Nikon D3
Or: How I Stopped Being Afraid of Darkness
(May 2008)
   
 
 

Before I start writing, I have to emphasize a very important fact: I decided to write this article from my own will. The article reflects my personal opinion about the product. Nikon did NOT ask me to write this article and Nikon did NOT pay for it. I bought the camera from my own funds with no discount, exactly like all other photographers!

Since 1980 I’ve used many different models of Nikon cameras. Usually I have used them in pairs: a professional camera for underwater photography and a semiprofessional camera for normal photography on land. I've used the models F3 and FM2, F4 and F90, F5 and F100, and D2x and D200. Recently I've changed the D200 for a D3, and so I'm now using two profession cameras: the D2x for underwater photography and the D3 for photography on land. I have to admit that no Nikon product has ever impressed me as much as the D3 did!

I do not want to repeat what has been said many times before about the D3, neither do I want to list the specifications of the camera. In this article I would like to write about what fascinated me personally in this camera and about the advantages (and disadvantages) of the FX (full frame) format. And because of the later (FX vs. DX) I will start with a comparison to its predecessor - the Nikon D2x. I've been using the D2x for almost 2 years for underwater photography and I find it very satisfying. The comparison is almost perfect, since both cameras have a similar resolution (about 12 Mpix)

 

The appearance and dimensions of the Nikon D3 are very similar to those of Nikon D2x. The buttons are on the same places, but what really caught my attention was the new 3" display (which is MUCH better than the screen on the D2x, mainly because it has an almost three times better resolution) as well as the bit higher and bigger prism. The handling of the new camera is almost identical and as an owner of the D2x it took me only a moment to familiarize with it. But not only you learn how to use it fast, you also get accustomed very fast and suddenly some characteristics of the D2x don’t seem good enough anymore (i.e. the display on the D2x immediately seemed too small, although I was completely satisfied with it  before!).

 

The D3 comes with a bundle of new features, which the user is invited to discover with the help of a 470 pages long manual. One of them is the Live view that enables framing the photo on the display and then there is also the active D-lighting, automatic correction of chromatic aberration and vignetting, a new (and better) bundle of tools for picture optimization, a virtual horizon and many more.

Why the D3? Because I wanted a camera that will (practically) never let me down in any lighting condition. On my trips abroad I dive four or even more times a day, consequently I usually run out of time to take photos on land. Often my only chance to do so is between the last dive of the day and the night dive. In tropic areas, where it gets dark already at six in the afternoon, the sun sets very fast and there is almost no time of twig light, the photographer is left in complete darkness in a matter of minutes. The rain forest with its lush treetops has very little light to offer, not to mention the dark-skinned locals that I was hunting down in dark alleys of mottled bazaars with my flash. An underwater photographer takes photos on land when he has an opportunity to do so and the lighting conditions are often far from ideal. But it is exactly in bad light conditions where the Nikon D3 triumphs without competition! Maybe some of you now think that I’m naïve and that I fell for Nikon’s propaganda, but don’t you worry: I tested the camera thoroughly before I bought it!

The Noise Test

Testing digital noise can be a very delicate matter, since in the hands of a dexterous tester, the results are easily manipulated. The amount of noise depends on many parameters, not only the ISO value, for instance: level of sharpness, quality of algorithm for noise reduction, algorithms for RAW conversion, contrast setting, color space, saturation setting, environment temperature etc.

In order to perform an objective test of comparison between two cameras, those parameters should be equivalent on both cameras. But since Nikon introduced new picture optimization protocols with the D3 / D300 generation, it is difficult to use the same parameters. However, the difference in the result of the D2x and the D3, when high ISO values were used, is so immense, that it practically made no difference to use default settings. My tests showed that the D3 beats the D2x for approximately three ISO values (that's eight times better!). With other words: the picture quality with the D3 at ISO 3200 is comparable to that with D2x at ISO 400!

 
 

The results speak for themselves, but what really impressed me, is that you can actually take the camera in your hands (without a tripod or something to put the camera on) and make a great, sharp and unshaken photo with no disturbing noise simply by using the moonlight. And when you magnify the photo on the display to check its quality, not only that it’s perfect, you even discover details that you did not spot before with your bare eye.

