The cat's out of the bag, time to answer some questions...
First up, we have the Nikon D3. As I tried to hint at recently, it uses a frame size about that of 35mm film (it's an teensy bit lower in height, but not enough to worry about). This is Nikon's first "full frame" (Nikon calls it FX) camera, making them the third company to mass produce one (Canon and Kodak being the first two).
What's so great about FX? I suppose if anyone were still shooting film and digital side by side it would be nice to have lenses work the same, but realistically, I can't point to anyone who seriously does that (I'm told that James Nachtway does, though, so perhaps there are a few). Thus, the usual "advantages" that are stressed are two: (1) DOF works as it did with film, and you can therefore achieve more subject isolation than you can with an APS-sensor body (all previous Nikon digital bodies since the D1 are APS); and (2) at the same megapixel count, a bigger sensor means that it should have better noise handling than a smaller one. Moreover, sensors are moving targets and still getting incremental refinements, so a new larger sensor should absolutely be better at noise handling than an older smaller sensor.
But for me, the real joy is that my closet of lenses acquired from the film days now works pretty much the way they were designed. I'll give you one example: the 28mm PC lens isn't very usable on a small sensor body, as it ends up giving a field of view equivalent of about 42mm. Suddenly, it can be used for the purpose in which it was designed: correcting linear distortions when shooting large architectural subjects. And my 85mm f/1.4 becomes a perfect portrait lens again (on my D2xs it's a tad long, pushing me further from my subject to get the head-and-shoulders type of portraits I used to do with it on film bodies). As most of you know, I've been harping on Nikon for having a somewhat haphazard lens lineup for awhile, and FX at least makes all those lenses that were designed for a specific reason work to that reason again.
(Side comment: Nikon's lens lineup still needs a lot of work. Sure, we got the expected VR in the remaining exotic telephotos, but so many lenses have been discontinued with no direct replacement that the used market is going to be hot for any number of lenses for use with the D3.)
From a sensor standpoint, the D3 looks like a winner in most respects. 35FF frame? Check. Lower noise? Check. Fast image handling? Check. 14-bit RAW? Check. One of the bigger pieces of news is that Nikon's ASIC and imaging chain has been revamped completely (and now named: Expeed). This is the second big change in that respect (the D2x had the first one), but this time Nikon thinks that they've upped the ante over Canon's Digic. But we'll have to wait for cameras to see if the supposed better high ISO capabilities are just JPEG/TIFF processing or both sensor and processing improvements. In other words, that ISO 6400 Nikon is pointing to: is it mostly going to be useful in JPEG/TIFF, or are the NEF files shot that way pretty clean, too? (It doesn't count if Capture NX automatically applies the same processing--workflows dictate that we need clean NEF files for ACR/Lightroom and other converters, too.)
The jury is still out on some of the other changes, too. Most of us didn't think the CAM2000 autofocus system (used in the F6, D2h series, and D2x series) was broken. What we have with the new CAM3500 system is completely different than anything Nikon has used before, so we'll all be in for a bit of trial and error in mastering it. Moreover, the "concentration" of AF sensitivity has moved from a sparse, wide field to a dense, narrow field (at least in FX; in the high-speed DX crop the sensors pretty much fill the frame). I'm not 100% sure I like that, but the proof will be in the pictures.
Other things of note in the D3 offering are the live view mode (with a compact camera-like contrast-based autofocus instead of the usual DSLR phase-based autofocus), the ability to recognize DX lenses and crop to them, the ability to tweak the autofocus for individual lenses, plus a bigger TFT LCD (3") with a near quadrupling of pixels (now 920,000). The dual Compact Flash slots are nifty, too, allowing you to just roll over to the second card when the first is full, create an in-camera backup as you shoot, or separate NEF and JPEG files onto separate cards. You can even copy cards using the camera, should you desire.
A curious change is that the D3 and D300 share a wireless transmitter option (WT-4). That makes for fewer accessories for most of us who have multiple bodies, a nice touch.
Finally, a few things should be noted that tend to be overlooked with all the big news. First, no new battery or charger: your D2 series batteries will carry over fine. Ditto with most accessories (other than the WT-1/2). The 10-pin and PC sync connectors get the F6 treatment: a rubber fold over cover that stays with the camera when you're using the connectors (finally! only 4.3 million lost caps later). The new autofocus system even has a D2 emulator mode, where the 51 sensors are reduced to a pattern that looks like the old CAM2000 11-sensor one. Compatibility is good, and the D3 has an awful lot of it.
Another thing that isn't talked about much is that the D3 supports a 10mp 5:4 crop. Why is that important? Well, if you're shooting for traditional 8x10" and 11x14" prints previously you've had to crop every image (and couldn't see the crop area in the camera without getting a marked focus screen, such as the Katz Eye).
