Impressions from the Nikon Solutions Expo 30.11.-01.12.2007, Cologne, Germany
Part 2: Nikon D3 first impressions
Nikon's most recent DSLR announcements were dominated by the debut of its first full-frame model, the D3. Sporting a 12.1 Megapixel ‘FX-format’ CMOS sensor, the new D3 made big claims over high sensitivity performance, and unlike Canon’s full-frame models, it could also accommodate lenses designed for its cropped-frame bodies – albeit with a cropped frame and at a reduced resolution.
Nikon fanatics the world over are understandably excited about the D3, and Camera Labs forum moderator Thomas is no exception. Self-confessed Nikon ‘fanboy’ and passionate owner of a D80 and numerous lenses including his beloved DX 18-200mm VR, Thomas was particularly excited about checking out the new D3 for himself.
So when the Nikon Solutions Expo came within 500km of his home town, we couldn’t hold him back. What follows is the second part of an interview between Thomas and Camera Labs Editor Gordon Laing after the show. Part one featured Thomas’ views on the D300, and now in Part two, he looks at the flagship D3. The views expressed here are those of Thomas and do not reflect those of Camera Labs or any future D3 review – they are the personal impressions of an existing D80 owner considering an upgrade and presented here as an alternative point of view.
GL: Hello again Thomas, let’s talk about your experiences with the Nikon D3 at the Nikon Solutions Expo in Cologne. Tell us about your first impressions in terms of design and build quality.
TR: Big, sturdy, heavy – but well laid out! Certainly not the gear that you lug around on every short trip or dog-walk. But if you’re already packing some serious lenses, the additional 600g of the D3 over the D80 doesn’t count much. As to the layout: I felt immediately at home and it was super-easy to get that camera working under stress.
GL: What stress?
TR: Well, you know: all those eager folks at the Expo shouting and pushing around, waiting for their turn to lay their sweaty little mitts on that new body! But both grips allowed me to hold on to the new camera even with the micro Nikkor VR 105mm mounted which results in a 2kg package. Very ergonomic design! But still, any prolonged time shooting with this package will make you yearning for a tripod.
GL: How about handling? How does the continuous shooting sound, and does it slow down when shooting in 14-bit?
TR: Handling of the camera was all natural for me. The portrait grip is a boon for shooting with the camera turned to the side. The responsiveness is also fantastic: everything is just “swish”, “whirr”, “snap”, “pop”, “click”!
One thing you have to look out for is switching off the vertical release. I encountered many times holding the body horizontally, when suddenly the AF kicked in or the shutter was released inadvertently because the palm of my right hand pressed against the respective buttons on the vertical grip.
As to shooting speed: I tried it, and to my ear there was no change in frames per second whether shooting in 12bit or 14 bit. Nine frames per second sounds incredible and I am hard pushed to envisage a situation where I personally need this. But if you do exposure bracketing from -2EV to +2EV in 1EV increments to make HDR images, than you can never be fast enough.
GL: As Nikon’s first full-frame model, you’d hope the viewfinder would be big and bright – how does it look compared to your D80?
TR: Well the size of the viewfinder looked the same to me so it should be a good one stop brighter due to the amount of light hitting it. But the human eye is quite adaptive and you must have the same lens mounted to both cameras to get an appropriate comparison. So honestly, I have to rely on gut-feeling, not being able to hold one camera to the right eye and the other to the left. I’d say, well, yes, perhaps it’s brighter. But it certainly doesn’t make the D80 viewfinder look lousy in comparison.
GL: Nikon’s launched into full-frame with ‘just’ 12.1 megapixels. Canon’s had this for two years with the EOS 5D and now boasts almost double with the EOS 1Ds Mark III. Is the D3 too little too late?
TR: A direct competitor to Canon’s popular full-framer is overdue. But Nikon opted to serve their pro-segment with their first FF/FX-body. The D3 is clearly aimed as a successor to the D2Xs and as such an excellent offer!
But in my humble opinion, resolution is overrated: the difference between the 12 Megapixels on the D3 or D300 and the 10 Megapixels on my D80 is hardly something to write home about. Those two extra megapixels just crank the linear resolution up by a measly 10%. And as there is a linear correlation between resolution and noise on a sensor of a given size, it was a good decision on the Nikon side not to outdo the Canon on resolution, but instead to try to capture the speed and sensitivity-crown. Whether they succeeded can only be seen in a direct Canon versus Nikon comparison.
And as to your comparison to the Canon 1Ds Mark III, That’s a completely different proposition because there are two segments in the high-end market: Sports photographers, reporters and anyone who needs a fast camera, versus studio and landscape photographers or anyone who needs the maximum resolution. And I’d assume Nikon will be answering the latter with a 20 Megapixel-plus FX-body sooner or later!
GL: Unlike Canon, Nikon’s allowed you to mount DX-format lenses on the D3 and crop the image. How well does this work in practice and is it something you could see yourself doing as a DX lens owner? And how about the drop in resolution to 5 Mpixels?
TR: I do love this as it allows me to leave a second body for my DX-lenses at home. There’s certainly the pain of being reduced to 5MP giving a 30% or 35% lower linear resolution than on a D80 or D300, and that’s a little disappointing. This gives you less leeway to crop and can reduce pixel density of prints from, say, 240 to 160 dpi. But honestly, if you buy a D3 you’ll certainly want to shoot with FF-lenses.