   
Nikon D3 and AFS 24-70 mm / f2.8 at 24 mm, 3200 ISO, f2.8, 1/15. A shot made from the hand, no tripod used. The picture is sharp, the moving people are not blured.

Resolution

Does it make sense at all to write about resolution, if both cameras have almost the same characteristic - about 12 megapixels? To be precise, the D2x has a resolution of 4288 x 2848 and the D3 4256 x 1416 pixels, so the resolution of the D3 is about 1% smaller. The path between a pixel on the sensor and a pixel on the picture is very long and depends on many factors, among others also on the algorithm for picture calculation. Pixels don’t "see" colors, that is why in front of them there are red, green and blue filters, who “see” a mosaic picture made out of 25% red, 50% green and 25% blue pixels. The color quality and sharpness of the picture depend mostly on algorithms, which calculate realistic colors.
 

But problems emerge with so called “border resolutions”, when the signal, sent from the lens to the sensor, is about the same size as a pixel. If for instance the lens makes a white line in the size of one pixel and this line falls on a red pixel, the camera will not "know" that the line was white; instead it will "think" it was red. That is how the undesired Moire effect or rainbow colors on spots, which should be uniformly coloured, happen. Anti-aliasing filters (but they reduce the photo's sharpness) and a set of complicated image processing algorithms, that take into account the values of several neighbor pixels, are being used to solve these problems. The better the algorithms and the weaker the anti-aliasing filter is used, the picture is sharper with no occurrences of the Moire effect.

The path from Nikon D2x till the D3 was (relatively) long - 3 years in which many things could have happened. My test photos, taken with both cameras and a 60 mm macro lens with a correction of the distance, because of the difference in sensor size, show that Nikon made an immense progress. In the field of border resolution we can spot a weak Moire effect with the Nikon D2x, but the picture of the D3 is entirely clear and without this undesired effect. When the camera can not differ between lines anymore, it draws an even grey coloring, which seems very natural.

   
Nikon D2x (100 %)
Nikon D2x (400 %)
 
 
Nikon D3 (100 %)
Nikon D3 (400 %)
 

Depth of Field and Diffraction

Depth of field and diffraction are inversely proportioned, connected with picture sharpness and depend on the aperture. The general rule is, that with closing the aperture we get more depth of field, but on the other hand the picture is more and more soft because of diffraction of light on the diaphragm.

According to theory depth of field depends on the reproduction ratio (the ratio between the size of a subject and its picture) and the aperture. Many people wrongly believe that depth of field depends on the focal length of the lens. This is simply not true: if you take a photo from the same distance with a 50 mm and a 100 mm lens (with the same aperture!), we will get a 2 times smaller reproduction ratio and consequently a 2 times smaller depth of field with the 100 mm. But if we move 2 times closer with the 50 mm lens, so we would get the same shot, the depth of field will stay exactly the same!

If we use the DX format (a bit less than 24 x 16 mm) or the FX format (36 x 24 mm) to take a photo with the same lens and from the same distance, we will get exactly the same depth of field. We could imagine the picture as a slide with a frame, that has a smaller hole (24 x 16 mm). The slide is still the same, but you cannot see the whole picture; the same goes for the picture with DX format, it has the same depth of field, but a smaller crop. If we want to get the same picture, we have to step 1,5 times more back or use a lens with 1,5 times smaller focal length. But with that we have changed the reproduction ratio 1,5 times and with that also the depth of field. In other words: if we take the same photo in nature with both formats, the DX sensor will have a 1,5 times bigger magnification ratio then the FX sensor (because the DX is 1,5 times smaller than the FX) and therefore 1,5 times bigger depth of field. One f-stop increases the depth of field for the root of 2 (meaning 41%), while the DX sensor increases it for 50% and functions as 1 1/3 f-stop.

The diffraction of light on the diaphragm is independent from the reproduction ratio and in a linear dependency relation to the f-stop used. Because of diffraction the same f-stop causes the same level of softness, which is perceived much better with sensors that have small pixels in contrast to those with bigger pixels. Since both sensors have a similar amount of pixels, the sensor of D2x has about 1,5 times smaller pixels (according to the sides not the surface) and is thus about 1,5 times more sensitive to light diffraction.

