|Nikon D4 Review
12 Becomes 16
|Add a comment or send Thom feedback on this article.|
It's hard to believe it has been a year since the D4 first started getting into users' hands (including me). A very quiet year, as it seems that virtually none of the usual Web sources has taken the time to review this top-of-the-line camera yet (including me). Is that good news or bad?
In my defense, I generally don't review a camera until I've had plenty of shooting experience with it. As it turns out, 2012 wasn't exactly a year in which I did much shooting that required the D4. Thus, it's taken more time than usual to get enough experience with it to say anything definitive. A number of folk apparently believed my silence (and that of others), has been because there's something wrong with the camera. No, not really. It's a complex product that takes a lot of work to get through even most of its features, let alone assess all its performance aspects.
Let me just say up front: a D4 in practice is an awful lot like a D3s with some modest improvements, but 4mp more data. That statement alone has some implications in it, as in "it might not be worth it for a D3s user to upgrade." We'll get to that. Let's first talk about the camera's features and changes.
So let's start with that battery change. Nikon claims that new Japanese market laws made it impossible to keep the old EN-EL4a in production. Something about maximum capacity of small Lion batteries now allowed; the EN-EL4a is over the limit for its size, apparently. Nikon also makes claims about the D4 being more efficient in power usage than the D3s, so everything should basically be a wash, maybe even an advantage in certain types of shooting. Okay, I'll buy all that. What I don't buy, and what pros everywhere didn't like, is that Nikon engineered a new battery shape and door that looks an awful lot like the old one, but is just enough different than the older battery so that there's no interchangeability between D3 and D4 bodies. Zero. As in, throw out all your old EN-EL4a batteries as they do you no good once you've switched to the D4 series.
Frankly, that was a dead wrong decision on Nikon's part. They could have engineered a solution (battery compartment door and new battery) that allowed use of older batteries, if you had them. Instead, here in the transition period where some of us have a D4 and a D3x, we're now in a two battery set situation. This is so customer unfriendly to Nikon's best customers, it's difficult to explain. Certainly Nikon hasn't explained it well enough. The net impact, though, is those of us who were D3s/D3x users that now are D4/D800E users, have to jettison perfectly fine batteries and keep two sets of new ones, one of which is worse than the previous one in performance. That's a big ugly start towards trying to convince pros to switch. There had better be compelling points elsewhere that offset it.
Let's look at some other aspects that are new. For example, the backlit buttons. Yes, they are backlit, but only if you hit the LCD illumination function on the Power switch or you turn LCD illumination On. Not all of the buttons are backlit, either. The playback, Delete, and AF-On buttons aren't, for example. As Nikon has configured it, you're faced with either an all-or-nothing choice (LCD illumination On) or an extra-step process (Power switch illumination). Certainly the latter can be learned, and you can argue that having backlit buttons that you can enable is better than not having them. Yet as implemented, it feels incomplete and not fully thought out (there's nothing that any of the backlit buttons relate to that is in the top LCD, for example, so having the buttons always linked to the top LCD illumination seems like a power waste in most situations). That's one of the problems of adding features to a pro camera: we pros want lots of flexibility and control; this addition gives us little, though it does give us a potentially useful thing.
I'll continue my discussion of new things in a more disciplined manner. Let's step around the camera side by side while examining the changes.
From the front, you'll notice some "curve" and slight size differences (which I'll talk about more in Handling), but probably not any quickly identifiable changes. There are actually two that come into play while shooting: the DOF Preview and Fn buttons are ever so slightly closer together on a D4, and the D4 uses the D7000-style Autofocus mode switch and button instead of the old C/S/M switch on the front with other controls on the back of the camera.
Up top, we have a new red Record Video button between the Mode and Exposure Compensation buttons that have been moved slightly. There's no longer a locking Metering Method selector on the side of the prism, but instead the L button (Lock) on the top button cluster has been changed to a Metering Method button. What happened to Lock? It's an assignable function to the programmable buttons (button plus front dial locks/unlocks aperture, button plus rear dial locks/unlocks shutter). There's also a dedicated lock function (CSM #F8).
