms of quantity, though consistency of quality has suffered a small bit. It's the higher end lenses coming out of Japan that are still an issue, though I see signs that things are improving there, too.
Still, as I've written before, Nikon has some design constraints that need to be considered before you can predict what is likely to come and how soon. Basically, the lens design process tends to take three years from start of design to first-ship-to-customer. And until recently, only six lenses were going into design each year, meaning that at any given time, about 18 lenses were in various stages of being readied. Nikon's managed to push that up to about 21 lenses in design the past couple of years, and I hear things from my sources that this has increased even a bit more. Let's say that it's now at 24 lenses being juggled in design at any given time. Since we've only had two lenses appear so far this year, we can expect five, maybe six, possible as many as seven more to be announced later this year. The first opportunity for that will be at the D3000 and D300s announcements in early August. The next opportunity will be at the D700x announcement later in the year. Next year, we'd see another six to eight lenses appear, as well. Eleven to fifteen lenses in the next 18 months would go a long way towards plugging some holes and fixing some weak choices if the designs chosen were wise ones.
First let's deal with the "when":
- August 2009: two to four lenses announced with D3000, D300s. Since these are both DX bodies, you'd expect such announcements to be lenses that might be useful on DX cameras. It wouldn't surprise me to see another DX mid-range zoom update, for instance. Also, we're long overdue for an AF-S version of the 80-400mm. Nor would it surprise me to see the final Micro-Nikkor refreshed, the 200mm f/4. (What we want to see though are a wide angle DX prime and a DX telephoto zoom [e.g. 35-135mm or 50-150mm]).
- Sept/Oct 2009: the remainder of the 2009 lens announcements should come with the D700x introduction. Here, I'd expect a few prime replacements and an update/replacement of the 24-120mm. (What we want to see are a 24-70mm and 70-200mm f/4 VR pair, a 70-200mm f/2.8 replacement, AF-S versions of several primes: a fast 24mm, a fast 35mm, the 85mm f/1.4 updated, and perhaps a 105mm, 135mm, or 180mm replacement.)
- First half 2010: two or three fill-in-a-blank lenses. These would be lenses that either update an elderly existing lens or fit into a random gap Nikon has perceived. If any of the lenses I've mentioned above haven't yet appeared, then this would be the likely time they would. Another likely candidate is a wide angle zoom of more modest specifications than the 14-24mm (either 17-35mm f/2.8 or 14-28mm f/4-5.6--basically the "filter friendly" wide zoom).
- Last half 2010 and first half 2011: more fill-in-a-blank and minor update candidates, I think. There isn't a big camera product launch in this time period on which to hang lenses, thus I think this period will see more fill-in-a-missing lens or fix-an-existing lens designs.
- At the D4/D400 launch in 2011: this is when I'd expect another barrage of targeted lenses that open up new lines for Nikon.
But as I noted, lenses are harder to predict with Nikon because there is more designer leeway in doing projects they're interested in rather than specific products. Because Nikon's lens sales have grown enormously in the past few years, I don't think that Nikon believes there is anything wrong with their current method of choosing lenses to produce; only their inability to produce them to demand is perceived as a weakness in lenses right now, I think. So I can't rule out our getting something completely unexpected in lenses. If a designer thinks that pancakes are the next great challenge they want to tackle, we might get pancakes. If a designer thinks that pushing lenses up another aperture is the next great challenge, then we'll get f/2 lenses instead of f/2.8. If a designer believes that they can make their mark by making something with a Canon specification that runs rings around it (e.g. 24-70mm f/4 VR), then that's what we'll get.
I should also mention that there's a sub-issue boiling for Nikon that could disrupt everything I've written above: video. The current Nikon AF-S designs aren't exactly video autofocus optimized. If Nikon truly perceives that video is not just a requirement but a major feature of DSLRs, we're going to get lens redesigns that are optimized for video (quiet focus/zoom, better real-time follow performance using contrast AF for video, etc.). That could show up as another mount addition (new contacts), meaning that much of the existing lens lineup once again is outdated for new cameras. Look closely at the next few cameras launched with video by Nikon for clues on this.
There you have it: the likely Nikon Road Map for the next two years. I could be wrong on a few details here and there (especially lenses), but an average of four cameras a year and seven lenses can't be far off what will really happen. I've seen no signs that Nikon is diverging from its established patterns, so I'm fairly comfortable that the broad picture I just laid down is close to right. At least until something dramatic happens that changes the camera world.