 

For the depth of field and diffraction test I've used a book with silver-grayish surface from flax and a page with text underneath it. I focused on the edge of the book, which is exactly 20 mm broad (so the page with letters is exactly 20 mm behind the sharpness level). With both cameras I've used a 60 mm macro lens and I was shooting from distances, which enabled me to get the same frame (the longer side of the picture is 128 mm long).

Pictures made with the D2x are completely sharp till f/11, then they start loosing sharpness as the aperture is closing more and more (for each f-stop about 40%). f/11 is the diffraction border for 12 Megapixels with the DX format for all lenses. This is the reason, why f-stops smaller than f/11 should not be used with such cameras, except when the depth of field is very significant (i.e. macro photography).

Confirming the theory, photos with the D3 are equally sharp until f/16 and only with f/22 we start noticing a slight decrease in sharpness. The 1,5 times bigger pixels give about 50% advantage, which is again about 1 1/3 f-stop. Considering diffraction the D3 has a big advantage in comparison to the D2x.

   

The test pictures also show differences in depth of field, which confirm the aforementioned theory. In this case the desires of the photographer have to choose the better fitting camera: for macro photography the D2x is better, while the D3 will take the upper hand in portrait photography.

Macro photography

All Nikon macro lenses focus till the 1:1 ratio, but because of different sensor sizes that means different shots. When you focus on 1:1 ratio with the FX sensor (D3), you get a 36 x 24 mm sized frame, but when you focus with the DX sensor (D2x) you get frame of about 24 x 16 mm size.

The test pictures of the ruler were taken with both cameras and a 105 mm macro lens on 1:1 ratio and with the same lens in combination with the 1,4x TC-14E II teleconverter. As you can see from the pictures, the new AFS VR Micro Nikkor 105mm / f2.8 gets even a bit closer then the 1:1 ratio.

   
Nikon D3 + AFS VR 105 105mm / f2.8 (1:1) Nikon D3 + AFS VR 105 105mm / f2.8 + TC 14E II (1:1)
Nikon D2x + AFS VR 105 105mm / f2.8 (1:1) Nikon D2x + AFS VR 105 105mm / f2.8 + TC 14E II (1:1)

With the D3 and the lens we get a 35 mm frame (longer side of the picture), with the D3 and the 105 mm lens and the TC teleconverter we get 24 mm, with the D2x and the lens 22 mm and with the D2x, the 105 mm lens and the TC teleconverter about 16 mm. In the field of macro photography the D2x is significantly better than the D3. In times before digital photography, underwater photographers equipped our 105 mm lenses with teleconverters in order to get better magnification, when we were shooting very small animals, but because of the converter we lost some sharpness and slowed down the auto focus greatly by making the picture darker for one aperture value. Underwater macro photographers have profited gravely with new Nikon digital cameras with the DX format, because it enables us to have a bigger magnification with only the lens, as opposed to before (and to FX format) with the teleconverter. And the cherry on top - we get more sharpness, as well as a faster and more reliable auto focus. There should be no doubt for the underwater macro photographer, that the D2x is a better choice in this field than the D3.

Tele lenses

The situation with tele lenses is similar to macro lenses. A smaller sensor is better, because it seemingly makes prolongs the range of the tele lens because of the crop effect. A picture shot with a 200 mm tele lens on the DX sensor seems like a picture shot with a 300 mm lens on a FX sensor. I intentionally used the word “seems”, because the depth of field of the 200 mm lens stays the same (if the same f-stop is used) on both sensors, and is therefore bigger in combination with the 200 mm + DX, than on the 300 mm + FX.

If we want to get similar frame with the D3 and a tele lens, like we had them in DX format, we have to buy a lens with a 1,5 times bigger focal length (which is bigger, heavier and more expensive) or use teleconverters (which is cheaper, but you lose the maximum aperture of the lens and some sharpness). Nikon offers three different teleconverters: 1,4x, 1,7x in 2,0x.

