initial review: 11/16/06
fixes: 7/7/07
added FUNC button comment: 1/29/08


  Nikon D80 Review

Is 10mp really better than 6mp? And can the D80 hold its own against three strong competitors?

  Add a comment or send Thom feedback on this article.

[Disclosure: Normally, all my reviews are based upon product that I purchase from local stores and my review is written using that sample and without consulting the manufacturer. This review is based upon use and examination of an early D80 I purchased, and when I discovered the amp noise issue (see Performance section), use and examination of a much higher serial number unit lent to me by NikonUSA. NikonUSA offered to repair my D80 or to let me keep the loaner and send back the body I bought; I accepted the latter offer, mostly because of timing issues. During both my testing process and the testing that I had other D80 users perform, I was in contact with NikonUSA about the problem. When I had a statistically significant sample size to draw conclusions on, I sent NikonUSA a summary of the results I found. At present, Nikon has no official stance on the issue I found. As regular site visitors know, I try not to put myself in positions where there could be a real or even perceived conflict of interest. When I do, as I have in this case by corresponding with NikonUSA prior to the review and accepting an offer of a replacement body, I make that known so that you can form your own opinion about my independence and credibility or my lack thereof. This review took longer to appear than normal because I wanted to be sure that what I said applied not just to the original body I bought, and that I could characterize what I found clearly and with confidence in the findings.]

Well, here we go again. As I write this, we have four enthusiast 10mp cameras on the market, with Canon, Pentax, and Sony being the competition this time around. Last time we had this type of burst of consumer DSLR product the rollout from the manufacturers happened slower, was at 6mp, and KonicaMinolta was the fourth entrant (now acquired by Sony).

While it seems like deja vu all over again, this time the stakes are higher. Canon wasn't actually first to the bar, and neither Canon nor Nikon gets any real time to build up installed base without competitors. Here's a crude generalization of the way things look at first pass:

  • Canon 400D (XTi): competes on price and brand loyalty
  • Pentax K10D: competes on price and deeper feature set
  • Sony Alpha 100: competes on feature set and brand name
  • Nikon: ???

While Nikon has a brand following, it isn't as large as Canon's nor is the brand name itself as visible as Sony's, so Nikon can't really win the branding game. Nor does Nikon's latest camera exactly win the feature set war or price war. So at first glance, the D80 seems overpriced and bringing a knife to a gun fight. Seems like bad news, doesn't it? Well, stick around, a product is more than the sum of its parts, as it turns out.


The D80 with the Nikkor 18-135mm lens mounted.

The D80 was a bit of a surprise to Nikon users, I think. A D70 replacement, while due, wasn't the next Nikon DSLR everyone expected to see. But that's what we got. The camera shipped in early September in most parts of the world, and my review is based upon one of those first bodies and a later sample provided by NikonUSA (see red section at start of review).

Along with the D80 we get another new "kit" lens, the 18-135mm f/4-5.6G AF-S ED DX, and a new vertical grip, the MB-D80. As you'll soon see, those are about the only "new" things we got: the D80 is an interesting amalgamation of existing Nikon parts and features. That's a good thing, actually, as Nikon seems to have chosen well when it consulted its part bin.

Highly Recommended

features
performance
build
value

Nothing is knocked out of the park, but nothing holds this camera back much, either.

The definitive D80 eBook is here! Click Here to find out more.



 

The Basics

The D80 is a refresh of the D50 body design with some transplants from the D200. Specifically, the body size, shape, and build is very much that of the D50, right down to the door for a Secure Digital card on the right side. As such, the D80 is slightly smaller and lighter than the D70s and definitely lighter and smaller than the D200. Yet Nikon has chosen to use the D200's viewfinder in the D80, and many of the controls that are missing on the D50 but present on the D200 have returned (the FUNC button, for example). We'll get back to what all that means in a bit.

Here are the D50, D80, and D200 side by side for comparison:


The sensor in the D80 is an ICX493AQA CCD made by Sony. The same sensor is used in the Sony Alpha 100 and Pentax K10D. The 10.2 effective megapixels mean 3872 x 2592 pixel images, enough to produce straight-from-camera prints up to about 11x16" without resizing. The base ISO of the CCD is 100, with third-stop increments up through ISO 1600. You can also boost ISO one more stop, up to an effective ISO 3200. While it has the same number of photosites and much the same technology as the D200 sensor, the sensor used in the D80 is a simpler two-channel device--it is not an exact match.

Noise in the D80 sensor stems mainly from three design elements: the APS frame size, the smaller overall size of the photosites as compared to the 6mp sensors, and the lack of in-sensor NR electronics. That said, Nikon has integrated a number of things into the D80 to address the potential for increased noise over the previous consumer cameras. For example, once again data is kept in 12-bit space right up until the final compression to an 8-bit JPEG. Second, the internal digitization engine has a noise reduction component in it (well, actually two different components, one for long exposure noise and another for high ISO noise). When we get to the results section of the review, we'll talk about whether or not Nikon was successful, but note that multiple techniques are once again being employed here rather than a single one.

The D80 is a poor IR or UV camera. Each successive Nikon round of DSLRs seem to have more and more IR and UV filtering, and the D80/D200 generation are no exception. If you want to shoot that type of image, you're either going to have to hold onto those old D1's and D100's, or have a newer DSLR modified to remove that filtration.