On the back things change a fair amount, as we now have a six-button stack to the left of the color LCD instead of a five. That stack now includes the Zoom In/Zoom Out button arrangement ala recent consumer Nikon DSLRs: no more button+dial to zoom. Live View gets the Still/Video switch, which made it big enough to displace the Microphone button down to under the lower LCD. Besides the slightly larger main LCD, the most obvious differences are the loss of the Focus Area selector (now done with the Focus Mode button on the front of the camera), and the addition of two miniature thumb joysticks. I'll have more to say about the latter in Handling.
On the left side of the camera (viewed while holding it) we have four discrete doors to the connectors instead of two. For certain uses, this minimizes door flaps sticking out and potentially interfering with your left hand position, for other uses you'll have more flaps open to contend with. Not sure there's a gain here. But we did gain dedicated WT-5 and Ethernet connectors for communications, while we lost the DC Power In socket (now you use a dedicated replacement for the battery when connected to DC). Plus, of course, we have a new battery compartment door (which doesn't have a name etched on it, which makes it difficult to figure out which door you have in your hand when you've got both D3 and D4 doors sitting around; do Nikon engineers actually use the cameras they design?).
All told, the observable external changes are again minimal between pro generations, though not as minimal as the D2x to D3 changeover.
Inside the camera we find the more substantive changes. You all want to know about the sensor first, I suppose, so let's deal with it. Instead of a Nikon 12mp sensor (D3s) or Sony 24mp sensor (D3x), we now have a Nikon 16mp sensor. And yes, it's a Nikon only part (NC81366W), with Nikon ownership etchings in the silicon. This is coupled with an EXPEED3 ASIC.
What you get from the combo is 16mp images (4828x3280 max) that can be rattled off at 10 fps for what seems like forever (13 seconds with JPEG Fine, even 10 seconds saving 14-bit raw files to an XQD card; more on that in a bit). That's one heck of a lot of data being captured and moved around with aplomb. By comparison, the D3s hit the limit at 9 seconds with JPEG Fine and 4 seconds with 14-bit raw files, which seemed perfectly acceptable at the time. Note that a Nikon V2 (14mp) will rattle off 60 raw images in a second using the EXPEED3, which seems to indicate that the ASIC has some bandwidth potential that's not being used in the D4, believe it or not.
The new sensor and EXPEED3 combo has some other tricks up its sleeve, too. Video has been completely reworked from the D3s's limited 720P. Nikon now gives us 1080P/24/25/30 and 720P/24/25/30/50/60 with very good H.264 compression (this current compression engine in the EXPEED3 is far better than previous generations). If that weren't enough, the D4 can output clean, uncompressed 8-bit video on the HDMI port for display or recording by an external device. With something like the Atomos Ninja or other external recorder, you can get broadcast quality video out of the D4. But wait, there's more! If you order now you can also shoot 2mp stills while recording video, plus you can get a very useful and clean 1:1 sensor output mode that doesn't have any video subsampling artifacts (2.7x crop).
If it's starting to sound a bit like a lot of the substantive changes are video related, they are. It's clear that Nikon spent some time thinking through and engineering pro-level video options on this camera, right down to the fact that you can output Broadcast range-limited video on the HDMI port (correctly set Black and White levels) and change the aperture continuously rather than in stops. Some of us probably wish that the same level of effort had been applied on the stills side, but some of you are probably happy to have a much more versatile pro body that can output both quality stills and video as needed.
We have many more internal features to hit on. One that engenders as much controversy as the battery change is the dual card slot. On the D3 series, we had dual CompactFlash slots. With the D4 we now have a mixed slot situation: one XQD slot and one CompactFlash slot. This is one of those good news/bad news situations. The good news is that CompactFlash is about topped out in speed, but XQD can take us much, much faster. Even on the D4 there is a difference: you can buffer about 16 more 14-bit images on the XQD slot than the CompactFlash one because the cards are just clearing data that much faster. If you're shooting at 10 fps, you'd notice the difference. If you're just shooting more casually, you're probably not going to notice much difference from the D3s other than the buffer seems bigger the few times you do press it.
