|Nikon D7000 Review
Nikon Rolls Another Lucky 7...
|Add a comment or send Thom feedback on this article.|
First, the 24mp sensor in the D3200 is every bit the performer the 16mp sensor in the D7000 is. The increase in resolution is sure to raise the eyebrows of all D7000 owners and those considering it for purchase, and it should: Nikon really needs to move the D7000 forward to 24mp, and soon. To have the lowest end DX camera outresolve a top end camera is a bit of a problem for them. I suspect we'll see the D7000 get updated within six months given that imbalance.
Meanwhile, the D800 is a camera that you have to look at if you're buying the D7000 for "reach." Basically a D800 is really no worse than a D7000 shooting in DX mode, plus it has the advantages of FX and a 36mp sensor when shooting FX. Plus you have 5:4 and 1.2x crops in between. Wildlife shooters used to buy the top end DX camera (D300) because it had more pixel density than the FX alternatives, giving them more pixels-on-animal when you couldn't get close. Then the 24mp D3x came and confused things a bit (though it was expensive). Next came the D7000 which had 16mp instead of the D300's 12mp, and a better dynamic range, to boot. But the D7000 had a slower burst rate and a smallish buffer. Still, it tended to be a better choice than the D300 because it shot better in low light and yes, there's that pixel density thing: it had more.
The D800 has essentially the same pixel density as the D7000, and this solved a problem for a lot of shooters. If you're one of those dual landscape/wildlife shooters, the D800 is the best choice for you. You get a wide range of wide angle lenses and 36mp of glory for your landscapes. Plus you get the "reach" of the D7000 with a bigger buffer and some other options (that 1.2x crop proves to be very useful, giving you slightly less reach if you can get by with it, and even more pixels-on-animal if the crop allows it.
Under normal circumstances, we would have expected a D7200 right about now. The Japan quake and Thailand flood seems to have put Nikon a bit behind, though. The D4, D800, and D3200 all appeared later than originally intended, so it's difficult to predict when the D7000 update will actually appear. I'm thinking soonish. It very well may be the next DX camera up for refresh.
This doesn't make the current D7000 a slouch and a camera you should avoid. If you're patient you can wait for the replacement, but if you're impatient, you'll still get a very good camera. In two years of use, the only real aspect of the camera that causes me issues from time to time is the smallish buffer. Shooting large NEF rapidly, especially with some features turned on, can put you down at the one-second buffer mark very quickly. There are a couple of things you can do to mitigate that. First, make sure the features that consume buffer space are turned off. Second, use state of the art cards in the camera, and use matching cards in both slots. Third, don't use NEF+JPEG saved on a single card.
That said, the D7000 is still a strong all-around performer. Don't avoid it just because it might be getting towards the end of its life cycle, just look for a good discount on it.
The D7000 was late. Right at the margin late.
The expectation had long been that we'd get a D90 replacement (more on that in a bit) in August 2010, shipping almost immediately. The August announcements rolled by and only the D3100 appeared. Photokina loomed at the end of September, and it was clear that Nikon wanted to show the D7000 there, as they made a very un-Nikon move and joined the flood of just-before-show announcements with the launch information on the camera. But was it ready for launch?
I think the answer to that question was: barely. I need to explain that a bit. When Nikon launches a new camera, one of two things are usually true: (1) the camera has been produced for a couple of months and there is a large enough quantity available to launch them simultaneously worldwide (typical of most consumer DSLR launches); or (2) the announcement precedes production and we all wait until Nikon builds enough cameras for simultaneous worldwide launch (typical of most pro introductions). The D7000 announcement turned out to be neither of those things. The camera was announced about the time it went into production, first production units were released in the US as soon as there were enough (approximately 12,000 by my estimate), and worldwide launch didn't happen until a couple of weeks later, when another big batch of units managed to get out of the factory. As I write this in mid-December, almost three months after the launch, Nikon is still struggling to catch up to demand, as new shipments are selling out almost as fast as Nikon can make them.
Given that the design of the camera seems to have locked on time in early 2010, the strange announce/launch pattern suggests to me that the new imaging sensor didn't arrive as fast or as early as Nikon expected it to. Or perhaps not in the quantity that they expected (e.g. low yields). This is an important thing to consider, as one wrinkle in the D7000 ointment is sensor related. More on that later when I write about performance.
Before getting to the camera itself, we need to address another issue: is this the D90 replacement or isn't it? Nikon says it is not. But I think they're being disingenuous here. The camera is sized and shaped like the D90, it continues and extends the D90 feature set, it comes in at about the same price point (adjusted for currency changes), and it clearly is directed at the same crowd that made the D90 a huge success. So why didn't Nikon say that the D7000 is the D90 replacement? Simple: they don't have enough D7000 bodies to sell in the short term and they've got plenty of D90 cameras left to sell. So, until the D7000 production ramps up to fill demand and the D90 inventory disappears, we have the D7000 slotting in slightly above the D90 in the Nikon DSLR lineup. (Note that the same thing was true of the D3100 and D3000 as I was writing this review.) But make no mistakes: the D7000 is the D90 replacement. When we get to the middle of next year and more of the lineup has refreshed, that should be clear. Just as the D90 sat at the top of the consumer DSLR lineup, so, too, will the D7000.
Which brings me to the camera itself. If you've been following the critical serious consumer DSLR model from the beginning, you'll know that Nikon has slowly added performance and features. The D70/D70s to D80 to D90 to D7000 progression hasn't changed the basic body size and style, the price point, or the base feature set. But each iteration has been pushing up the image quality, performance, and slowly added some of the pro-level feature set.
Thus, it won't surprise you find that the D7000 and D90 sat next to one another look remarkably the same. Casual inspection from the front won't reveal any tangible changes to most people (though there are some). Even careful inspection of the back reveals only some modest changes in buttons and something new sitting under the Mode dial. It sure looks and feels D90-replacement to me.