Outside the sensor, the D80 feature set is also state-of-the-art. With surprisingly few exceptions, essentially it's the D200 feature set. Really. So much so, that here's a fairly complete list of the differences between a D80 and D200 (for a full comparison, see here):

  D80 D200
Sensor 10.2mp two-channel output 10.2mp four-channel output
Drive 3 fps 5 fps
Grip Optional MB-D80 Optional MB-D200
Autofocus 11 sensor CAM1000, no group AF, only central wide area 11 sensor CAM1000, group AF options, wide area
Card Secure Digital Compact Flash
Meter 420-pixel matrix meter; Scene exposure modes, AI and AI-S lenses mount, but don't meter 1005-pixel matrix meter; no Scene exposure modes, AI and AI-S mount and meter
Other Compressed NEF only, 80ms lag, 1/4000 top shutter speed, 1/200 flash sync, some additional bracketing limitations, D70-type wired remote, Retouch menu Compressed or full NEF, 50ms lag, 1/8000 top shutter speed, 1/250 flash sync, M-UP, 10-pin wired remote, GPS support, interval shooting

Shocked? You should be. Yes, this means that the D80 doesn't give up a lot when compared to the D200. Indeed, most of the differences boil down to modest performance differences (3fps versus 5fps, AF selections, shutter limitations, etc.).

A better way to look at things is to compare the D80 to the camera it replaces, the D70s:

  D70s D80
Sensor 6mp 10.2mp
Buffer 14 JPEG, 4 RAW 23 JPEG, 6 RAW
Grip No vertical grip option Optional MB-D80
Autofocus 5 sensor CAM900 11 sensor CAM1000
Card Compact Flash Secure Digital
Viewfinder Pentamirror, .75x magnification Pentaprism, .94 magnification
Meter 1005-pixel matrix meter 420-pixel matrix meter
Other 106ms lag, 1/8000 top shutter speed, 1/500 flash sync, FULL speed USB (slower), single group remote flash control 80ms lag, 1/4000 top shutter speed, 1/200 flash sync, mirror delay, FUNC button, HIGH speed USB (faster), multiple exposure, retouch menu, B&W shooting modes, noise reduction, 3-group remote flash control, TTL FP support, mirror delay, more

Most of the changes are improvements, even where they don't at first appear to be. The D80's 420-pixel metering system is revised from the one in the D70 series, faster to calculate and has more pattern matching. The 1/200 flash sync speed seems like a step backward, but the D80 doesn't have the D70's electronic shutter (which can cause blooming), instead adding TTL FP sync at any shutter speed with the current Nikon Speedlights.

The big changes, though, are the viewfinder and sensor. The D80's viewfinder is bigger and brighter than the D70's, and the sensor has nearly 30% more resolution on any axis. Things get better than that, though, when you look through the details.

The D80 adds intelligent battery life tracking ala the D200 via the EN-EL3e. We get a FUNC button and the ability to see the ISO value in the viewfinder. We now have a mirror delay option to limit vibration in telephoto and macro shots. A vertical grip option is available, complete with full controls and the ability to use two batteries at once or AA batteries in a pinch. Overall, as you explore the D70->D80 changes, the higher model number correctly suggests a wide number of additional features that D70 series users would envy.

The US$999 price tag for the body puts the D80 a bit above the Canon, Pentax, and even Sony competitors. So perhaps a look at what those cameras bring in addition to the basic feature sets and 10mp is also warranted. The list of major items seems at first to be daunting:

  • Canon: antidust mechanism
  • Pentax: antidust mechanism, antishake sensor, weather sealing
  • Sony: antidust mechanism, antishake sensor

Antidust mechanisms, while useful, aren't perfect. You'll eventually have to do a direct sensor cleaning on all three of the competitors, just not as soon as you might with the Nikon D80, all else equal. For what it's worth, Nikon has put a new anti-static coating on the filtration pack in front of the sensor, though for some reason they don't seem to mention that in their marketing (hey, Nikon, how many times do I have to say you need to improve your marketing?). I don't see sensor cleaning as a big deal in the first place, and most of those frightened into purchasing a camera with antidust mechanisms aren't likely to be the type that shoot in ways where they'll even see the dust (most Program exposure modes don't use small apertures, electing to keep the shutter speeds high to prevent camera shake and subject motion). We live in an age where fear is used to market products (and politicians), and dust is a fear issue rather than real issue, in my opinion. When you encounter it, you clean. But you should be cleaning your DSLR anyway, so adding another rather simple step to the regular maintenance list doesn't seem daunting to me. It shouldn't to you, either.

Antishake is more useful. There will be situations where having a image stabilization system allows you to shoot handheld where you might otherwise get unacceptable results. But image stabilization is situational in use: it doesn't stop subject motion, for instance. Pentax and Sony are using sensor-based stabilization systems. The advantage is that they add stabilization to all your lenses. Their primary disadvantage is that they add mechanical complexity at the point of highest cost of repair in your camera. Nikon and Canon have chosen to use lens-based stabilization systems to date. Lens-based systems work a bit better with longer lenses than do sensor-based stabilization systems and they allow you to see the stabilized image in the viewfinder. Unfortunately, lens stabilization costs more since you end up purchasing the stabilization with each lens you purchase rather than once with the camera. Unfortunately, Nikon's kit lens offering with the D80 is not one of the lenses with VR built in (Nikon's stabilization system). You'd have to purchase a more expensive lens, such as the 18-200mm Nikkor to get this feature. So Pentax and Sony have a feature advantage here.

