Nikon D800 in Praxis (Part 2: Underwater Photography)

 
(December 2012)
Introduction
 

I use Seacam housing for my underwater photography, which I've reviewed in my next article. This review will be devoted to my first impressions of using the D800 for underwater photography. Until the date of writing, I made about a hundred dives using the camera in Egypt and Indonesia and I believe this was enough to form a balanced opinion about it.

At the time of writing this article, the Nikon D800 is in my personal opinion without a doubt the best Nikon camera for underwater photography. As soon as I put it in the underwater housing, I realized that I'm holding equipment with a great potential in my hands and I will try to prove that in this review.

Let me clarify in the beginning that out of my two cameras (D800 and D800E), I only use the D800 for underwater photography and never the D800E! In underwater photography stopped down apertures are used. In macro photography we stop down the aperture value in order to increase the depth of field, while in wide-angle photography we do it to reduce the optic aberration of the dome port. Because of the diffraction effect all underwater photos are slightly unsharp (on a pixel level!). In addition to that, there is a lot of plankton in the water, which influences photo sharpness on a micro level too. Light diffracts on their microscopic bodies or, if they are transparent and consequently function as a sort of micro-lenses, the light refract on them. The longer the path of light through the water, the more interactions the light has with plankton, the softer the photo will be. Due to all the above mentioned underwater shots are always slightly less sharp than land photos. The D800E would have an advantage over the D800 under water only when shooting on very short distances with wide apertures (i.e. 1:1 macro photography with f-stop f/5.6), which is not used in real shooting conditions, except when we want to achieve a special effect.

As soon as the D800 had been announced (much before I ever held it in my hands), I was already thinking what the ideal f-stop for macro photography would be to achieve the best compromise between diffraction and depth of field. I was also worried, how the 36 MP FX sensor will handle the optical aberrations of the dome port, which was much more visible on the 12 MP D3 sensor, compared to the 12 MP D2X (here is an explanation why). Unfortunately, I only received the Seacam housing a few days before my departure to Egypt (or maybe I was actually lucky to have received it at all, considering the great demand for this product!), so I did not have time to carry out tests with the camera in the domestic sea (or at least in the swimming pool), to get some answers to my questions.

 
Domeports and Wide-angle Photography
 

I was much more worried about wide-angle photography than about macro. A few months before I've received the housing for the D800, I remodelled the frontal part of my old Seacam F5 housing in a way that it could house the D800. Back then I was mostly interested in how my dome port for fisheye lenses (Seacam Fisheye port - FP) will function as compared to the big dome port (Seacam Superdome - SD), which I borrowed for this test. I used the DX 10.5 mm f/2.8 Fisheye lens for all tests. The FP is an almost ideal half-sphere (my measures showed that it's lacking only 6 mm; the outer diameter of the glass is 158 mm, and the height is 73 mm), so it has an almost ideal position for the Fisheye lens. But due to its smaller radius of curvature, it bends the picture more (field curvature), thus causing a greater unsharpness in the corners. The SD (the diameter of the glass is 228 mm, the height 81 mm, and the outer radius of curvature (calculated) 120 mm) on the other hand, has a greater radius of curvature, meaning that it bends the picture less and enables greater sharpness in the corners. But since it's not a half-sphere, the Fisheye lens cannot be positioned ideally in the dome port (in its spherical centre), so the angle rays refract when entering, which causes a reduction of the viewing angle and additional chromatic aberration. My calculations showed that the viewing angle of the 180° Fisheye lens is reduced by the use of the Seacam SD dome port for 5° (on each side), meaning that the new viewing angle of the lens is 170°. Tests in the bathtub clearly proved my calculation. So what is the influence of these two opposing effects (field curvature and chromatic aberration)? The Superdome is the clear winner! There is a negligible increase in chromatic aberration because of the refracting of side rays, while at the same time the unsharpness in the corners is reduced due to the greater radius of curvature of the dome port. True, the difference is not big, but it is noticeable. Therefore, when I ordered my new housing for the D800 I also bought the Seacam Superdome.

 
 
Seacam Superdome (SD) and Fisheye port (FP)
 

When the new housing together with the new dome port finally arrived just a few days before my departure to Egypt, I repeated the bathtub test. This time I only tested the Superdome with Fisheye and wide-angle lenses. A new dilemma arose: should I use DX lenses and get better corner sharpness, but a lower picture resolution (15 MP), or FX lenses which enable me a greater resolution (36 MP), but less corner sharpness (again, you can find the explanation here)?

 
 

In the array of Fisheye lenses, I've tested two of my own: the Nikon 10.5 mm f/2.8 and the Tokina 10-17 mm f/3.5-4.5 as well as two borrowed ones: the Nikon 16 mm f/2.8 and the Sigma 15 mm f/2.8. I sold my old 16 mm Fisheye lens soon after I bought the D2X, since it turned into a "curved" 24 mm, and I was highly satisfied with the new 10.5 mm lens.

My tests showed that the Nikon 10.5 mm is notably sharper than the Tokina 10-17 mm, even though I think (but I am not sure...), I've seen some reports on the Internet, which claim the opposite. Maybe this is again sample to sample variation of the same model, which would mean that I have a good Nikon and a bad Tokina.