But what a replacement. Nikon appears to have decided to defend the high-end consumer DSLR camera with everything they've got. The changes inside the camera are massive, and make for a far better camera than I think most people expected at this price point.
The D7000 is a top DX performer in almost every way, and is more than most people expected in a D90 update. But like some consumer cameras before it, Nikon has let a few minor things get in the way of perfection. At a slightly lower price it might have been five-stars across the board. Consider it four-and-a-half stars if you'd like. It's definitely good, very good.
The "Strongly Recommended" is new wording for me. It suggests that the product is somewhere between "Highly Recommended" and "Recommended."
Help Support This Site
Help support this site at no cost to yourself. Use my preferred vendor, B&H, to make your purchase (they ship Internationally as well as to the US). See this page if you'd like to know how this works and why it costs you nothing but benefits me. Meanwhile, here are the direct links to the B&H order pages for the two primary buying options. Use these links to support this site:
My thanks to all that have supported this site in the past.
Review product source: two purchased D7000 bodies.
As noted earlier, at first glance the D7000 seems like only a modestly updated D90. The body appears externally, with only a modest number of changes, literally that of the D90. Same size, same basic layout, same eyepiece, and same flash. From such a casual glance, the D7000 underwhelms. Don't let that fool you. The camera has undergone massive changes from the D90 it follows.
Those changes will make even the D90 owner salivate and consider updating, which I'm pretty sure was Nikon's intention. They touched a nerve with the D90 and brought a lot of new (or returning) Nikon users onboard, but the D7000 gives even those folk some reason to consider upgrading. The only change that'll cause groans in that list of changes is the battery change. Even there you might find a tiny bit of good news: the new batteries can't easily be discharged accidentally due to their connector protections, and Nikon has tried (only partly successfully) to mimic the Apple-style chargers, where you can go either way: direct to wall socket or use a cord to extend reach. As I understand it the battery change was necessary, as new rules in effect for Japanese export make it so that you can't have Li-Ion batteries with exposed contacts, as did the old EN-EL3 type batteries.
Overall, an impressive list of changes. So we've got a lot to explore, both in describing the changes and in determining whether they make any useful difference in performance or ease of use.
Let's start with the sensor. It's clear that the sensor was fabbed by Sony. Nikon started out quietly saying that this is their own sensor and unique to Nikon. From the specifications that appears to have some truth in it, as the video specifications are different than the equivalent sensor in Sony cameras. But since the appearance of other cameras with what looks like the same sensor and the fact that Nikon has backed off using the word "Nikon" in association with the sensor, I'm guessing that very little is different with this sensor from the ones being used by Sony and Pentax, if anything. I did note some unusual changes to the Bayer filtration with this sensor: blue photosites are very different on this sensor than on previous Nikon/Sony sensors. You can see that very clearly when looking at the RB coefficients under various lights and white balances.
One key element of the new sensor is 14-bit ADC support built-in. Like the D300 and D3x sensors, the ADC is built into each photosite on the sensor. Unlike the previous iterations, there is no loss in frame rate performance by enabling 14-bit. The video attributes of the sensor still seem to lag behind the times to me, though. Contrast autofocus performance--which is what the camera uses in Live View and Movie modes--is driven by the frame rate coming off the sensor. Drive at 24 or 30 fps and you get 30 data sample points a second, drive at 60 fps and you get double the sample rates. One of the reasons why the new Panasonic GH2 is so fast in video focus performance is that it is driving its sensor at 120 fps. Nikon seems a generation or two behind in their implementation here, and 1080P/24, while better than any previous Nikon DSLR, is still not up to the level of what some other DSLR cameras can do. State of the art right now should be 1080P/24/30/60. The Nikon is 1080P/24 or 720P/30. At least the frame rate values are broadcast correct (23.976 and 29.97 respectively).
On the plus side, once the data is off the sensor and into the EXPEED 2 architecture, some good things happen for the video. Nikon is using a relatively modern H.264/MPEG-4 compression (Part 10), complete with so-called B frames. Many similar compressors don't use the forward/backwards looking compression of B frames because it takes so much computational horsepower, but having B frames makes large motion components look better than they would without them. If you're ever read about "mushy-looking" video when lots of motion is involved, it's likely that the compressor is not using B frames. Videographers will want to know what the bit rate is on the D7000, but Nikon isn't talking. I've heard estimates of about 20-24mbps, which is decent, but not earthshattering. Personally, I go by the visual look more than the actual number. We'll get to that in the performance section. Audio has been vastly improved with the D7000, too. It uses a stereo 16-bit Linear PCM recording at a 48Kbhz sample rate. Of course, with the monaural microphone located next to the lens, it picks up virtually every zoom or focus operation; you'll want to use external stereo mics on this camera.
The focus system is all new, too. And it integrates with an all-new metering system. Both these things were pretty much unexpected. Nikon rarely updates these systems between pro cycles on consumer bodies.
The focus system is a 39-area system similar to the 51-area one used by the D300 series. It has 15 cross-type sensors at the center, for better low-light handling, but the interesting thing is that it has more actual sensing area than the 51-sensor version. Not overall area of the image frame, but sensing area of each individual detector. That means the line detectors are slightly longer and fatter on the D7000 than on the D300 series, which is a good thing.
Coupled with the new focus system is a new metering sensor in the viewfinder area. This new 2016-pixel part does color discrimination that integrates into the Auto Area and 3D Tracking aspects of the focus system. The focus system itself is controlled in a new way. Instead of extra buttons and switches on the top and back of the camera, the Autofocus mode lever next to the lens (labeled AF and M; used to be S, C, and M on high-end cameras) now has a button in the middle of it. Press that button and the Front Command dial controls the Autofocus Area mode and the back button controls the Autofocus mode (yes, I know Nikon uses the words "focus" and "mode" in seriously confusing ways with different functions all having labels that look the same until you read the words really carefully; it's one reason why my books sell so well, as I explain the differences clearly and precisely ;~).