But to my thinking, the most impressive extra feature of any of the competitors is the Pentax's weather sealing. Not only does this help keep dust transfer into the camera down, but it allows you to basically ignore things like mist and mild rain. It helps keep condensation issues in check (though not completely--the lens will still have issues in rapid humidity changes). In general, it provides an extra layer of protection against the elements for the somewhat delicate electronics inside. I find that more valuable than antidust or antishake mechanisms.

Of the three 10mp competitors to the Nikon D80, the Pentax K10D seems the most impressive and complete in terms of feature set and build, and the only one that I think clearly warrants a look to a potential D80-purchaser who isn't already committed to lenses (the four competitive cameras all use different lens mounts, and thus, different lenses). I know I'll get grief from Canon and Sony proponents on that statement, so perhaps I should elaborate. The Canon Digital Rebel XTi is a mildly reworked Canon Rebel XT (350D in other parts of the world). It didn't get as much feature and functional changes as I think that body needs, and the viewfinder is dim and small compared to the Nikon D80. In terms of actual shooting, it's a competent camera and finally has a respectable buffer, but I think it appeals more to the all-automatic crowd than those that really want the full flexibility of a DSLR. I still don't like the Canon Custom Setting design--Nikon gets this function right (though they keep renumbering options). The XTi has a slightly awkward grip and hand position, in my opinion. I do like the rear LCD display of the now-missing top LCD information; very nicely done. The card door problem is still unfixed, which makes me wonder just how much time Canon is spending on engineering user-proof solutions versus easily marketed changes. Side by side with a D80, the Canon XTi just doesn't quite seem to accommodate a serious user quite as well (check out the differences in how Auto ISO is implemented); it misses the target a bit in flexibility, speed, and directness. And the image quality is arguably slightly worse than the Nikon at high ISO values (some might argue otherwise, but I find the Canon blocks up the shadow areas and muddies colors compared to the Nikon). The Sony Alpha 100 seems to have the worst build of the bunch, which is surprising since it isn't the lowest priced. But I think the Sony's biggest liability simply is that the image quality starts to suffer more in comparison to the others as you bump up the ISO. The difference is easily seen at ISO 1600, but it's there beginning as low as ISO 400.

Features aren't the only thing you evaluate a potential camera purchase on (otherwise this review would be almost done, wouldn't it?). To me, the bigger issues are handling (ergonomics) and performance (image quality). So don't count the D80 out against the competitors just yet...

The bottom line for a dedicated Nikon user is this: the D80 adds nicely to the D70 series feature set, yet doesn't exactly sacrifice a lot from the D200 feature set.

The bottom line for the first-time DSLR purchaser is this: Canon, Nikon, Pentax, and Sony have all produced mid-range consumer DSLRs that have 10mp and a raft of interesting features. You're not going to go far wrong purchasing any of them. But you might want to keep reading this review to find out how the D80 does on the bigger issues of handling and performance.

Handling

I've written a lot about the handling of Nikon DSLRs in my other reviews, and for the most part, things haven't changed much with the D80. Rather than repeat myself too much, I suggest you look at my D200 review and read the Handling section there. I'll confine most of my comments here to differences between the D80 and D200.

While the D80 is a smaller camera than the D200 (and D70s for that matter), Nikon has once again gotten the "grip" right. The same one-handed (right hand unfortunately for you lefties) hold still feels comfortable and secure, and reaching the key buttons while holding the camera is natural (all except for the Depth of Field Preview button, which has migrated to bottom of the lens mount, an arguably okay position, but a position some will think is a long reach). For such a small camera, it doesn't necessarily feel small (and you could always opt for the optional MB-D80 if you want to bulk it up).

Nikon's button-and-dial interface remains largely intact throughout, though the location of buttons moves around a bit. For example, on the D80, the bracketing button is on the front of the camera near the lens release button (on most other Nikon bodies, any bracketing button is either on the top plate or top rear of the camera). Fortunately, the Metering Method and Exposure Compensation buttons are where we expect them.

What might bother a few (it doesn't bother me) are the Frame Advance and AF buttons to the right of the Top LCD. Most people read Nikon's documentation and think they must be pushed over and over to get to the setting you want. Well, you can do that, or you can simply hold the button down and use the Rear Command dial to change the setting, as usual. The one problem with the "push for next setting" option is that you need to make sure that you don't accidentally hit these buttons--you can change your camera's setting without knowing if you're not careful. Fortunately, these buttons are recessed, but still, you could bump your camera against something and find that now you've got the Self Timer set or your AF settings have been changed.

Overall, the AF options have been severely reduced, and if there's anything in the camera's handling that feels compromised, this is where you'll find it. There's no group autofocus functions, and we've only got AF-A, AF-S, and AF-C for our primary, easily set focus setting choices. Single Area, Dynamic Area, and Auto Area (really a renamed variant of Closest Subject Priority) are buried in the Custom Settings, so changing the character of the AF system is not something you can do without getting into the menus (you can, however, assign it to the FUNC button). We only get a center wide area option, so the focus system really is always an 11-segment one. These are all logical subsets of the D200 abilities, but I wish the controls were a little more thought through (we could have had the Front Command dial control AF Area and the Rear Command dial control AF Mode using the dedicated AF button, for example).