Nikon did an excellent job of masking the viewfinder for 5:4 or DX-format. The masked portion is grey but still semi-translucent, so you can see what’s going on outside of the frame. And another good feature is: if you want, you can switch off the auto-crop that recognizes DX-lenses, because some of those lenses can fill the 5:4 crop or even the complete FX-frame. So you’re free to use the lens as intended or capture more.
GL: Let’s talk about specific lenses now – you did take some full-frame and DX-lenses with you. How did they look and feel on the D3?
TR: The DX-lenses naturally don’t fill the FX sensor, but I was curious whether there was a small chance at some focal lengths of filling the 5:4 (30x24mm) frame. What I didn’t try was my Sigma 10-20mm as it’s reported to have a smaller image-circle than the other wide-angle zooms around. But the Tokina 12-24mm, the Nikon 12-24mm and the Tamron 11-18mm can cover full-frame from roughly 17mm onwards. So there should a good chance to fill the 10 Megapixel 5:4 crop of the D3 even a little earlier or with better corner quality.
My 18-200mm covered the 5:4-crop from 150mm onward, but was never able to touch the full-frame. You have to live with a little vignetting, but in situations like in this shot it is not critical. That is quite interesting as it allows you to use more of the valuable real-estate of the FX-sensor.
And I brought the Micro-Nikkor VR 105mm f2.8 because that lens was giving me some trouble with focus-hunting on my D80 body. I just wanted to know, how well its AF works in connection with the new bodies. Well, I had the impression that the micro-Nikkor focuses more reliably on the new D3. But I’ve still got shots that were clearly out-of-focus with this lens.
GL: Ok, down to business again: image quality. First tell us what settings you used for the following pictures.
TR: Same as with the D300: almost everything set to “standard”, except recording 14 bit RAW + large normal JPEG with optimal quality. Well at least I thought I set it to NEF+JPEG, but when I came home I realised no NEFs were recorded, only jpegs. So unfortunately something must have gone wrong in the heat of the trials.
GL: It can certainly be hard with all the people jostling for a go at a trade show. Now let’s examine the image quality. One of the issues facing full-frame Canon owners is vignetting – not just from the lens, but from the sensor itself. How did this look on the D3?
TR: I took a lot of shots with the brand-new 14-24mm at the wide end plus wide open and couldn’t see any vignetting on more evenly lit subjects. But that is certainly not conclusive. There’s rumour on the web that with certain wide-angle lenses there is a severe light-fall-off problem that seems to be a result of the tricks Nikon played with the FX-sensor and perhaps the arrangement of the micro-lenses on it.
GL: How about noise levels and dynamic range? Were they noticeably better than the D300 and your own D80?
TR: It was obvious in some cases and not so clear in others I have seen images that look another stop better than the D300. That would make it two stops better than my D80. And I’ve seen images that look comparable to those taken with a D300 at the same ISO. To get a feeling for the noise look at this composition of D3-shots made at different ISO-settings. It shows (from left to right) the same subject shot with ISO 400, 800, 1600, 3200 and 6400.
The following shot is a 100% crop from another ISO 6400 shot with the D3 showing details plus very low noise.
This is quite impressive, but you have to remember that by default, the D3’s high-ISO NR kicks-in at 2000 ISO, whereas with the D300 the NR starts at 800 ISO, and with the D80 at sensitivities above 400 ISO. So if you compare a D3 ISO 1600 picture to a D300 ISO 1600 picture with the default settings, the latter is with NR, and the former not. Looking at those, it is not so obvious that the noise of the D3 is clearly lower than of the D300. See the following composition of D80+D3 at ISO 800 plus the D300+D3 at ISO 1600.
Those shots are not completely comparable, so you have to concentrate purely on the noise/grain in those 100% crops. This is why I love your noise-comparisons, Gordon! With those you can compare noise and detail at the same time.
If you look at the sensor-sizes one would expect a 1 to 2 EV noise advantage of the FX-sensor over a DX-sensor. This is what I believe is achievable and constitutes the single biggest advantage of the D3 over other Nikon cameras in my eyes. Just be careful with the different NR-settings if you want to compare apples to apples.
With regard to dynamic range I cannot make any comments, as that is hard to test and verify outside a lab, but I asked Nikon-employees. They said that there is no official word about DR of the D3 but that it might be around 12 stops.
GL: Before we go, any titbits or rumours about what we might expect at PMA next year? A new D80 perhaps?
TR: Nope! Unfortunately those Nikon-guys kept schtum. So we have to wait and see. I personally would be very interested in Nikon’s answer to the rumoured successor of the Canon 5D!
GL: Hmmm, you’re expecting another Nikon FX so soon? We’ll have to wait and see! And finally, now you’ve tried the D300 and D3 for yourself, will you be upgrading to either of them anytime soon?
TR: Nope, I’ll be happily shooting with my D80 until Nikon announces further plans with their newly born FX-line. Perhaps PMA in January will shed a little light on their plans.
GL: Ever the optimist Thomas! I’m betting on a D80 replacement at PMA and perhaps a successor to the D40 or D40x myself. As always, watch this space, and once again thanks Thomas for your report from Nikon Expo.
For the official Camera Labs report on the Nikon D300, see our Nikon D300 review and video tour.