The viewfinder that you're looking through to see the focus system settings (yes, you can change focus settings without taking your eye from the viewfinder--hurray!), is now 100% view, which is nice. It's still a little lacking in some information, like the D90 was, but you get used to that.
Three very useful additions come to the D7000 from the higher-level bodies and set the D7000 even further above the D90: AF Fine tune, Non-CPU Lens Data, and Intervalometer. Bravo. Those of you with older manual focus Nikkors (AI or AI-S only) will find that they work with the meter in the D7000; those of you with exotic long lenses will be able to tweak the camera/lens combination to its fullest; and Intervalometer makes for an interesting addition for all those video freaks using the D7000 (allows you to take frames for a time-lapse video). Very nice, Nikon. And overdue.
Overall, a lot of significant feature updates, many of which move the D7000 to the top of the DX camera ladder.
Which brings me to the D300s. I've noted a lot of people getting confused by the fact that the D300s now seems to be lagging in some areas of performance to the D7000, thus they believe it can't be considered the top DX camera any more. Given that the D300s is really a three-year old camera (the s revision really was a very modest mid-life tweak), it's to be expected that the sensor and image processing will be a generation behind the D7000. The same thing will likely happen to the D7000 when the D400 gets introduced and moves another step forward, probably sometime late next year. About half the DX lineup iterates at a time, and there's a constant leapfrog effect that happens in the lineup because of it.
But let's not confuse something: the D7000 is not up to the level of the D300s in some critical ways: build-quality, feature set, buffer use, and more. Little things, like the material used over the color LCD, can make somewhat substantive differences in whether a camera manages to take full time pro abuse or lives only up to a lower level of consumer handling. The materials, durability, ruggedness, and other build-quality aspects of the D300s are executed at a slightly higher level. If you're bouncing your camera around in the Outback all year long, the D300s is going to stand up to that better than the D7000, though the D7000 will stand up to it better than a D90 would. Plus the D300s still has some performance advantages: 8 fps maximum frame rate and a critically larger buffer, for example. On the plus side, the D7000 has come mighty close to the D300s level in terms of build quality and features, so I'm sure there will be people who opt for a D7000 over a D300s. But I don't consider the D300s obsolete because the D7000 appeared. It's still an excellent camera, though starting to show its age a bit. I fully trust Nikon will address that next year and then the D300s/D7000 debate will just go away entirely. Until then, I think you have to pick the D300s if your camera handling is going to be abusive and rough, the D7000 if you value image quality and performance (other than frame rate and buffer size) over ruggedness.
Put simply, the D300s leans pro, the D7000 leans consumer.
Surprisingly, there are many handling differences to write about, despite the fact that the D7000 doesn't show all that much external difference in controls than the D90.
I've already written about the focus settings. Bravo. Every future Nikon should have this new Focus Mode switch/button design. On the flip side, Nikon needs to do a little better at describing what's being changed in the viewfinder. The use of the LED strip at the bottom of the viewfinder means that there aren't many characters that can be displayed, so you get cryptic AF S, AF C, 3D, and other indicators, and it gets worse with Dynamic Area AF settings. If you understand the AF system well, you won't be confused by the shortcuts in naming, but newcomers have a lot of homework and catching up to do, and the use of confusing acronyms (AF S) isn't going to help. It's past time for Nikon to address their naming (and please, pretty please, stop calling everything a mode!).
Shooting Method is now set via a locking dial underneath the Mode dial. This new sub-dial operates much like the pro camera Shooting Method dials do, and includes Self Timer, Remote, Quiet Mode, and Mirror Up positions as well as the single frame and continuous frame settings. This is another big step forward for the consumer body. It's not without a bit of a new problem, though. The Remote setting (for the IR wireless remote) still has a time out, so you can actually get the camera into the appearance of a non-functional state. Likewise, the Remote setting doesn't allow the main shutter release to work at all, which I think is a design mistake. Modalities are always a problem in complex systems, and I think Nikon missed a better solution here (shutter release is active, plus a partial press resets the Remote timer clock).
Most of the (useless) 21 Scene modes (there's that word again) are set in a slightly confusing way. Either my manual is wrong or my cameras are wrong ;~). The exact control sequence is: (1) rotate the Mode dial to Scene; (2) either press the INFO button or rotate the Rear Command dial once with the meter active; then (3) rotate the Rear Command dial to select the Scene setting you desire. Nikon's manual leaves out Step #2 or my cameras are broken. I'm guessing the former. Which is going to be really silly design because true beginners, who might be tempted to use the Scene modes, will be slightly confused by this. Worse, if the camera isn't active (exposure meter running), rotating the Rear Command dial doesn't always bring up the Info display, which is why you need all three steps. Whoever is designing Nikon's beginner user interfaces simply isn't designing them to the right level of discoverability and usability. My old friend Alan Cooper would straighten this design silliness out in a heartbeat. Hint, hint, Nikon (at least have someone read and take to heart his seminal book on UI design: About Face).
I'm a bit on the fence about the U1 and U2 (user settings 1 and 2) positions on the Mode dial. Yes, this means that you can have your camera set up to as much as three different configurations with a turn of a dial (PAS or M, U1, and U2). But there's a bit of clumsiness in the design there, too. We also have a Save/load settings option in the SETUP menu, and this is located on a different page from the Save user settings that puts your settings into the U1 and U2 positions on the Mode dial. What!?! Was someone on drugs in the firmware department the day they added this? There should be one Save Settings function (U1, U2, card in Slot 1, card in Slot 2), and one Load Settings function (U1, U2, card in Slot 1, card in Slot 2). Plus you should be able to be set on U1 and load the settings from U2 (and vice versa).