Long-term bythom site visitors know I'm not a fan of the optional vertical grips. The MB-D80 is no exception. Indeed, with the fewer buttons (no AF-ON button, for example), you get less flexibility in setting up way things work. Because of the long reach from the vertical grip to the AF Direction pad, you'll almost certainly want to set the AE-L button to also let you control the AF Area via the Front Command dial. I find this somewhat distracting, and it does place a limit on what you can assign to the button (you can't assign FV Lock to it, for example). On the plus side, the vertical grip does have the full set of controls (both dials and the AE-L button in addition to the lockable shutter release) and has the same multiple battery options as the MB-D200 did (six AA or two EN-EL3e).

One handling aspect that D70 users will welcome is that the ISO value now displayed on demand in the viewfinder (via the FUNC button). What isn't welcome is that if you press the ISO button to set the value, the value isn't displayed in the viewfinder. This seems like a silly design, to me. Why assign it to the FUNC button, taking up a valuable ability when we have an ISO button on the camera? Sure, the ISO button doesn't control ISO when an image or menus are displayed on the Color LCD, but frankly, just who is looking through the viewfinder trying to figure out what ISO you're at when the Color LCD is active? A quick half press of the shutter release would turn the Color LCD off and get you to setting ISO anyway. In other words, I like that Nikon thought to give us the ability to see the ISO, I just don't like where they put it.

Pretty much all the D200 image quality options are present (including the full and quite large range of white balance options), but we've got one very significant new one that I like a lot: the ability to set "filtration" for black and white image shooting. That's right, you can tell the camera to "add" a red, yellow, orange, green, or blue filter when shooting in-camera B&W. It works, and it's nice to finally be able to able to take a few shots and see how filtration changes the overall tonality of your B&W images without having to physically add filters to the camera. The one D200 image quality option that isn't present is the separation of Color Space and Color Mode. The D80, like the other consumer cameras, gives you only Ia, II, and IIIa as combined Color Space/Color Mode options (that's two sRGB options and one AdobeRGB option for the uninitiated).

Unlike the D200, Custom Settings are back to a single scrolling list of 33 options (Reset is the 33rd) instead of being "batched" by function. Moreover, the default of the camera is "only display the first few"--you have to use Setup to display the full set of menu options before you see all your options. I continue to dislike this "feature" of having a simplified menu system, though I do like Nikon's latest tweak: MyMenu. MyMenu allows you to define which menu options appear. Thus, for all those set-once-and-forget options, you can banish them from the menu system after setting them and not have to scroll through them time and time again.

Another new menu option is the Retouch menu, which replaces the D200's Recent Settings menu (Nikon giveth, Nikon taketh...). Retouch allows you to do quite a few editing types of things: D-lighting, Red-eye correction, Trim, Monochrome conversion, Filter effects, Small picture, and Image overlay. Most of these are things I wouldn't want to do in camera, as the color LCD just isn't good enough to evaluate D-lighting and Image overlay effects, for example. Trim is nice, as it allows you to do after-the-fact crops. Filter effects/Color balance is probably the most interesting, as it gives you a two-dimensional CIE color space to shift as you see fit. This is much more flexible than Hue adjustments. In all cases, the Retouch options create new versions of your photo (your original remains intact).

While we're talking about images, let's talk about image review. Nikon has changed the way magnifying an image works. We now have zoom in and zoom out buttons, essentially. This simplification works quite nicely. Press zoom in and you zoom in. Press zoom out and you zoom out, and once you're at Fit Screen view you then move to four thumbnails and then nine thumbnails with subsequent presses (zoom in gets you back to single image view). Much simpler to explain and understand, though not quite as fast as the old system. I think it's a perfectly appropriate design change for the intended audience of this camera.

The new D200-derived viewfinder is much better than the D50/D70 viewfinders. It's bigger and brighter, uses a better eyecup, yet still supports all the consumer goodies like overlaid grid lines.The downside is that eyeglass wearers are going to have a more difficult time simultaneously seeing both the frame area and the data area underneath. I can, but I also have very thin lenses in my glasses and a thin frame. I suspect that people with large prescription corrections and thick frames won't be able to see everything in the viewfinder at once. One other nice touch on the viewfinder: the diopter adjustment is a small knurled wheel just to the side of the viewfinder--it's easier to adjust the position, keep it, and even see what that position is in than it was with the D50/D70/D100 diopter adjustment.

Nits are few. I still don't like the Secure Digital door mechanism. In extreme handling it has a tendency to pop open on me, just like on my D50. There's not enough viewfinder information for my taste (where's the metering method icon?). The Image Review On option is inflexible: images appear for only four seconds after being taken, and you can't change that value.

One final thing I noticed: on the D70 Nikon advertised it as capable of "shooting until you fill up the card." The D80 has an interesting limit: put the camera in continuous frame advance and hold down the shutter release. Depending upon the speed of your card, you'll get some number of frames at the maximum frame rate, then the camera slows a bit to whatever your card is capable of sustaining. But in no case will it go beyond 100 frames that way. When you hit 100 consecutive frames, you need to lift your finger off the release and press again to continue, even though it appears the buffer has room.

 

top | home

Cleaning the Sensor I've posted an article on how to clean sensors.