It's the mix of slots and the non-acceptance of XQD by others that is grating on most Nikon pro minds. We'd all prefer to have two matched slots. For a lot of reasons. First, when you have both slots active, the slower card is what determines the performance of the camera, and that will always be the CompactFlash slot. If you don't have UDMA-7 type CompactFlash cards (e.g. are using non-state-of-the-art ones), you may very well notice right away that something's amiss. When you shoot with Secondary Slot Function set to Backup, performance is determined by the CompactFlash card. Yes, I know this means you can still use some older cards in the D4, but you will suffer a performance hit in doing so. Most of us pros I think would rather have matching slots, as it means only carrying one type of card with us. Consider if you're shooting with a D3s or D3x and a D4: different batteries, different cards, different DC dongles, different WiFi adapters. That's just enough to make it a logistical nightmare for someone carrying along both bodies and needing all the options. Yes, progress in technology does lead to obsolescence, but Nikon picked some strange form of overlap-but-not-quite-overlap and made things a little more difficult than it had to be, I think.
Then there's the stunning bit: either Nikon was way ahead of the pack in adopting XQD, or they took a wrong turn. Here we are over a year later, and we've not had one other significant product embrace XQD. Not one. Co-developer Sony touted these cards as being great for video use, but the only products Sony makes that can use the cards are high-end XDCAM models, and then only via an adapter; essentially they're a secondary option on these pro video cameras (as are SD cards and Memory sticks). So at present we have two card makers (Sony and Lexar), one still camera that uses them (D4), and a small handful of high-end video cameras that can use them via adapters. Doesn't seem like the world is moving to XQD, does it?
Back to the changes.
The focus system, as noted, uses the front button/switch with feedback in the viewfinder to control all major focus settings, ala what was first introduced with the D7000 and now has appeared on all other models at that level or higher. Personally, I'm all for it. No taking the eye from the viewfinder to change modes. It does take some getting used to, though. One focus setting we need to have the same level of control for is Focus Tracking Lock On. Unfortunately it's still buried in the menu system. Why does it need to be at the shooting level? Because the subject you're shooting and the backgrounds against which you're shooting make you want to change it sometimes. Take BIF (birds in flight). Blue sky as background allows a different Lock On timing than bird flying low with trees in background. Same thing happens in sports, events, and other situations with complex movements present.
Another change in the focus system is that f/8 lenses (and lenses with adapters that are effectively f/8) now support autofocus. There are some penalties for this: from f/5.6 to f/8 not all sensors are active, and at f/8 only a small subset near the center are active. Still, if you're at 1000mm f/8 (500mm f/4 with TC-20EIII) you're probably not trying to focus at the edges of focus area, and using the central area is probably within the depth of field of your subject across the frame. You do have to learn what sensors are active, but I don't see this as a large penalty.
The focus system is also even more tightly tied to the metering sensor in the viewfinder, and that sensor itself has been upgraded to 91,000 pixels (from 2016). This, coupled with the D4's fast CPU makes for an autofocus system that sometimes seems to have a very smart mind of its own. Again, like the D3 models, the big benefit comes with some of the automated focus choices when following humans in action (skin tone detection, apparently). Focus also works at 10 fps full frame, a slight increase from the D3 models.
A few new menu options (some seen on other models between the D3s and D4) appear, too: Timelapse, Exposure compensation for flash (separate flash exposure and exposure compensation), and a new Store points by orientation function that has the camera move your selected focus point when you move the camera between horizontal and vertical orientation. In other words, if you had selected a left-hand point when shooting horizontally, moving the camera vertically will still have a left-hand point selected. One surprising bit: the new additional GameBoy-like thumb pads can be programmed separately from the main Direction pad, and pressing these small controls into the camera is a programmable button action! That gives us five customizable buttons (six if you count the button in the center of the Direction pad). If you can't customize a D4 to your usage, I'm not sure which camera you can.