So I applaud the ability to quickly set the camera configuration differently, but there's a unusual complexity to it and gaps in what we can do that come about via the menu design. It's the old Nikon giveth, Nikon taketh thing all over again.
I've complained about this next one before, but on a consumer camera it is glaring: no one seems to understand which menu functions you use most. Role played by card in Slot 2 is the fourth function on the SHOOTING menu while Set Picture Control is the 11th function and on page two of a scrolling menu screen. Which do you think a consumer is going to use most? Burying ISO and White Balance are a little more forgivable since we have button equivalents to adjust them quickly, but still, Nikon is totally oblivious about hierarchy and organization in the menus. One could even argue that Role played by card in Slot 2 and File naming are actually SETUP functions, not SHOOTING functions. Simply put, Nikon's done nothing to clean up their menu system, and with so many settings involved on this high-end consumer camera, it just feels awkward in use. The good news is that the pros moving down to a D7000 will find the menu structures pretty much what they're used to. I guess we pros haven't complained enough to get Nikon to do some long-overdue cleanup. Now the consumers will suffer. (Lest you think this is an invalid complaint, go pick up a D3100. Nikon has reorganized the menus on that camera, and Set Picture Control is number two on the SHOOTING menu there, while White Balance and ISO are also on that first page. Hey, whoever re-arranged the D3100's menu seems to understand that there might be a preferential order for menu items, maybe Nikon should let them play with the older engineers and the higher end designs, too. ;~)
Going into Live View or recording videos has been simplified from the D90, which is nice. The display is getting more cluttered during these uses, though, and we're still missing manual control of apertures and exposure. In fact, Live View can be very frustrating in Manual exposure mode, as it really only catches up to your changes when you take a picture or exit and re-enter Live View mode. I'm pretty sure this isn't the behavior people expect.
Unfortunately, Nikon has once again made small changes that'll stop multiple body users cold if they encounter them. For example, on some previous Nikon's the NEF(RAW)+JPEG image quality was supported for White Balance bracketing. Yes, you guessed it, instead of adding NEF(RAW) to the image qualities that you can bracket white balance with, just like the D90 Nikon removed any use of NEF. I'm not sure what Nikon is thinking here. Apparently they don't understand some users' workflows. Those of you who were extracting the embedded JPEG to save space on your cards have always been shut out from bracketing white balance, but now any use of NEF kills the ability. Moreover, there's not even a "not supported" message from the camera should you set your camera to bracket white balance, select a raw image quality, and then press the bracket button. Instead, nothing happens. You'll think the bracketing button is busted on your camera if you haven't caught this nuance.
As if all that isn't enough, Nikon has once again made small changes to the way flash and automatic ISO work. Flash exposures are indeed more accurate than ever, but with automatic ISO active, figuring out who's doing what has taken on a new level of complexity. In the old methods (there were two), Nikon either ignored the flash with Automatic ISO active or ignored Automatic ISO until the flash output would be exceeded. The D7000 does something in between ;~). It uses the flash sync speed as the new "minimum shutter speed" in Automatic ISO, regardless of what you entered. We don't need Nanny's, Nikon, we need flexibility. At worst case, a nanny should be optional.
Nikon now allows MUP to be set with Interval shooting, but I wonder if the engineers have completely lost touch with the reasons why we want mirror up in the first place. Because metering is done via the sensor in the viewfinder, the camera brings the mirror down after each shot is taken. Just before (and I mean just) the next interval shot is taken, the mirror is pulled up and there's a very short delay before the shutter is opened. So brief that I can still measure the vibrations in my Gitzo 1325. So, basically, useless. You're better off using Custom Setting #D11 (Exposure delay mode) with Interval shooting. Note to Nikon: if MUP is selected when Interval shooting is started, take a final exposure reading, flip the mirror up, wait a second, and start the intervals, keeping the mirror up during all the interval shooting. Yes, that would mean that we can't use i-TTL flash with the combo (well, we could if you treated it like FV Lock was active after the first pre-flash). But someone looking for this combination knows what they're doing (or should). What they're looking for is absolute rig stability, not all kinds of fancy automatic handholding.
Fortunately, the engineers did manage to get the message elsewhere. The infrared remote now can do Mirror-up! First press is lift mirror, second is take the photograph. Simple and effective. Given the second rear IR detector, the ML-L3 is probably the remote control of choice for the D7000. Even nicer: it's cheaper than the wired remote, by far. (And if you like "real cheap," I've found some ML-L3 compatibles on eBay for as little as US$4.95.) But why did Nikon not leave the shutter release active when remote is set? Because remote times out, you can get into a state where the camera seems inoperative (you have to press the shutter release partway to reactivate the timer, though for some reason the metering seems to still be on). This needs to be fixed. The shutter release should be always be active when the camera is active. I thought that Nikon had learned that after the D1 modality fiasco, but apparently not.
One step backwards (or at least sideways) comes in the form of the MB-D11 optional grip for the D7000. On the plus side, it has a nice overall feel and a full set of controls (including a joystick-like Direction pad). For my hands, the Direction pad and AE-L/AF-L button feel a little off in placement, but not enough for me to worry about. If you have big hands or operate the grip with gloves, I suspect both will be problems, though. The method of attachment (strong connector and two long guide pins) should make it more secure, which is a good thing. The bad news is bad, though. No support for the EN-EL4 batteries and...here's the big drawback: if you're buying the grip to get two-battery capability, you'll be taking the grip off every time you want to charge the in-camera battery. My problem with detachable grips has always been that they are a bit of a weak point, especially for those that use brackets or tripods with them. That weakness is amplified by constantly taking the grip on and off, in my experience. But that's exactly what we have to do to get to the battery in the camera to charge it.