 

 

Performance

Battery Life
Battery performance is either quite good or fair, depending upon how you shoot. If you're a JPEG shooter, you'll get very good battery performance--I easily got more than 800 shots per charge on all my batteries every time I shot JPEG only. If you shoot NEF or NEF+JPEG, you won't be as happy. My range of shots per charge in actual use is significantly lower. Granted, I've got a VR lens mounted a lot of the time, but I suspect so will many D80 users since the 18-200mm seems like such a an excellent match for the body for general shooting. Still, only having a few hundred shots per charge means that I have to carry at least one extra EN-EL3e battery with me for a day's worth of shooting. If I were shooting events, I might need many more batteries to get through the day. Good thing they're small and light.

Writing to Card
The Secure Digital write performance is generally very good to excellent. Indeed, the same cards in my D80 perform much faster than they do in my D50, so it's clear that Nikon has goosed up the Secure Digital write performance. You can achieve about 9MBs per second write speed with the D80 if you choose the correct card and are shooting JPEG (NEF writing performance tends to be about 10% slower than JPEG on the D80 as far as I can tell). Rob Galbraith has tabulated card write performance on his site, and my results are again pretty much in line with his. Even with some generic cards I was getting 5MBs per second performance, which is pretty decent. Remember, you've got more data to store with a 10mp camera, so if you're used to shooting with a 6mp camera you'll probably want to invest in larger capacity cards. A 2GB card is relatively inexpensive these days and just about right for the average D80 user to standardize on.

Autofocus System
Autofocus performance is very good, but not excellent. The number in the part name (CAM1000 versus the old CAM900 in previous Nikon consumer DSLRs) indicates the number of overall points that are sampled for focus information, so it should be obvious that the D80 should be better than a D70 (CAM900) but worse than a D2 series camera (CAM2000). Indeed, it is just that, though I'd put the performance more towards the D2 series end than the part number would indicate. There's definitely a significant improvement over the D70 in terms of autofocus speed in most situations.

Surprisingly, the D200, which uses the same AF system, can be configured to perform somewhat better than the D80. This has to do with the option sets, such as Group AF and most significantly, the ability to fully control release priority versus focus priority on the D200 (but not on the D80). The D80 has a simplified set of AF controls that don't fully allow you to get to the underlying performance of the AF sensor.

Like the D50/D70/D100, the D80 works best in low light if you can keep the central AF sensor the active one. The outlying sensors aren't quite as bad as the ones in the earlier cameras, but there is a decided drop-off in performance as you move outwards from center in low light.

Overall, I'll stick by my initial statement: very good, but not excellent. The demotion comes primarily due to the fact that AF on the D80 works better the closer to the center of the frame you're trying to focus and the fact that you don't have as much finite control over the way the AF system works. This actually may not be a problem for most amateur photographers, as they either center their subjects or focus-and-reframe. But for the more sophisticated user AF performance degrades as you move outward, and compared to the focus-almost-anywhere aspect of the D2 series, the performance drop is obvious.

Color
Color integrity is very good in Imatest's color checks, even with the slightly distorted sRGB color spaces that Nikon uses (Ia and IIIa). Saturation is a little higher than I expected, resulting in punchy colors for a Nikon body (AdobeRGB gets you closer to accurate). Here's the color charts at ISO 100 and 1600:

Note how close they are. Amazingly close. Saturation drops from 118% to 116%, and the color error stays virtually the same (and quite low overall). Note that the blues and the deepest red drift away from accurate in this sample (1a color space, the default), but all colors move just a little bit away from neutral. If you're using an inkjet printer, you may find it difficult to get the blues the D80 can produce onto the paper, though. Still, the overall performance is extraordinarily good and consistent at the ISO values you're most likely to use (100 through 1600). In particular, the ISO 1600 performance is nearly jaw-dropping--we're so used to losing color saturation and fidelity at higher ISO values that the D80's excellent ISO 1600 performance is refreshing.

White balance isn't quite as good. My normal test situation uses mixed lighting sources as sort of a worst case scenario. No camera I've tested manages to get perfectly flat white balance in this situation, even using Custom white balance settings. The D80 is no exception, though it seems to miss by a bit slight bit more than my D200 does. In no case was any deviation more than 3 mired, which means the deviation is relatively low, but in almost every case the brightest values tended to miss on the warm side while the darkest values tended to miss on the cool side. I'd characterize both the Auto and Custom white balance abilities as good, but not perfect. As you get more unbalanced lighting (especially on the red side), they do less well than they do at mid-temperature (daylight) ranges.

Overall, color is very saturated and quite good even in the Normal and less saturated image optimizations, a bit over the top if you use any of the enhanced optimizations.

Noise
Let's start with amp noise, because it's going to be the most controversial part of this section. On a 15-minute exposure, amp noise is excessive. You'll see it in both top corners of the screen (image from loaner D80 from Nikon):

Thus, the D80 is not a suitable camera for astronomical or long night exposure work. If that were my only conclusion, I wouldn't have said much other than that last sentence. Unfortunately, it's not that simple. Shortly after the D80 was in customer hands I got a couple of images forwarded to me that showed clear amp noise in what appeared to be something that wasn't a long night exposure, and in a location other than the ones I just showed. Upon further testing with long exposures, it was clear that my original D80 had amp noise in three places (15-minute exposure on my purchased D80):

Moreover, I noticed that I was getting strange full-image noise testing results from my own D80. Unlike other sites, I not only test the usual small patches on a properly exposed MacBeth ColorChecker chart for noise, but I also do full image tests of a gray card (it's part of my dynamic range testing regimen).