Overall, this looks like a very nice follow up for both D2h and D2x users. It gives the D2h user the megapixels they asked for, it gives the D2x user the noise handling they asked for. Some will like the FX decision, some won't (for those, there's always the D300, see next).
What's with the D300?
Well, that was the big surprise I hinted about long, long ago: that the D3 announcement would also come with a twist. The twist is that the D2xs just got better, smaller, faster, and significantly cheaper. That's the way I think of it. That's a pretty impressive feat, actually.
Using a variant of Sony's new CMOS sensor, Nikon has essentially melded the feature list of the D200 and D3 into one new D200-sized body. Most of the key differences have to do with the sensor:
- The D300 is DX (same as the D200), the D3 is FX.
- The D300 has sensor cleaning shake, the D3 doesn't.
- The D300 only reaches to ISO 3200, the D3 supports 6400 (both support HI1 and the D3 supports HI2, so basically the D300 has two less stops flexibility and perhaps slightly higher noise due to the smaller photosites).
- The D300 doesn't crop, the D3 supports 5:4 and DX crops.
I'll limit myself to comments that involve new features the D200 didn't have (essentially, all of the D200 key attributes carry over, including the viewfinder [though it is masked differently], unless they were replaced by a D3-type attribute):
- The D300 is faster. 6 fps normally, and 8 fps if you use the new MB-D10 vertical grip and battery pack. Caveat: if you record in 14-bit NEF form fps drops to 2.5.
- The D300 has a better autofocus system. Basically, the D3 system is brought over intact (51 point, scene recognition, many group modes).
- All the new Expeed functions (Active D-Lighting, for example) are brought over from the D3, giving you more image processing options.
- UDMA support. Faster writes to the latest state-of-the-art cards; indeed, buffer sizes are becoming less important as card speeds increase.
- The 3", 920k pixel TFT LCD. Finally, you can check focus and actually see something. Likewise, there is HDMI output for video.
- The MB-D10 allows use of both EN-EL3e and EN-EL4 batteries, meaning that someone carrying a D3 and a D300 could standardize on a battery. Finally!
Overall, take a D200, which is a very nice handling camera, and stuff some key new components into it and you have something that's as fast as a D2x, has as many pixels as a D2x, has a better focus system than a D2x, writes to cards faster than a D2x, has cleaner high ISO results than a D2x, and, oh yes, sells for US$3200 less (at list prices, at street prices the difference is hundreds of dollars less). What's not to like about that?
What about Pixel Density?
Actually, the D300, like the D2x, is right up there as having the highest pixel density currently available in a DSLR. Why is pixel density important? Well, if you're a wildlife photographer (and some types of sports photographers), you're usually constrained in getting "close enough" to your subject. That means that the quality of your results is limited to how many pixels you can capture it with. If you shot subjects from afar with an FX body--even one with a high number of pixels like the Canon 1Ds series--you're cropping an awful lot of those pixels to get those you want to keep. Even the 1DsIII cropped down to the old D2xs high-speed crop area would provide fewer pixels on the same subject as the D2xs.
Some photographers seem to believe that putting a TC on the camera will "solve" the lack of pixel density problem on the D3. No, I don't think it will. First, putting a 1.4x TC on the D3 loses you a stop of ISO. Is the D3 more than one stop better than the D300 at high ISO values? We don't know yet, but you'll have to take that into account. Second, you're compromising autofocus performance, especially with the f/4 lenses (the 200-400, the 500, and 600). Third, you're compromising the optical performance (especially on the 200-400). Even without seeing the pixel performances of the respective cameras I'd wager that the D300 would still be preferred.
So, if you're a wildlife photographer, it's a Canon 40D or Nikon D300. Of the two, the Nikon seems to be the better camera feature wise. Let's hope it lives up to expectations image wise.
Isn't Nikon Still Behind Canon?
Well, let's look at how things line up for the moment:
||Relatively equal product in most respects. Nikon also has a price leader version (D40).
||No equivalent Canon product.
||Looks like the same game as the 30D/D200 one: slight feature edge to the Nikon at a slight price increment.
||No equivalent Nikon product. Or is there? Look at the products on both sides and note that the 12mp 5D is nicely bracketed by Nikon. You can get a less expensive APS version that holds its own (D300) or a more expensive version (D3) that adds capability.
||Personally, I like FX better than the 1.3x APS-H crop Canon uses. From the claimed image quality and other changes on the Nikon side, this looks like relatively equal products.
||Nikon has no equivalent product.
Other than at the high-pixel count end (Studio Pro), I don't see any big liabilities in the Nikon lineup. Of course, I'm making some assumptions about image quality, but given Nikon's changes in that regards in recent models, I'll give them the benefit of the doubt for the time being.