   
Nikon D2x + AFS VR 70-200 mm / f2.8 (@ 200 mm)
   
Nikon D3 + AFS VR 70-200 mm / f2.8 (@ 200 mm)
   
Nikon D3 + AFS VR 70-200 mm / f2.8 (@ 200 mm) + TC 14E II (effectively 280 mm)
   
Nikon D3 + AFS VR 70-200 mm / f2.8 (@ 200 mm) + TC 20E II (effectively 400 mm)

The test pictures were made with a AFS VR Nikkor 70-200mm f/2.8 at 200 mm with both cameras and with the 1,4x and 2,0x teleconverters. They came as a pleasant surprise: Nikon’s optics are so sharp, that even with the use of teleconverters you barely lose sharpness. A picture made with the 1,4x teleconverter and the D3 (280 mm) is almost as sharp as the picture made with D2x and the same lens without the converter. A slightly smaller contrast is noticeable. The picture made with a 2x teleconverter (400 mm) on the D3 has the biggest magnification but also a clearly smaller contrast..

Nevertheless, I find it arguable, if the DX format is really better than the Nikon D3 concerning remote photography. While this is an undisputable fact in macro photography (while taking photos from close you can easily control the lighting and consequently use low ISO values), many things depend on lighting conditions in remote photography. We cannon influence the lighting of a subject while taking photos of it from far away, so we are bound to use the available lighting. When the lighting circumstances are bad (twilight, indoor photography etc.), the D3 can boast with its low noise level at high ISO values. In good lighting conditions the Nikon D2x is better choice for remote photography, while in bad conditions the D3 proves to be better.

                                         

Underwater wide-angle photography

After the arrival of Canon’s 1Ds in 2002 some veteran underwater photographers with a set of Nikon lenses started selling their equipment and switched  to Canon’s full frame system because of wide-angle lenses, to remain their wide angles. But I thought they’ve not made a wise decision. In my personal opinion the smaller (Nikon DX) format is more suitable for underwater photography, if supported with suitable lenses. Since Nikon launched its DX Fisheye lens (10.5 mm f/2.8) any need to switch to the full frame system virtually vanished. The new 10.5 mm lens enables the photographer to see the world from the 180° point of view, and is better than the old 16 mm f/2.8 lens because it can focus from closeer distances.

When taking photos with wide-angle or fisheye lenses under water we use correctors - dome ports – which acquire rays of light from the water to the air and into the lens under an almost right angle and with that they minimize several aberrations. The optic center of the lens should be set in the center of the dome. The dome itself, which successfully abolishes aberrations of normal planparalel glass (which is used in underwater macro photography for small picture-angles), forms a strong negative lens together with the water. Two new problems appear – a focus change and curving of the field. The first problem makes very distant objects appear much closer than they really are and the only remedy is a lens, which focuses from very close (infinity usually starts at 0.5 m). Because of the second problem the picture isn't projected on a flat field (of sensor) anymore, but on a spherical surface. The picture is very sharp in the middle and loses sharpness progressively (not linearly!) when you move to the edges. The only solution to this problem is closing the aperture and thus making the depth of field bigger, so that the curved surface of the picture is still entirely in the (acceptably) sharp area. The smaller the dome, the more obvious these problems are, but on the other hand, the higher the ratio between the radius of the curve and the focal length of the lens is, the less aberrations appear. In practice the radius of the dome should exceed the focal length of the lens at least five times. Theoretically speaking, the 10.5 mm lens should give better results with the same dome then the 16 mm lens.

 

When I started using the Nikon D2x for underwater photography in 2006 I noticed, that the photos with a 10.5 mm lens and f8 are about as sharp as photos made with the 16 mm lens and f11 aperture value (on film), if the same dome is used. I never did a comparison test, simply because I’ve never inserted a film in my F5, since I started to shoot digital.
 

Now that the new full frame D3 came out, finally the time arrived for me to prove my theory. But to my surprise the new D3 proved to be very differently build than its predecessor D2x. My first disappointment started when I found out I cannot put it in my Seacam underwater housing, which I use for the D2x. A bunch of buttons and similar details are located in a slightly different places and the levers, adjusted for the D2x, were in the way, so I couldn’t fit it in the housing. And despite the higher body, the lens mount of the D3 is 2mm lower than that of D2x! The later is very important, because the optical axis of the lens has to be even with the optical axis of the dome, or the correction is not adequate. But despite all the trouble I did not give up, I started modifying the housing (or better yet removing the unnecessary elements). I better not start explaining how much time it took me to reassemble the housing ... I managed to mount the camera on the front part of the housing (I inserted a 2 mm ruler under the camera because of the calibration with the dome). Of course I could not close the housing, but this way at least I could use the camera to take photos down, the dome being fully submerged in the water.