And then there's the ugly: PC-E lenses. Nikon claims the PC-E lenses are compatible. I want some of what they're smoking. To even get my 24mm PC-E on the camera I scratched the underneath of the flash housing, it's that tight. Once on, the shift lock knob is basically inaccessible, and Nikon's build quality on the PC-E lenses absolutely requires that you lock things down tight to get correct alignment. Don't plan on using PC-E's with the D7000. Of course, none of them are the focal lengths you'd really want with a DX sensor, so it's not a huge loss. Still, it's getting silly. The D700 was tight, the D300 is tighter, the D7000 is even tighter. It's as if Nikon doesn't actually want to sell any PC-E lenses to anyone but D3 series users. Where are these kinds of design decisions coming from? Left hand, meet right.
Finally, there's the issue of connector access. Nikon changed the design of the flaps covering the connectors on the D7000, making it near impossible to remove the doors temporarily. This comes into play in a number of ways, but the flaps tend to stick out now when you've got something plugged into the left side of the camera (as viewed from the rear). It becomes a real issue with L brackets: connectors just start to be impossible to use if you've got an L bracket on the camera. Yes, I'm aware that some L bracket makers are trying to "adjust" for this by letting you slide the bracket out (which compromises support) or removing the L temporarily, but neither of these things are the answer. Simply put, camera designers have no clue as to how we're using their products and what we mount on them. While the new door design is really nice for casual users who only access the USB connector when home downloading their images, the new door design is a pain in the butt for anyone who shoots with the camera with something connected (GPS, wired remote, WT-4...hey, wait a minute, those are accessories Nikon wants to sell us!). Right hand needs to meet left hand and get together here.
Still, if you're a D90 user--or any recent Nikon DSLR user--the D7000 will prove to be remarkably like what you're used to. Things that you probably take for granted in handling you can probably still take for granted. I would appreciate Nikon going a step further and not introducing so many small glitches, though. I'm pretty sure there will be something that catches even Nikon faithful by surprise when they pick up a D7000, and everyone will get a bit frustrated by the continued disorganization in the long menus.
Looking for useful D7000 accessories and want to help support this site?
Here's a list of accessories that Thom has tried and recommends, and which work with the D7000. Be sure to read my short notes in the wishlist for further comments.
Time for the rubber to meet the road (okay, the photons to meet the electronics). Since so many things have changed internally from the D90 there must be some differences, right? You bet your sweet bippy there are.
The one drawback to the new battery system is that it isn't exactly a fast charging system. Run a battery down fully and expect to get it back topped off in two-and-a-half hours. So we won't be calling the MH-25 a Quick Charger.
Writing to Card
Raw shooters will not be terribly happy with the buffer size. While you can theoretically get to 15 buffered images shooting raw, you're throwing away highlight data (Compressed NEF) and bits (12-bit) and avoiding a lot of features of the camera to do so (though they'd be applied only to the embedded JPEG). This isn't a sports shooting camera in raw, IMHO. You'll hit the buffer limit real fast. Set for optimal image quality, you may have only a one second buffer in raw. This, if anything, is one clear difference between the D7000 and the D300s. Even a short delay in hitting the buffer can be meaningful to the burst shooter.
Mirror up, well, that's another story. While Nikon touts the full-time contrast autofocus ability of the D7000, it's marginal in performance, at best. The D7000 is not going to replace your autofocus video camera. I consider the contrast AF to be a focus (slowly) once system, not a usable full time one. Yes, it's that bad. If you don't believe me, pick up a Panasonic GH2 and compare it side by side with the D7000 on moving subjects while recording a movie. It's a night and day difference.
I've noticed a bit of chatter on the net about "overexposure." But that's not what's really happening with the D7000 metering system. No, it's that color matching and pattern matching coming into play. And correctly, I think. Let's say, for example, that there's a skin tone in the foreground of your scene. Perhaps the person with that skin is even a bit backlit. Well, the D7000 certainly sees that skin tone and knows where to put it on the tonal scale. But in previous Nikon matrix meters, if the background was producing values that would blow out the histogram, the matrix meter tended (but not always and not completely) to preserve highlights. I don't see as much of that with the D7000 (except in single servo AF). It's not going to preserve those highlights at the expense of what it thinks is "subject." It certainly won't preserve them as much as previous Nikon matrix meters, even when it decides to do so. Two other things play into the "overexposure" issue. First, there's gamma. People coming from older (pre-D3) Nikon bodies and seeing Picture Controls for the first time are reacting to the mid-range boost that the default Picture Control applies compared to the old style image settings. Second is contrast. The defaults (and many of the other Picture Controls) push contrast a bit, and that has a tendency to make bright seem brighter.
The corollary is that if you pop up the flash for some fill, the D7000 seems to get that exposure just a little more on target than previous consumer cameras. Nikon's obviously done a lot of tweaking, and for those of you coming from another Nikon DSLR, there's going to be a learning curve before you manage to fully grock the new matrix patterns and tendencies.
However, all isn't perfect. Be aware of one very big caveat: when the scene you're metering hits 16.3 EV, the matrix metering system gives up and sets its value for 16.3 EV, no matter how much more light there may be. EV 16.3 at ISO 100 is f/11 at 1/500, which is barely beyond Sunny 16. This won't occur all that often in your shooting, but it does occur sometimes, so make note of that. In really bright light conditions (snow, beach, etc.) you probably need to be in centerweighted metering.
White balance is decent to good, especially in mixed lighting. The direct Kelvin settings once again didn't match my Minolta Color Meter or my Imatest measurements, but they're closer than some previous Nikon DSLRs. The range over which Auto WB works well is more limited than Nikon suggests, though from about 4000K to 6500K it can probably be trusted. I'm not a fan of the new Auto Warm setting, but I understand why Nikon put it in.