Don't get me wrong. Things change over time. Canon will have a 5D replacement at some point, and I can't believe that they're satisfied with only one true consumer DSLR, either. Meanwhile, Nikon is still missing that flagship high-end studio type of DSLR (but prototypes of one have been floating around--think D3 with more pixels). Back a couple of years ago I predicted that when all was said and done both companies would probably have a six-DSLR lineup, and I still think that's likely. So things'll change as new models get introduced.
But my point is this: all those emails I get about differences in Canon versus Nikon lineups really don't seem to have much validity other than if you need 21mp today. For 98% of the photographers, the two lineups are now about as closely matched as you could expect, with only modest differences in price points and features. The camera exists today (okay, in November) that does what you need. So buy the one you like and enjoy.
4mm Increase Isn't Much?
One common complaint I've already heard is that the new 24-70mm doesn't make much sense because it only goes 4mm past the already excellent 28-70mm (a lens that is also AF-S already). I guess if the difference between 74 degrees angle of view and 84 degrees isn't enough for some people. But in my experience, that's a significant and useful difference. Indeed, most of us pros don't think wide angle even starts until 28mm on an FX body. Another interesting change is that the new lens is a little slimmer and trim than the 28-70mm, which is very welcome. Personally, I applaud the changes (though don't like the idea of buying another expensive lens).
Moreover, I've been asking Nikon to get their lens lineup readjusted correctly, and that's exactly what they seem to be doing. Before today, the FX lens lineup would have been 17-35mm, 28-70mm, 70-200mm. Now it's 14-24mm, 24-70mm, 70-200mm. We lose the overlap at the wide end and gain more angle of view. Basically we now go from 12 to 115 degree coverage in three lenses, something we couldn't do before. Heck, you can go even further by just adding the 200-400mm. Four quality lenses to cover pretty much 99% of the range any of us ever shoot in. Moreover, the lenses don't have focal length overlaps (that could be a good or bad thing depending upon how you shoot, but since I'm not a fast shooter, I'll take the extra coverage over overlap; besides, if you want the overlap, just keep the old lenses).
Of course, what's missing is what made a lot of pros fall in love with Nikkors in the first place: fast primes. Where are the 24mm, 28mm, and 35mm sub-f/2.8 lenses? Where's the 50mm and 85mm f/1.4 AF-S lenses? Perhaps Nikon thinks that ISO 25800 substitutes for those, but I don't--there's a unique look you get with the fast glass that you don't get with higher ISO and slower glass. Moreover, the DX lens lineup is still mostly a consumer zoom one, and VR is missing from so many lenses that it is starting to become a liability with all the new sensor-IS cameras coming to market.
Short answer: the two new lenses are nice, but that don't come close to correcting all the missing holes in the Nikkor line.
Additional Comments on the Press Conference
Most of what you've seen on the Web as I write this is rehash of the press releases and specification sheets. The press conference which introduced the products had a few other interesting quotes (I've slightly altered them in translating them into English, but I think they're accurate as to intent) and tidbits that are worth commenting on:
- "The D300 is the highest peak of DX format performance." Nikon seems pretty certain that the D300 will eclipse all previous APS-sized sensor bodies in image quality and handling performance. I'm still a little hesitant on this one having not seen the image quality yet and looking at the NEF buffer figures, which seem a little low. On the flip side, with a UDMA card, the write speeds may be fast enough to eradicate the buffer as a big factor and perhaps the new sensor is better than previous ones.
- "The [prosumer] target is the most important group Nikon caters to." Yep. They seem to get that.
- "We would like to be the leading company for DSLRs." Specifically, Nikon wants a 40% market share and will probably get it in the near term with these latest offerings. The latest Nikon DSLR prediction for their current fiscal year is 3.2m units, by the way. Still, the pro world won't be 40% Nikon, I don't think. There's still some missing elements there.
- "FX and DX formats will continue to advance in parallel." That means new cameras in the future in both the small sensor and larger sensor formats, and new lenses for the smaller sensor format, as well.
- "Internal processing is now done at 16 bit levels." I'm not 100% sure on what's going on here--the data is coming in as 12-bit or 14-bit from the ADC (depending upon how you've set the camera), so they are probably stuffing the data into the high bits of a 16-bit field and processing from there to avoid rounding errors.
- "The 14-24mm zoom exceeds the quality of a prime lens." [I think I have that right in translation.] Well, zooms have gotten better over time, and this isn't really surprising, but remember also that the new cameras have a built-in chromatic aberration correction function, too. We should be able to get stellar results from all our lenses.
- Production: D300 is pegged at an initial production of 60k units a month, D3 is pegged at an initial production of 8k units a month. Translation: even if they start now and build at capacity until the November launch, these products will be immediate sellouts and hard to obtain (my guess).