I took photos of a ruler on the bottom of a plastic household container. My lighting came from a glowing table, which I used in the "good old days" to check slides. So the light entered from below into the bottom of the container. The test pictures, made with both cameras show how the DX (small) format is leading for at least one f-stop according to the sharpness on the edges, in front of the FX full frame format. Frame A at f/8 with the D2x looks sharper than the same frame at f/11 with the D3! In frame A with the D2x we notice a clear picture degradation because of diffraction at f/16, while this is much less visible at the same frame with the D3 (I did not spot it in the film at all!).

 

On the other hand there is a lot of noise on pictures made with the D2x (especially in frame C), as I was forced to use higher ISO values because of low lighting conditions. The unsymmetrical unsharpness in frames A is also interesting, it gets bigger in the tangential direction, rather than in the radial direction. The unsharpness looks as if the photo was shaken. I did many shots at each f-stop, but I always got the same pattern of unsharpness.

A rich underwater photographer, that could afford housings for both cameras, would use the D2x to shoot scenes of a well-lit coral reef with many details on the whole picture area, and the D3 to take photos of badly-lit wrecks at the bottom of the dark ocean.

Fisheye lenses

 

To the benefit of (mostly underwater photographers) Nikon has fisheye lenses for both of their formats: 16mm f/2.8 and DX 10.5mm f/2.8. Both lenses make the picture on the fitting camera over the whole format with 180° diagonal angle of view. But the new DX 10.5 mm is much better than the old 16 mm: it focuses from much closer distance – only 4 cm from the front lens! This feature enables underwater photographer infinite creative options in shooting extreme close-up wide-angle photos that would never been possible with the old 16 mm. And moreover, shots where everything is sharp from 4 cm till infinity are only possible under water with a very small dome. Out of water, the depth of field even with f/22 doesn‘t cover the whole area, while under water with a small dome infinity is moved much closer and even f/16 gives enough depth of field for the whole picture to be sharp.

   
Nikon D3 + Fisheye 16 mm / f2.8 (focus on the shortest distance) Nikon D2x + DX Fisheye 10.5 mm / f2.8 (focus on the shortest distance)

Nikon D3 + DX Fisheye 10.5 mm / f2.8 (auto DX mode turned off)

Nikon D2x + DX Fisheye 10.5 mm / f2.8 at underwater photography.
The size of the sea urchin in front is about 8 cm.

Even if Nikon launched a new 16 mm lens for the D3 with same option of close by focusing, as the DX 10.5 mm has it, the DX format would still be better. In order to fully use the option of underwater close-up shots, we have to use a small dome and small domes have much poorer correction features than big ones. So again the DX format proves to be better than the full frame.

Zoom lenses

A professional journalist photographer uses three zoom lenses in his job: wide-angle, standard and a short tele lens. While the tele zoom lens can be he same for both sensor formats (i.e. AFS VR 70-200mm f/2.8), it is recommended to have separate wide-angle and standard zoom lenses, so that you can fully use the features of your sensor format.

To tell the truth Nikon only has one professional DX format zoom lens: DX 17-55mm f/2.8. The wide-angle zoom lens DX 12-24mm f/4 is according to its construction and optic features maybe semiprofessional, everything else is amateur optics. The 12-24mm f/4 is sharp in most focal lengths (except for 12 mm), but this focal lenght is in my opinion the main reason to buy this lens. Its housing is not as so well build as that of  DX 17-55mm f/2.8, maximum aperture is “only” f/4, it doesn’t have a rubber protection around the lens mount for shooting in the rain (so it is not weatherproof), it does not focus from close enough and keeping in mind all of the above, it’s way to expensive. It is a pity that Nikon never offered a real professional DX wide-angle zoom lens, like the legendary AFS 17-35mm f/2.8 was in the good old days of film! According to recent development (when professional photography slowly but surely moves to the FX format) it looks like there is never going to such a lens.

The story with FX format is completely different. With film cameras we had two marvelous professional zoom lenses to choose from: AFS 17-35mm f/2.8 and AFS 28-70mm f/2.8. But the new Nikon D3 proved, that these are not sharp enough anymore. Knowing this, Nikon launched together with the D3 two entirely new professional zoom lenses: AFS 14-24mm f/2.8 and AFS 24-70mm f/2.8. Their optical and constructional features outrank the predecessors by far. The new lenses with their focal length range fit also very well on the old AFS VR 70-200mm f/2.8.