One last new bit: an Ethernet connection on the camera. This is a bit more than it at first looks. Believe it or not, there is a Linux-based computer sitting at the base of the Ethernet system, which means that there's potential for much more than just moving images off the camera. The D4 moves images off the camera very effectively via Ethernet, too. On the order of perhaps a full 14-bit lossless NEF every three to four seconds. It's a little difficult to measure exactly, but it's very much faster than using a direct USB connection, and even a little faster than the WT-5 WiFi option (maybe 5.5MB/s for Ethernet versus about 5MB/s for WiFi).
Even more interesting is that the D4 has an HTTP server mode, so you can fire up the Web browser on your computer and connect to the Live View and perform camera control of the D4. That's right, you can make some camera control changes via Ethernet, and get Live View with only a slight lag. Nikon is also fond of showing a connection to an iPad, but here things slow down a bit, as you are communicating on the iPad's slower WiFi to a computer, and then via the computer's Ethernet connection to the camera. However, it's still a usable tool for some. I'm sure by now a few enterprising studios have mashed up some interesting setups via all the communication capabilities that are possible. If you're interested in the Ethernet abilities, be sure to read Nikon's Network Guide.
But wait, there's more. If you have more than one D4, you can set one to act as a master release to up to 10 others (synchronized release). You need a WT-5 on each camera, but this is an interesting solution for the pro sports or event photographer who has set up remotes at a venue and wants to control them from the main camera they're carrying. Of course, most of us long ago started using PocketWizards for this same function, and they're cheaper than buying a bunch of WT-5's and easier to configure.
A pretty strong update of the D3s, though many of the changes might not appeal to all shooters (all that video capability, for example).
Review product source:
Support this Site
Since we've now had three "big" FX performance bodies, it probably is useful to look at a table of the primary differences:
First up, we have to address one major change that most people don't notice on first visual inspection: Nikon has once again worked with Italian designer Giugiaro to tweak the camera's basic body shapes. This relationship started with the F3 many decades ago, and has persisted through most of the pro generation changes.
If you're coming from a D3 model you'll immediately feel the changes in your hand. In particular, the shutter release area (on both the horizontal and vertical grip positions) has more angle to it, and the hand grip itself is a bit deeper. I don't have large hands, so now the right-hand grip feels a bit too big to me, but it's likely to please some, and it does work a bit better with gloves. But the slope of the shutter release had a lot of folk complaining early on (the D800 shares this). That part, I believe, is clearly for the better long run, as it puts the index finger in a more relaxed position overall.
More important, however, is that Giugiaro finally convinced Nikon to think of the horizontal and vertical holding positions as "the same." That's not perfectly accomplished on the D4--I'm not sure you can really do it perfectly short of making the body a bit ungainly--but it's very close. Here's what got improved: the AF-On button is in the same relative position to the hand position on either grip, and so are the two new sub-controller pads (thumb pads). No more reaching for a button in an awkward place or stretching to move the focus position when shooting with the vertical grip. Of course, we don't have an echo of the Depth of Field Preview and Fn buttons on the vertical grip, which would have completely rounded the shooting situation to not favor either position, so Giugiaro still has a bit of convincing to do with the Nikon designers.
One other thing about those thumb pads: they're not the level of quality of the rest of the body. Every other control and button is pretty hardy. But I'm worried about these two little pads and the main Direction pad. First, they stick out enough from the body that they tend to catch on bags and other things as you're handling the camera. Second, they don't have the quality feel of the rest of the controls, especially the thumb pads. I've already heard from pros who've had the thumb pads break off on their cameras. My guess is that this will become one of the more frequent D4 repairs from rough use.