At ISO 1600, amp noise started showing up at shutter speeds I wouldn't it expect it to. Specifically, on my original camera, that happened somewhere around the 1/15 second mark. Any shutter speed longer than that would have some amp noise buried in the top center of the image. Generally, this wasn't visible in images until I got to at least a 30 second exposure, but if I did any type of post production that pulled up pixel values in that area of the image, I sometimes started to see the problem in images even at 1/15. The noise was there even if High ISO Noise Reduction was turned On (the default at ISO 1600). Here's an example, shot at 1/10 at ISO 1600 and then using Auto Levels to make the underlying noise obvious:

Now very curious, I asked a number of D80 users to run a series of tests to see if other cameras had the same problem or if mine was simply defective and needed repair. Within the first 10 results coming back, I saw a clear pattern: virtually all of the other D80s being tested had some variant of this issue, though the point at which it appears and how badly it appears seemed to vary a bit. In all cases, the underlying amp noise was not visible when examining the image normally. But if you run Auto Levels you would almost certainly see a very clear underlying amp noise pattern. On some cameras it is first visible at 1/60, on some it takes a longer exposure to begin to see the amp noise. It seems clear to me that some electronic component near the bottom of the sensor (remember, images are flipped at the sensor) is running hot or mounted too close to the sensor.

Next, I contacted NikonUSA about what I had found (and test results from others kept coming in, giving me a better handle on the problem). They were concerned and said that they didn't see the same thing in bodies they tested that they had on hand. I asked them to lend me such a new body so that I could do a comparison. Indeed, the body I received from Nikon clearly performs differently in respect to that third amp noise location. By this time, I also had enough tests from others to be able to statistically predict to the NikonUSA D80 population, and I noticed something else: the results changed somewhere around the 3050000 serial number mark (first generation US bodies have serial numbers starting with 30; the success of the body has forced Nikon to now also use the prefix of 31 for US bodies). Bodies before that had a clear third amp noise location, bodies after that didn't. Moreover, I was better able to quantify the actual problem. Here's a summary of the more elaborate set of results I sent to Nikon:

  • Amp noise was always visible with a proper exposure on a maximum of 10% of the samples, and it really only started showing up at 1/8 second at ISO 1600 in those cases.
  • A large majority of the samples showed amp noise at all tested shutter speeds (1/60 to 8 seconds) at ISO 1600 and at about 2 to 3 stops under the exposure (i.e. it takes significant underexposure and/or post processing to see).

So let me state what I think I know: on early D80 bodies at ISO 1600 and above, you'll start to accumulate a little amp noise along the top central edge of the (horizontal) image at longer shutter speeds. Under a normal exposure, you're not likely to see that in your pictures. If you do see it with normal, well-exposed shots, I'd say that your body should be looked at by Nikon. But even if you know what to look for and expose in a way to try to make it more visible (e.g. put black in that area and underexpose) you still may not see it. But there's noise living in that area on those early D80s, and if you do any major post processing that dramatically changes the pixel values in that area, you may start to see it. The question at hand is whether this is a camera flaw. I think not, but I certainly don't like the fact that my purchased camera had this trait embedded in it (which is why I took Nikon up on the offer to swap cameras). I don't know of another DSLR that exhibits amp noise quite as early in the exposure spectrum as these early D80s (all DSLRs exhibit amp noise at some extreme exposure). As many of you know, I shoot predominately outdoors and in wild areas (backcountry). I often shoot pre-dawn and post-sunset, so I often am using very long exposures (though not usually at ISO 1600). What I see in the amp noise tests would keep me from using the D80 in such situations. Even if the noise isn't "visible," I have Nikon cameras that do so much better in very long exposures that I can be confident that I'll never post process an image and suddenly discover that I have purple grunge in a portion of the frame. Thus, I characterize what I found as a clear limitation, not necessarily a flaw. Those of you with later D80 bodies don't have to worry about that limitation (though I should still note that amp noise shows up in the corners in very long exposures; in other words, there's just a different limitation point--a better one--in the later D80 bodies).

My advice: don't over obsess about this issue, but be aware that your camera has limits at high ISO values. That's exactly how I'd word it: the D80 has a limit that other Nikon DSLRs (and the D80 10mp competitors) don't. If the problem is clearly visible on your sample on proper exposures, have Nikon look at your camera. That's true of any problem, though: if you're not getting correct results with proper technique and exposure, then the camera should be looked at for problems.

For general noise I have better news to report. I'm getting more and more leery of the "numbers" approach. All the noise tests you see on the net or in magazines are done basically the same way, using 100x100 or 200x200 patches of a test chart and measuring things like Standard Deviation (or in the case of Imatest, more sophisticated statistical analysis). Those numbers don't usually tell you a lot, but if you really want to know, Imatest reports the D80 being a tiny bit better than the D200 at ISO 100, and clearly a little better than the D200 at ISO 1600.

But numbers don't really tell you whether the images look free from artifact. Noise is actually more a subjective problem than an objective one. Some cameras have had low numbers but get a little watercolor-type effect on edges, others have lots of color noise that is easily seen and distracting but little luminance noise, still others have clear patterns to their noise (as opposed to a randomness) that makes it obvious that the image comes from a digital camera. I'm going to give you two small 100% samples that appear also in my eBook to illustrate my point (and how well the D80 handles noise).