Surprisingly, the Red and Blue channel raw responses are a bit different than we're used to in Nikon DSLRs. I was quite surprised when I started putting together my UniWB file for my Complete Guide: something seemed different in the channel responses. Clearly, there are Bayer filtration changes on this camera compared to other Nikon/Sony DSLR/sensor combos. The Red channel is not as good, the Blue channel is better. This is likely one of the reasons why people are commenting about the improved noise (most low light situations are warm light, so the Blue channel is often the first and highest noise producer). But this means also that if you're a raw shooter and you've got ACR/Lightroom/whatever presets that you've dialed in for a D5000/D90/D300, you're going to be redoing them.
Which brings me to my final comment about color: it's ever so slightly subdued (some would say flat) from what I see in my D3x. This tends to happen on very tough to capture colors, like golds and bronzes, but it can happen in other colors, as well, including skin tones. It's a very subtle difference, and many people won't notice it. But if you're pushing for every last little bit of capability from a raw file, you're likely to notice it at some point. I'm still trying to figure out if this is just a metamerism that we have to live with or whether you can overcome the slight flatness via post processing.
Hot pixels, on the other hand, may be a small step backwards. There's much more to that statement than at first appears. The D7000 is somewhat prone to hot pixels on longish (but not long) still exposures or when the camera is hot and shooting at high ISO values. Live View and Videos are sort of the worst of both worlds, as the sensor is on for long periods of time and gets noticeably warmer with constant use. And Live View/Video use ISO boost to set exposure. In such cases, a few hot pixels may show up. Much less so with just shooting long exposures (over one second). Hot pixels, of course, bothers the heck out of those that like examining images at 100% view, but the hot pixeling is just strong enough that it may bother others, too. More on that in a moment, but first we have a slight detour to make.
Nikon has changed their hot pixel suppression scheme (yes!). All of us pros have long complained about the crude nearest neighbor filter that Nikon applies on all long exposures, even in raw data. The old method tended to obliterate small bright detail. For example, dim stars in a night sky. In some cases, it would put a black pixel in the middle of a small bright object. Well, Nikon is using a new method now, thanks partly to Marianne Olelund. A number of us who go beyond pixel peeping (all the way to pixel examination and digestion) have been complaining about this problem for years. With Marianne's permission, when I met with Nikon engineers in Japan in early 2010 I presented her improved method and the case for using it. It was the one confrontational aspect to my presentation, as to demonstrate the problem I had to essentially show that the old method was incorrect, something you don't normally do in a meeting full of Japanese peers, as I did. Someone in that room lost face because of my presentation. But I was willing to take the risk (and got about ten minutes of defensiveness and frowns all around in response). Fortunately, it appears that Nikon got the message, and has adopted something very similar to what Marianne proposed. This brings both good and bad news. The good news is that I'm seeing highlight detail that I don't see from the older Nikon DSLRs in long exposures. It's especially noticeable in night sky shots. The bad news is that even with Long exp. NR turned On you'll now sometimes see hot pixels. Not many or often, but the new routine is not a brutal repression scheme that kills all bright data. The new routine is allowing true neighbor photosites to actually deliver more of their data into the raw file, and sometimes you might encounter adjacent photosites producing a hot pixel. Doh! I guess we have to be more careful what we ask for. Still, this is a very positive step forward for long exposure raw shooters. And no, this is not really related to all those discussions of "The D7000 has a hot pixel problem." Not in the least. What I'm talking about here is in very long exposures (one second or more). What all those forum posts are complaining about is...well...um...well, what are they complaining about?
As best I can tell, most of the "hot pixel complaints" seem to boil down to a few things. First, there's the "don't quite know what I'm looking at folk." These people leave the lens cap on, fire up Live View and zoom all the way into the image. "OMG! I see strange pixels and patterns!" they proclaim. Uh, yeah. The camera is trying to boost ISO high enough to give you a picture (of the inside of the lens cap), because that's how Live View works: it changes ISO to change the exposure for the video stream you're looking at. It doesn't stop there, though, the video stream isn't the data stream for a still picture. It's a lightly processed data stream that is actually fairly crude. Simply put, you can't evaluate hot pixels from that stream. The other problem is that, in addition to Live View, most of the complainers I've seen are also running long video streams and basically have the camera jumping through hoops as they constantly pixel peep to see if they've got a problem. Paranoia has this nice circular nature to it: the more you think there might be a problem, the more paranoid you get, so the more you examine things in ways you don't actually use the camera. I can just see someone doing the Lens Cap Test, then getting freaked out by that and running video and more Live View and more video and more Live View until they've got the sensor up to a nice toasty froth.
Personally, hot pixels don't really bother me, and the two D7000 samples I've been testing have to be provoked to a very high degree before I see any. (You need to provoke a D7000 much more than the D80, for instance, to see a hot pixel.) I can also develop techniques and strategies to mitigate or eradicate them. If you shoot raw files, many converters, such as Adobe's, will remove them automatically (though, of course, not Capture NX2 ;~). Still, the Internet masses continue to complain about them as I write this, so there's a group that do find them obnoxious, to the point that they're returning perfectly fine cameras and exchanging them for another camera that's likely to be essentially the same as the one they turned in.
So let me say this: chillax.
The D7000 is around mid-pack in the long line of consumer DSLRs in terms of hot pixel tendencies. One of my samples stopped producing one hot pixel after using it normally for a couple of weeks. Yes, that sometimes happens. Some camera manufacturers actually go out of their way to make that happen (i.e. if they detect a hint of hot pixelness in a data stream, they just start mapping out that data and putting false data in its place). But a true hot pixel is not necessarily permanent (a permanent hot pixel would be called a stuck pixel). If you do end up with a stuck pixel, you can have Nikon map it out. But I'd never have them do that until I'd used a camera for a month or more and knew that it really was a stuck pixel. Try actually shooting with the camera normally for a week or two before coming to a decision about what to do with it. And I mean normally. Let's lay out the whole hot pixel bit:
If you examine that list, you'll find that, in actual practical use, hot pixels shouldn't be an issue. This is why examining a zoomed in, slightly processed video stream in Live View over and over with the lens cap on is futile in predicting whether a camera actually does have a real hot pixel problem. You're out of the bounds of regular shooting, and you're not looking at what the camera would produce in an image.