Plenty of other subtle body tweaks are present, too, most of which haven't really been written about or discussed by photographers, but which most certainly make for subtle improvements in handling. For example, the buttons on the left edge of the color LCD all have a subtle indent from the left edge to the button, making them easier to find by feel. The ISO, QUAL, and WB buttons are now labeled on the button rather than above the button. Unfortunately, Nikon still hasn't figured out that we need a raised dot on certain buttons to find them by feel. This is especially true of the ISO/QUAL/WB buttons: a raised dot on the middle QUAL button would allow us to find all three by feel (find the dot, go left for ISO, right for WB, or press for QUAL).
Flash users will find good news (this change is also on the D600 and D7100): Exposure comp. for flash is a new Custom Setting that allows you to remove (ambient) exposure compensation from flash exposure calculations in i-TTL modes. Let's see, I first mentioned the problem of interlocked exposure and flash exposure compensation back in my Nikon Flash Guide, published in 2001 (see "Warning" on page 315, for example). Here we are over a decade later and Nikon has finally addressed this, though it's buried in a Custom Setting and the default is the old way that's problematic.
There's other work to do, though. Two things still bother me about the current implementation: more "which button" problems and the "button with glove" problem. They are related. Consider the top button cluster (Flash, BKT, Metering). Nikon has put a very subtle "outer ridge" on these buttons. Too subtle, I think. By feel it isn't always easy to find the right button while keeping the eye at the viewfinder. Moreover, with thick gloves on you can't always hit them accurately (not enough extension above the main surface). You see this on the vertical grip AF-On button, too: the horizontal grip version of the button is raised above the surface and unmistakable to find with your thumb. The vertical grip version doesn't have as much ridge; with even light gloves it's more difficult to find.
We're into the ninth generation of Nikon pro cameras, and we're now into a world of lots of potential connections, too. While I commented about the multiple flaps hiding the connectors on the D4, I'm a little disappointed that the design is still: plugs sticking directly into the camera with no support. Shooting video I have three plugs coming out of the left side of the camera and maybe one coming out the left front (remote control). That's a lot of wire left unsupported and making for no good hand position on the left side of the camera. It's time for some camera maker to realize that we need a solution that works both for the hand and to keep connectors from getting stressed. The simplest solution I have to that is to use right-angle connectors into the camera and a cable channel that has a lock bar at the bottom of the camera. Unfortunately, using anything plugged into connectors on the D4 just gives you cables that are easily dislodged, snagged, and have stress on them, and which get in the way of the hand.
While we're on that subject, Nikon did address the WiFi problem as it applies to cords: the new compact WT-5 plugs into a new special connector on the left side of the camera. Coupled with its small size and position high up in the connector set, it's reasonably out of the way and less likely to be snagged accidentally. Basically you just add a small hump to the side of the camera, one that has locking connector. Score that one right for a change.
If you're still using a D3 model and the D4 in a single shooting session, as some pros do, cognitive dissonance raises its ugly head a bit, as usual with all the cheese Nikon keeps moving. While much on the cameras are alike in handling, there's enough different that isn't that sometimes I have to think to remember which camera is in my hands. If you're in this situation, here are the big differences that'll cause you to want to know which camera you have in your hand: autofocus controls, metering method, AE-L/AF-L (the D4 doesn't have this button), voice annotation, and zooming on image review. Everything else should feel pretty natural in terms of transition.
Other subtle handling issues you might not have noticed: continuous shooting now extends to 200 images, not the 130 of the D3 models. In Live View, the aperture is active and the exposure shown on the color LCD is settable to be accurate to the ambient light settings or not in Manual exposure mode (OK button switches).
Finally, I like XQD cards. Pros are in a hurry (usually). Removing and inserting an XQD card is very reliable and easy. No chance of bending pins when you're in a hurry, as sometimes happen with CompactFlash. No extra button to press to release the card. Put differently, the XQD card changing experience is exactly what a pro wants: simple, fast, and reliable.
That said, be careful. With a GPS active, the WT-5 active, or the HDMI port active for shooting video to record externally, you're chewing the battery harder. Still good performance, though. I was able to run a D4 continuously for HDMI recording for a three hour performance (and surprisingly, the sensor didn't overheat during that time, either). So the battery is still very good for high-drain uses, just not as good as the old EN-EL4a was.