Concerning zoom lenses, the D3 is in a much better position than cameras with the DX format. Not only it has a much better wide-angle zoom with a substantially larger viewing angle and a larger maximum aperture, it also has a better sequence of focal lengths, with no gaps or unnecessary redundance. In DX format the collection of zoom lenses has focal lenghts 12-24, 17-55 and 70-200 mm, which translated into full frame measures means about: 18-36, 26-85 and 105-300 mm. The wide-angle and the standard zoom lens have some redundant area, while there is some gap between the standard and the tele zoom lens. In FX format on the other hand, the focal lengths sequence is very harmonic: 14-24, 24-70 and 70-200. No unnecessary redundance, no gaps, all three lenses are of professional quality in optical as well as in mechanical characteristics. They all have f/2,8 maximum aperture and are weatherproof. The only advantage of the DX format in zoom lenses is the longer effective focal length in tele lenses.

Move the mouse cursor over the picture!
                                         

Automatic correction of chromatic aberration

One of the best new features of the D3 (and the D300) is the automatic correction of chromatic aberration of the lens. Till now this feature was possible only when shooting in RAW format and converting it in Capture NX. Nikon D3 (and D300) have a powerful processing unit and use the same algorithms as Capture NX, hence automatically correcting chromatic aberration when generating JPG pictures. JPG photos taken with the D3 (or D300) look sharper than photos taken with other Nikon digital cameras, even if taken with the SAME lens!

   
Move the mouse cursor over the picture!

Virtual horizon

 

Whenever I bought a Nikon camera, I always added a type E focusing screen with referential lines, intended for architecture photography. Even though I am not an architecture photographer, I’ve used it to take photos of seaside landscapes, so that the sea was level.  With the Nikon D3 I do not need this focusing screen anymore. I simply programmed a function button of the camera to show me an “analog” virtual gyroscope in the viewfinder (which is by the way incredibly precise) and with it I can sea the tilt of the camera to determine when it is level. The gyroscope functions just as well for horizontal as for the vertical format (if the camera is tilted over 45°, it automatically changes from horizontal to vertical and vice versa). You can see the gyroscope also on the last display, but in my opinion it is much less useful then the one in the viewfinder. And this is how the sea level became level again.

Speed

Speed is one of the most essential features of the D3, but to me personally it doesn’t matter so much. Nevertheless, I’ve tested weather Nikon's claim that the camera can do 9 shots per second in full resolution is true. The nine sequential photos shot at maximum speed show a stop watch. The difference between the first and the last is 1.01 second....

Autofocus

There is absolutely no doubt that the Nikon D3 has a great autofocus, which has surely been improved since the D2x. I’ve noticed that it works very well even in darkness (in condition where the exposure meter gave up long ago, the autofocus still worked for static objects!).

In my work the greatest challenge for the autofocus are small and swift fishes, which I want to shoot with a macro lens. Since their speed is very high, you could compare it to taking photos of bigger, but very fast-moving objects with a telephoto lens. But since I cannot fit my D3 into the underwater housing of D2x, I will never know how well it’d have done with making sharp photos of those tiny fast fish.

Because I had no other option, I've tested the dynamic autofocus on a cyclist with the 200 mm telephoto lens and at the speed of 9 shots per second. Because of the cloudy weather I’ve used ISO 2500. The f-stop was f/4 and shutter speed between 1/800 and 1/1250 sec. The average speed of the cyclist was 5,5 m/s (20 km/h) and at the closest point she was about 2 m away from me. I know that it is not the most difficult of challenges for the autofocus and I’m aware that the fishes would’ve been much more demanding, but the percent of sharp pictures in six rounds was more than 95%! The most common mistake that occurred, was sharpness on the ear and not on the eye. It never focused on objects in the background. Even the pictures from the closest point were sharp, even though the relative speed is the highest there. I believe, despite the rather simple task it had to master, I can say that the autofocus is very quick and reliable.

What is bad about the D3?

Every decent test should write about the bad sides of a product, just as well as about the good ones, and that is the hardest part of the test. Is there anything bad about the D3 at all? Well, since nothing is perfect, even the D3 has a few bad features. I do not intend to write about physical or geometrical features of the sensor (i.e. not as good for macro photography as the DX format), but about technical features of the product, which could have been better.