Let's talk about the noise that you want to know about: basic pixel noise (shot noise, read noise, and so on, which shows up as luminance and color variances in your images). Wow. Yes, one word: wow.
Before people misinterpret that word, let me be clear about something: the D7000 is not down to D3/D700 noise levels, let alone D3s noise levels. Nevertheless, it reaches unmatched levels for a crop sensor. Considering that we're talking about 16mp here, that's saying a lot. Most of us were expecting a very modest improvement in visible noise from the D90's 12mp sensor, if any improvement at all. The reality is that the D7000 made a pretty significant, and clearly visible, step forward. The interesting thing is that the actual numeric values produced may not seem all that incredible. But I've long pointed out that numeric values in tests (usually standard deviations in middle gray targets) don't necessarily tell you enough. It's the way the noise looks that's just as important as its numeric value. Maybe more important. It's on this latter bit that recent Nikons have been a dramatic break from past ones: starting back around the D3 Nikon seems to have gotten a strong handle on the "look" of noise, and chroma noise is tightly controlled while luminance noise is allowed to show through a bit.
Everyone expects my basketball shot by now, so I won't let you down. Here's a 100% view of a slightly underexposed NEF taken directly out of Capture NX2 (no NR applied, camera defaults):
There's some color noise in there, but not as much as we've seen in previous DX Nikons. The overall level of noise is clearly visible (again, NR is set to Off), but not obnoxious. Here's the same NEF run through Adobe's ACR 6.3 conversion and trying to balance sharpening against noise reduction:
A very clean, usable result, despite some smudging by the noise reduction. JPEGs out of the camera with noise reduction set to Normal are actually slightly better than this. Remember, this is 401 actual horizontal pixels from a 4928-pixel image. On my 30" display this shows as about 4", so I'm looking at the equivalent of a 49" print here. (Your equivalent will almost certainly vary; just measure what you see and multiply by about 12 to get a rough estimate of the overall image size this 100% view represents.) (Further note: I didn't like all that smudging, so I spent more time with the image and was able to get an even better rendering by some very careful tweaking of sharpening and noise reduction. I won't show that here, but let's just say that my "better" version, above, is not the "best" possible version.)
Overall, the D7000 is the best cropped sensor camera I've seen to date in low light (i.e. set to higher ISO values). It's not perfect, nor is it a D700, let alone a D3s. But it's better than a D90 or D300s in the same situation, and clearly so.
You'll see a lot of numbers in various camera tests and reviews on the net. Most of them are some variant of standard deviation tests. You take a patch of plain color (or gray) and measure the standard deviation of the pixel values (or some variant on that). This gives you numbers that you can compare. You'll note that i've backed way off reporting numbers these days. That's because two standard deviation values that are the same can look very different. I've learned to trust my eyes more than my calculator (any professional photographer who says the opposite is really an accountant ;~). With that in mind, I'm going to make a generalized statement: the D7000 is about a stop behind the D700 and two stops behind the D3s. Visually. Put another way, if I'm happy with my results on my D3s at ISO 6400, I'm happy with my results on my D700 at ISO 3200 and on my D7000 at ISO 1600. Those numbers are actually pretty close to the values where I'm very comfortable shooting without worrying too much about noise (there will be noise and some loss of dynamic range, and I'll have to do something about that, but it's not difficult post processing work). That's not to say I wouldn't push any of these cameras higher in a pinch, but then I know I'll be faced with some serious post processing work to get the quality I want.
I suspect most of you reading this review will actually be happier with the D7000 than that previous paragraph might imply. First, it's not likely that you're running ISO 3200 shots up to 49" prints ;~). Just to prove that point, I just brought the full image without noise reduction used for the samples above into full view on my 30" display (due to Capture NX2's window methods, that doesn't quite stretch the whole display--I end up at 45% view). At this size you can still (barely) see that there's some noise in the image, but it just looks like a small grain texture. At 25% view I can't see the noise texture at all. Similar things happen in prints. By the time I get down to about 10" prints from the entire frame (which represents 490 pixels per inch given to the printer, or a resize by half and giving the printer 240 pixels per inch, which is still high quality), most of the noise is pretty much buried where you have to get out the loupe to see it. Thus, I'd say that you can use very gentle noise reduction techniques in post processing and get very printable ISO 3200 images out of this camera in sizes up through what your largest desktop inkjet printer can handle (13x19"). How many of you need more than that?
But let me put this in no uncertain terms that everyone should be able to understand:
While it's the best performance of any Nikon DX body has produced to date, the D7000 is not in D3s territory.
Some of you probably want to know how it compares to the competition (Pentax K-5, Sony Alpha 55, Canon 60D). In a word: similar. It's impossible to do apples-to-apples comparisons because there are plenty of things that get in the way of that. But in playing with raw files from each of the cameras with the same converter, what I wrote in the bullets just above tends to be fairly similar across the cameras. The Sony and Canon seem to "lose it" slightly faster than the Nikon and Pentax when you start pushing above ISO 800, but all produce a "usable" ISO 3200 result. The Pentax seems to hold color slightly longer than the Nikon, but it also clearly starts adding noise reduction (even in raw) at the highest ISO values, which means that edges look more anti-aliased. Realistically, if you're nitpicking noise between these four cameras, you're nitpicking. I would guess that if you need ISO 6400 in a DSLR, lenses are going to be a bigger determinant of your final results than the noise levels. That said, I don't like the look of the Pentax files above ISO 6400, or the look of the Sony or Canon files above ISO 3200. The D7000? I don't like the look of its files above ISO 12,800. But if you think about that for a moment, we're talking two stops maximum between all the contenders, and probably less (I doubt those of you reading this are nearly as picky as me). That's the difference between using a kit lens (f/4 on average) and a modestly fast prime (f/2). If you're not already maxing out a fast prime, well, you really shouldn't be pixel peeping noise. And like I just wrote, if you are using a fast prime, the lens you're using is going to determine your final results as much as the sensor's noise.