Writing to Card
There's no doubt in my mind that the D4 has the best write performance of any Nikon camera to date. If you're a buffer bully, you're going to have a difficult time intimidating the D4.
CompactFlash, not nearly as great. If you have an older UDMA-enabled card, it probably won't work faster in a D4 than it did in the D3. I only seem to be able to get a speed bump with UDMA-7 cards. Even then the max I've seen is 60MB/s (my older SanDisk Extreme Pro cards are about 50MB/s).
Which brings us back to the mixed card slot issue. Once you've shot with an XQD card in heated action, you won't want to go back. You'll be one of the ones wondering why there aren't two XQD slots in the D4.
One thing I didn't touch on in handling but definitely comes into play when we start looking at the focus system is the red focus sensor position indicators. Gone are the black in bright light, red in dim light sensors. They're always red in the D4. Moreover, the overlay that drives this leaves tends to "ghost" all the sensors (more visible in some light than others). Some users find this distracting, others find it useful (there's always been a subset of folk that ask why all the positions aren't always shown in the viewfinder; the answer is: it can be distracting, so if such a thing were to be implemented, it should be an option, IMHO). I'm of mixed opinion on this. Sometimes I find the ghosting useful, sometimes I find it distracting. Unfortunately, because it's an artifact of how the overlay lighting works, it ain't going to change and there won't be an option involving it via a firmware update.
Noise and Dynamic Range
Here's my usual underexposed JPEG at ISO 3200, direct out of camera:
Looks a little drab, as we'd expect an underexposed JPEG to, but you're not seeing much noise, are you? Let's take the same image in raw form and run it through absolutely basic Adobe Raw conversion:
Wowsa that's good. We're pretty close to ISO 6400 here after the exposure correction in Adobe Raw. I'm used minimal sharpening and noise reduction in the ACR tools. There's a small bit of residual noise, but look at the net: detail is still present. You can see it in the ball, too. So here we have an underexposed ISO 3200 image that cleans up terrific. How's that work out in actual shooting? Here's a slightly cropped frame:
Pretty darned good for ISO 3200 work. Note that the blacks look black, not a pile of dark noise. About as good you're going to get from a current production camera.
But you're all wanting a bit more about image quality, and that's where it's going to get a little difficult to describe. We have two prime candidates to talk about:
If you're not already at the pro camera level, then we have other things to talk about. US$6000 is a pretty big commitment to a camera body. The question you need to ask yourself is whether you're committing to the body or the sensor. Put another way, what problem are you really trying to solve moving up from, say, a consumer DX body?
As you might notice, it's tough to justify the big jump from an older, lower level body to the D4. I'd tend to say that you need to have all four of those "wants" I just described and plenty of cash in the bank to really make the leap from a lower, older model. Seriously look at what you gain by choosing a D4 over a D600: body build, double the frame rate, better AF system, lots of dedicated buttons, deep buffer, bigger battery are the main things. But you also lose some things, too: nearly US$4000, 8mp, and a built-in flash, for example. It's difficult to make the cost justification of a D4 over a D600 if you just want to go FX and have better low light performance than you've had on a DX body.
You've probably heard a number of things about the D4's video. Let me see if I can guess what they are: (1) the video is "soft" compared to other DSLRs; and (2) the D4 can create uncompressed HD video streams. Both of those are true, but not exactly good or bad.
I'm not entirely sure why, but the D4 video created in camera (H.264 compression) is clearly "softer" than that which comes out of the D800. I'm not sure that's a bad thing, though. The D800 is prone to motion artifacts on fine detail. Pan across something that is close to the pixel pitch on a D800 and you'll get annoying "sparkle" on edges as they move. That's not true of the D4, at all. It doesn't help that the D800's sub-sampling is wider spread across rows and columns than it is on the D4. Personally, I don't find the D4 video to be "bad." Just softer on edges than other cameras are producing (the Canon 5D3 is between the D4 and D800 in this respect). I wouldn't get too excited about this attribute of the D4 unless you're primarily considering it as a video camera. If video is your main use of the D4, you very well might be better served by another DSLR (though hold that thought until you've read the next paragraph).