One of these deficiencies are the AF sensors, which could have been more far apart, like in the D2x (and D300). In the D3 they are concentrated in the center of the format, which can be disturbing in some compositions, when the main object is far from the center.

The second thing is the sensibility of the exposure meter. It is as sensitive as in other Nikon cameras, but because of the amazing possibilities of shooting in low lighting conditions the D3 offers, this is simply not enough! For instance when I shoot at night at ISO 6400 on "A" (shutter priority), the message "Low" appears in the viewfinder. Nevertheless I take the photo and I get a well-lit shot! From the picture information I can see, that the camera used 1/4 sec. What a pity, that I did not see this information in the viewfinder before - it's very inconvenient for the photographer not to know, how long the exposure will last.

When to choose D3 and when D2x?

D3 has an advantage over the D2x in low lighting conditions and everything connected to it. It has higher ISO sensitivity values and less noise at the same ISO value. It is faster, especially when shooting in high resolution. It’s loaded with new features, which help you shoot better photos. It has more professional lenses to chose from, especially wide-angle lenses, which give it another advantage over the DX format.

On the other hand, the Nikon D2x is better at macro photography (especially underwater macro photography), at wide-angle underwater photography when using dome ports , and has a better fisheye lens. Additionally, the telephoto lenses have a longer effective focal length (usable in good lighting conditions).

Which camera to choose (OK, the D2x is "history" by now, if you’re buying today, you will probably be deciding between the D3 and the D300), depends mostly on our needs and our way of photographing. Finally it also depends on our budget. In the ideal case, you should buy both....

Did I act to hastily buying the new D3?

There was much said about the Nikon D3 on the internet, among it many speculations and guesses about the camera’s successor, even though the D3 hardly arrived on the shelves. The biggest resentment some (mostly theoretics in my opinion) have towards the D3 are the “humble” 12 Mpixels. Once upon a time, before I started digital photography with my D2x, I was one of these people, thinking that 12 Mpixels is simply not enough. Even a slide, scanned at 4000 dpi has about 21 Mpixels! But at that time I didn’t know that pixels from a digital camera are much more “effective” then scanned film and consequently the pictures are sharper. I printed many photos from my D2x in 120 x 80 cm format and I couldn’t possibly say that they’re not sharp enough. Definitely they’re much sharper, than any 35 mm film magnification ever was!

Because the D3 has “only” 12 megapixels, its pixels are relatively big and this is one of the reasons why it excels in photographing with low light conditions. If Nikon would build in a sensor with the same technology, but double the amount of pixels (24 Mpixels), they would gain only about 40% resolution, but lose 50% of picture processing speed (maximal speed would be 4,5 shots per second) and the noise would worsen for one ISO value. Such a camera would also come with other bad features of the DX format: the picture degradation because of diffraction would be visible from f/11 on and the pictures would be about 40% more shaken, because of smaller pixel size. If you wanted to use all of such a camera’s resolution, you would've had to shoot from a tripod. But the core of Nikon D3 is its agility; it’s made for action and for situations when there is no time to set a tripod. A camera with 24 Megapixels would be more useful (maybe only) in studio and in landscape photography.

Should I forget about the D3 and rather wait for the new D3x? For now, D3 is Nikon’s only full frame camera and the choice to buy it (except for the financial point of view) was anything but hard. In case I had to choose between the D3 and a hypothetical (abovementioned) version of the D3x (with 24Mpixels) it would have been more difficult to pick the better one. But considering what I need the camera for (when I'm on land and not under water), I’d probably have chosen the D3. With almost 30 kg underwater photo equipment, that I have to "smuggle" on board of a plane (I hope no airport personnel is reading this), I definitely don’t have space for a tripod. I always shoot from the hand and I hardly ever have time to wait for better lighting conditions. I admit, a higher resolution could come handy at times, but the price for it (loosing so many great features of the D3) is to high for me to pay. And in case I had both cameras, I’m sure the D3 would be my travel companion.

Post scriptum

Once I was taking night photos of Piran, an old Slovenian sea town, when a stranger came to me and said: "Although I didn’t see what it says on the camera, I knew you’re shooting with a D3! In such darkness, you can't shoot from your hand with anything but the D3."