The AA filter in the D7000 also seems "tight" to the Nyquist, unlike some Nikon DSLRs where it seems like it is set clearly below Nyquist (and thus robs a bit of acuity). Like the D3x, there's a tiny bit more "bite" to the edges the D7000 produces with a good lens than you might be used to with previous Nikon DSLRs.
"Practical DR" is a whole 'nother story, but unfortunately any such measurement is often subjective (an alternative view on that in a moment). Moreover, it'll also be determined somewhat by how you convert your images. I define my own DR numbers by "held bright sky detail" to "start of clear visible noise in shadow tones." When we started the digital DSLR era, that number was well below 7 stops. For much of the middle of the era, in was in the 7-8 stop range. The D3 (and D3x interestingly) pushed me up to perhaps 9 stops I consider fully usable. So where's the D7000? My preliminary use tells me it's right there with my D3x.
Remember that dynamic range (both engineering and practical) reduces with each boost of ISO. This is where we see some clear difference between the D7000 and the Nikon-sensors in the FX bodies. For example, on the D3s I see virtually no tangible change in practical DR out through ISO 800. On the D7000, there is a clear hit to practical DR at every ISO boost. That makes it very difficult to produce finely ramped tonal ramps as you boost into a high ISO, and it means that dark colors get muddied first and progressively brighter colors lose their accuracy as you boost ISO. This all ties in with noise (after all, noise is what is defining the low end of our DR). It's one of the reasons why I say that ISO 3200 is quite usable on a D7000: the practical dynamic range is still in a reasonable range and colors and tonal ramps haven't been visibly compromised. But don't get me wrong, if you need ISO 3200 for birds in flight before sunrise at Bosque, you don't have a lot of dynamic range to work with. That's exactly where a D3s starts to outshine the D7000 big time.
Ditto with the jello effect. The D7000 definitely has some rolling shutter issues, but not nearly as severe as those in the D3100 (which I characterize as the worst I've ever seen on a DSLR), but also not as good as I see on the Panasonic GH2. The rolling shutter issues are acceptable for all but the high-action video level (note that Apple's iMovie has a rolling shutter correction feature, and it works very well on D7000 videos). If you're thinking about shooting skateboarders whizzing by you from a moving skateboard platform, this might not be the camera for you. But as long as you keep your pans and subject motion down to, say, Woody Allen levels, you'll be fine.
The new compression engine for the most part shines. It takes a great deal of motion to start getting "compression mush" at the High Quality settings. So again, you Woody Allen type directors out there will find that the D7000 serves you just fine. The new audio support, however, is awesome, and puts the audio quality right where it should be (assuming you're using an external microphone and have boomed and socked or muffed it). While you can crudely set audio levels, you'll probably want to run an external mic through something like the Beachtek mixer to get optimal results.
As I noted earlier, the D7000's autofocus during video is sub-par: expect to be pulling focus manually (if you don't know that term, then I suspect that you'll find the D7000 is a fine video camera /sarcasm).
Video is a step forward for Nikon, but still not at the level a dedicated videographer wants.
A Note About Quality Control
As some of you know, I'm proactive in following up such complaints. While I no longer do it publicly on forums, I still send messages and emails to people with complaints in an attempt to see if they're really valid. Many of those "doesn't focus" problems actually turn out to be things like not understanding that Single Point AF does not seem to track motion quite the same as Dynamic Area. Likewise, the "overexposes" complaints often end up being not liking the change in gamma that the Standard Picture Control imparts coupled with things like Single Point AF (there it is again) telling the matrix meter to put more emphasis on the thing that's being focused on. Unless that's skin tone or brighter, the camera will "overexpose." It's one of the reasons why my books are so well received: I point out these things and what you should do about them.
But the real issue here is that we've got a lot of people buying into very sophisticated equipment--and the D7000 is extremely sophisticated--and expecting magic "just set to auto and shoot" results. If you're thinking about the D7000 as your camera, expect to spend some time learning it, mastering it, and setting it appropriately for what you're trying to do at any time.
Meanwhile, all those folk who returned D7000's for reasons that turned out to be not valid are probably fueling a "refurbished" market coming sometime in early 2011 ;~).
Should You Get a D7000?
Bottom line: the D7000 an excellent consumer camera. Nikon has upped the ante with the D7000 and produced a real winner. It's very hard not to like the camera, and it takes images that ought to be good enough for just about anyone.
FX does have its pluses: that's where the pro bodies live, so if you really need a pro body, that's what you'll buy (at least for now ;~). The viewfinder of an FX camera seems brighter. You've got more wide angle lens choices and flexibility with FX. You can isolate depth of field more with FX. And...oh, there is no and anymore! On the minus side, FX bodies are big and heavy. They don't have the pixel density that wildlife shooters seek. They cost more ;~).
So we've come to a time when you need to make a decision: am I DX or FX? DX is still missing the pro body (but I'm predicting that the D400 will be it) and some good wide angle primes, while FX is still missing the small, light, simple body (but I'm guessing that gap, too, will be eventually filled, though probably not at the price point all of you are thinking it'll be at).
I'm going to put on my Galen guessing hat: he'd be DX. And he'd have scooped up the D7000. I'm not 100% sure what he would have stuck on front (hey Nikon, are you listening? DX WIDE ANGLE PRIMES PLEASE; yes, I'm shouting). And that's what I'll be doing on my upcoming Patagonia workshop: emulating what I think Galen would be doing. Because there's no way I'm hauling my heavy D3x up the those steep and long trails.