And yes, the D4 can produce uncompressed video streams. What you're going to do with that is another story. Can you really afford 12GB a minute? The costs add up fast when you capture uncompressed video. Very few folk really need that. On the other hand, the uncompressed stream from the HDMI port is useful with external devices that record into usable, more cost friendly formats, such as ProRes HQ. I outline this in another article. I've found that ProRes HQ nets me about 100GB per hour of recording, and very good looking video that can be further refined in Final Cut Pro X or Adobe After Effects. Moreover, I've found the results a bit less soft than with the native in-camera compression (though it took me a bit of experimentation to find the exact camera settings I liked best with my video workflow).
I'd call the D4 a highly competent video camera with an external recorder. To get better HD output you'd need to spend a lot more money. However, the audio side in-camera leaves a bit to be desired, so there, too, we're going to go off camera. I personally prefer to capture both video and audio with my Ki-Pro Mini than to do so directly in the D4. As in "much prefer." The microphone amp in the camera is noisy, at best.
Let's face it, beginning with the D3 and now certainly with the D4 and D800 we've moved into a different world. The real question is "how much camera do you really need?" I would argue that the D3s/D3x combo was all a pro really needed for a considerably long period forward (remember, that period started in 2008/2009). While one can say that the D4/D800 combo eclipses the D3s/D3x in many ways, the question is "do you need those changes"? I'm betting that the answer should be "no" for far more of you than would actually say "no." It's easy to get trapped into the "new is better," "more is better" game, after all.
I've long been a proponent of the "upgrade every other generation" notion. To do otherwise is to spend a lot of money on smaller changes and not get the true useful life out of a piece of gear. Some pros think they live on the hairy edge: they need every small advantage they can get in order to stay ahead of the fast moving crowd behind them. I'm not so sure. I'd be willing to take on a D4 user with my D3s. The extra pixels would be nice, but thoroughly knowing my camera and not having to master something new counts for something, too.
This trend is going to worsen until we have the next major disruption. One can imagine the D4 replacement being 24mp, for example, another 20% gain in resolution with perhaps no loss in noise handling. One can imagine a D3x replacement being 36mp or more, though past 36mp the gains are going to be difficult to capture and decline in visibility in normal usage.
Look at it another way: 2.5mp D1h in 2001, followed two years later by the 4mp D2h, followed four years later by the 12mp D3. Those were big moves in almost all respects. Then we had the D3s two years later and the D4 three years later. Much smaller moves. A pro trying to keep up with all of this would have spent US$26,000 on bodies, less what they could get for their older bodies on the used market. If they kept one generation as a backup, the total spent would still have been over US$20,000 in 11 years, or US$1800/year to stay current. Of course, well-heeled amateurs don't think that way, they just buy the latest and greatest as bragging rights. Still, this is a US$6000 camera (up from the old US$5000 price of the pro generation cameras). You really need to know this is the right camera for you and that you're going to get a return for that money.
I'm not trying to talk you out of a D4. I'm trying to make you justify getting a D4. If you can justify it (or have infinite disposable income), then great, you'll get all the things that I've described as differences above. But as I kept looking the D4 specs, and now that I've shot with it for awhile, I'm not as sure that I could justify it for my shooting. It's a nice step forward, but a relatively small one in almost every aspect I use most frequently. Some full time PJs and event shooters will probably find a bit more justification than I do, maybe even a lot of justification. I'm just cautioning that you don't make the knee-jerk reaction to buying a new generation just because there is one. Had the D4 been a 24mp camera with the same attributes it has, the step forward would be a little more clear. But it isn't. It's the same 12mp to 16mp step that I cautioned a few about getting too excited about in the consumer world. A step. But only a step, not a leap. Up through 2008 we tended to get leaps. Now we're getting steps. That's all I'm saying.