Remember, you're looking at 100% view here, so these are very small crops from the full image. The things I want to point out are these:

  • Overall, the granular effect that most people associate with noise isn't particularly visible (it certainly isn't when printed). But if you want to look for it, look at the texture of the barrel the parrot is sitting on. At ISO 100, the barrel has the bright glossy sheen of the plastic it is (Lego sets, folks). At ISO 1600, the barrel has taken on a slight matte type of finish. This is visually distinguishable at ISO 1600 at pixel peeping views, but doesn't really show up in prints.
  • The blue in the parrot has lost a little of its royal blue sizzle, though it is still remarkably saturated in color (the XTi I tested took the blue down to a darker color, more like the D80 at ISO 3200). The really punchy colors will lose just a bit of their zing due to the underlying noise structure, but not as much as you see in most DSLRs.
  • Edges are certainly crisper at ISO 100 than they are at ISO 1600, but they're still edges at ISO 1600. There's none of the watercolor fuzziness you sometimes see with noise reduction systems, and none of the edge destroying tendencies of the D2xs at high ISO values.

In short, the D80 is producing very consistent results across the spectrum of ISO values that users normally would set (100 to 1600). Indeed, I can't really find any tangible difference from 100 to 400, and at 800 the results are still so good you have to look hard to see the small differences. The performance is so good that I can recommend some users leave Auto ISO set with a max of 400 or 800, depending upon your tolerance to low level changes.

What's that in the back? What about ISO 3200 you ask? Okay, here's the same thing at ISO 3200:

Yep, here the tell-tale impacts of noise are clearly showing themselves. You can see grain effects across the picture, the parrot is changing color, the barrels are even more matted (and a bit of color noise is peeking through), and look at the parrot's back: edges are starting to have problems even with normal sharpening. Still, you're at pixel peeping maximum on a 10mp camera--this is still quite good performance. Usable in a pinch (and hint: it's pretty darned nice with the B&W shooting mode; put another way, better than Tri-X pushed a stop).

Exposure
Good news, bad news, folks. Let's start with the bad news, since there are plenty of contentious threads across the Nikon forums on the Internet talking about the issue: the matrix meter is more sensitive to what's going on in the central area.

Say what? That's bad news? Yes, it is for those of you who've shot with the D70, D100, D200, D1 series, D2 series, or any of the Nikon film bodies. That's because all those other bodies are remarkably good at not being coerced into changing exposure based upon smallish non-middle gray things in the central area of the frame. The D50 was prone to changing exposure in those situations, and the D80 is even more prone to changing exposure due to middle subject tonal value. If you're not following me, try the following: cut a diamond shape (like the AF sensor pattern) out of white paper, middle gray paper, and black paper. (I'm having you use a diamond shape so that you'll frame them the same.) Now place each of those against a middle gray background, carefully framing so that the diamond exactly fits the AF sensor pattern. Voila: the gray background is underexposed with the white diamond in the center, properly exposed with the gray diamond in the center, and overexposed with the black diamond in the center. It's that last one that's problematic, in my mind. Overexposure means blown highlights that can't be recovered, and the D80 is going to be prone to do just that with the matrix metering system any time you've got darker-than-middle-gray subjects in the center of the frame. Photo of the bride only? A little underexposed but easily savable. Photo of the groom only? Overexposed and not savable. Oops.

Other than that, Mrs. Lincoln liked the play. The center-weighted and spot meters are their usual excellent selves, and the matrix meter, while a bit different than those that came before it, is predictable once you know what it's doing.

Surprisingly, the step test (wedges of white to gray to black in 22 distinct steps) produced a result I hadn't seen on other Nikon bodies before: it sags in the middle. Put another way, the shadow detail seems to be sacrificed a bit for better mid- and high-range tonal ramps. This also explains the excellent noise results: Nikon isn't pushing the shadow details as aggressively on the D80 as it probably could. Antennas just went up: so does that mean reduced dynamic range? Not that I can tell. Shooting raw, which is the best way to get a real handle on the maximal dynamic range of a camera, I couldn't detect any real difference between the D80 and D200. There's at least seven usable stops, and probably more depending upon your tolerance to noise (remember, you can't really define dynamic range until you establish a noise floor; mine tends to be pretty conservatively stated, so I tend to produce lower dynamic range figures than others do). In my book, anything from 7 to 8 stops is fine, and the D80 lives in that range. Less dynamic range than that and you have problems getting good tonal ramps and shadow detail. More dynamic range than that and you'd better be really good at post processing, since you're going to have to play with contrast curves to get the range to fit well with most output devices.

Resolution
Which brings us finally to resolution. What do those 10 megapixels buy versus the old 6mp standard? The answer isn't quite as simple as it seems. First, note that in any given axis, you're only getting an increase of a little less than 30% more pixels with the new sensor. That's a significant improvement, but not a huge one. For the most part, you're not going to see the change unless you're printing big.

But the smaller photosite size also brings with it a new issue: diffraction limitations. Without going into the details, visible diffraction starts to show up somewhere between f/11 and f/16 on the D80, while it's about a stop lower than that on the D70s. If you were taught to "set f/22 and hyperfocal distance" for landscape photography, you'll want to relearn that it's "set f/13 and hyperfocal distance" if you want to maximize acuity and take full advantage of that resolution boost.

10mp also means getting better at shot discipline. Fortunately, the D80 has both a delayed shutter ability and both infrared and wired remote release options. Learn to use them (on a stable tripod, of course) if you want to maximize what the camera can produce. Moreover, poor lenses tend to show their imperfections more clearly on the D80 than they do on the D50 or D70s. I've got a couple of lenses where chromatic aberration didn't really show up when used on the D70s, but is clearly present when I use them on the D80.

Still, 10 is better than 6. At low ISO values with proper technique with images printed at the maximum size a desktop inkjet allows (usually 13x19") you should see more detail resolved than you did in the past. If you don't, suspect your technique first, your support second, and your lens third. It's not the camera.

Final Comments
I'll repeat what I said earlier: we now have four 10mp DSLRs in basically the same class, with similar feature sets. All of them perform well, and the differences between them in terms of image quality are almost non-existent at ISO 100. At higher ISO values, the image quality differences start to become more obvious (note that I haven't yet tested the K10D using my usual test suite, so I won't comment on its image quality). To my eye, the D80 is at the top of the heap with the Canon close behind (the Canon tends to sacrifice shadow areas and lose color saturation at the highest ISO values). The Sony falls clearly behind the others at high ISO values. But even with the differences, I'm not sure they're enough to base a buying decision on. If you're nitpicking pixels so much that you can see these differences, you probably should be shopping for the next better class of camera.

The more clear differences come in ergonomics, usability, and other performance. The D80 simply doesn't get anything wrong in any of those categories. It's a well-rounded design that does everything well, but nothing perfectly. The other 10mp cameras all seem to get one thing or another not quite right, so I'd say they do most everything well. But you really need to handle and test these cameras yourself to see if anything holds you back on any individual camera. My advice is to try as many of the models as you can and choose based upon your visceral reaction to the designs. You'll clearly prefer the way one or another is laid out and the controls managed. That's the one you should choose. Because, when all is said and done, these cameras are all going to get very similar results in the hands of the skilled. Thus, the key differentiator is whether or not you're going to feel comfortable "getting skilled" with that camera. How fast can you change controls? How much (and how clear is the) information you are getting from the camera? Has Thom Hogan written a book about it that really tells you how the camera works? (Okay, shameless self promotion, that was. What can I say, I woke up this morning with as much testosterone as Floyd Landis, apparently ;~) Does the feature set have the controls you'll want as you grow in shooting confidence?

I know this will cause some controversy, but here's my personal order of preference and the reasoning behind it:

  1. Pentax K10D. A clear bargain for the impressive feature set. (But I reserve judgment to change its place in this list if the image quality doesn't match or exceed the D80's.)
  2. Nikon D80. Slightly more responsive (very fast shutter lag, for instance) than the others, no glaring faults. Excellent image quality in all ways I test, and perfectly usable at ISO 1600.
  3. Canon Digital Rebel XTi (400D). The low-price leader, but it shows, I think. Just compare viewfinders with the above two. Canon no longer has the high ISO image quality advantage.
  4. Sony Alpha 100. Should be priced like the Canon, but the biggest liability is the higher ISO image quality, which doesn't match that of the other cameras.

But I repeat: you won't go far wrong with any of them.

Drawbacks

  • Wide isn't wide. If you do scenic work, as I do, you need DX lenses to restore your wide angle due to the 1.5x angle of view change. We've now got enough pixels that the flaws of wide angle lenses, particularly chromatic aberration, are more apparent. The 12-24mm is a decent mate with the D80, but you'll see that it has a bit of CA you weren't seeing with the 6mp bodies.
  • Noise at higher ISO values. While I'm perfectly happy with the ISO 100 to ISO 800 performance, some will want even better results. High ISO noise is definitely there, though JPEGs are mostly free from chroma noise.
  • White balance encryption. Despite being dealt with by the mini-SDK, it still is a lurking thorn (see my D2x review for more details).
  • Diffraction. Being diffraction limited at f/13 or so is a bit limiting.
  • Battery lack of life. NEF shooters are going to be disappointed. You'll be lucky if you can reach 500 images shooting NEF+JPEG, especially if you're using a VR lens.

Positives

  • Sufficiently Fast. For a consumer camera, the frame rate, mirror return, shutter lag, and autofocus are fast enough to feel snappy. Indeed, this is a clear asset compared to some of the competitor's 10mp cameras.
  • Flexible and Controllable. New controls, more options, and improved user control all around. Some of the stuff is modest (like ISO view in the viewfinder on demand), but it still adds up to making for a very flexible and controllable consumer DSLR.
  • Rugged enough. Decently built. While not specifically weatherproofed, like most Nikon bodies it repels the elements decently. And say what you want about "plastic" bodies, in practice I find nothing wrong with them--the D80 handles drops and bumps much better than you'd think.
  • Flash Dance. i-TTL works, and it restores multiple flash ability. Having a built-in flash that can do the Commander mode thing for two remote flash groups and settings is slick, and unequaled in the marketplace. FP sync, repeating flash, and the other goodies are just icing on the cake.
  • Excellent Image Quality. 10mp is a lot of data, and the D80 in some ways does it better than the D200 (at least for JPEGs). Acuity is very good with careful sharpening, color is excellent, and noise performance is excellent at low ISO values and more than acceptable at higher ISO values. With the right settings and discipline, this camera performs near the state-of-the-art.

 


 

bythom.com | Nikon | Gadgets | Writing | imho | Travel | Privacy statement | 2008 Thom Hogan. All rights reserved.