An entirely different story is the 16 mm Nikon and 15 mm Sigma lenses. Despite the general popularity and good reputation of Sigma, the lens I've borrowed was an utter disappointment. The 16 mm Nikon was far sharper than the Sigma! The difference between them was so big, that I decided to test the lenses outside of the water as well and even there the Nikon was clearly better than the Sigma. Maybe I borrowed a faulty specimen, but ever since I don't even want to hear about Sigma Fisheye lenses! The Sigma 150 mm f/2.8 macro lens however, is nothing like that – it's an excellent lens and optically fully comparable to Nikon's macro lenses.

Let's return to the comparison between DX and FX lenses on a 36 MP sensor. The 10.5 mm DX lens is the absolute winner, as I have also expected it to be. The 16 mm has the advantage of resolution in the central area of the picture, but the 36 MB sensor shows a great and in my opinion unacceptable unsharpness in the corners. The lenses are comparable in corner sharpness only at f-stop f/22, when diffraction sets in. I've noticed that many underwater photographers prefer FX Fisheye lenses, but to me an even sharpness throughout the entire photo is very important. And besides, isn't 15 MP, which the D800 enables in its DX crop, enough for most applications? The effect of a field curvature can be reduced with FX fisheyes to a certain extent by not focusing on the central sensor, but instead on the most extreme sensor points, which are kind of half-ways between the centre and the corner of a picture; however this only lessens the effect, there is no way to fully annul it. My simple test in the bathtub lifted all doubts: for me the only acceptable lens for underwater fisheye photography with the Nikon D800 are the Nikon DX 10.5 mm Fisheye lens and the Tokina 10-17 mm!

Surprisingly, among wide-angle zoom lenses, the FX 16-35 mm f/4 did unbelievably well compared to the DX 10-24 mm f/3.5-4.5 lens. Of course the DX lens is sharper in the corners, but the 16-35 lens is not bad at all! I only tested the widest angle available. Consequently I decided that I will use the FX wide-angle zoom lens with my D800 and not the DX anymore. The reason for this is mostly my personal philosophy on the use of wide-angle lenses in underwater photography. I use Fisheye lenses mostly to shoot underwater landscapes, so sharpness in corners is very important to me. Because of their flexibility I use wide-angle zoom lenses mostly for shooting bigger animals. These swim in the open water or lie down on the sandy bottom and therefore, there is no important details in the corners. Consequently I give priority to central resolution before corner sharpness.
In order to achieve good results, a good dome port position in regard to the lens is of a vital importance. In the ideal case, the spherical centre must be positioned exactly in the optical pupil of the lens. Harald Hordosch (the owner of Seacam) suggested me to use 70 mm extension rings with my 16-35 mm lens and the Superdome. However, my measurement showed even more distance is required, since the lens should lie deeper in the dome port in order for the light rays to fall through the glass as perpendicular as possible. I've tried with 90 mm and I did actually achieve a slightly better result than with the 70 mm rings.

I've also tested the 14-24 mm f/2.8 lens at 14 mm, but I was not satisfied with the corner sharpness.

In the Red Sea I shot many photos with the 10.5 mm f/2.8 lens with the Superdome and I can really say that I'm extremely pleased with the results. Bellow you can find and example from the Red Sea in which you can clearly see how the sharpness in the edges is no smaller than in the central area of the photo (10.5/2.8, f/11, 1/125, ISO 200, DX crop).
 
 
 
 
I've only used my 16-35/4 lens on one dive, and even that more to test it. It would be much more useful in the Bahamas, where I was shooting sharks and dolphins, but unfortunately I did not have the housing for the D800 yet at that time. Back then I was using the 10-24/3.5-4.5 lens (with D2X) on practically all of my Bahamas dives. The example below taken in the Red Sea clearly shows, how the photo is sharper in the central area compared to the corners, but for me this is still an acceptable result while shooting big animals, where there is no relevant details in the corners (16-35/4, 16 mm, f/13, 1/40, ISO 200).
 
 
 
 
Macro Photography
 
 

This photo plainly shows how small certain "monsters" of the sea can be. (105/2.8, f/22, 1/250, ISO 100)

When I compared the 12 MP sensors of Nikon D2X and D3, I wrote that the DX sensor has certain advantages in macro photography compared to the FX censor, due to a greater pixels density, which means that a camera with the DX sensor captures the same object in nature with a greater number of pixels than a camera with the FX sensor. Many things changed with the arrival of the D800, considering that its DX crop features a greater resolution than any DX camera before 2010 (before the launch of the D7000). Only the newest D3200 and D5200 models with its 24 MP on a DX sensors enable a greater pixel density than the D800, however we must keep in mind that they are entrance-level cameras, which are in general less suitable for underwater photography.

Because of the high resolution, the cropping options and excellent autofocus, the Nikon D800 offers great possibilities for macro photography. The only unsolved problem (a problem of all sensors with high pixel density) is that of the depth of field. If we want to increase it by stopping down aperture, the photo becomes soft due due to diffraction. Macro photographers need to be exceptionally careful and meticulous when focusing. In order to avoid the problem of the small depth of field, the photographer needs to anticipate where the plane of sharpness will be and hold the camera as perpendicularly to the object as possible.

The old technique of focusing in macro photography (which I used with the D2X), where I focused with the central AF sensor on the eye of a static animal and then slightly corrected the composition and took the shot, is definitely out of question with the D800. The animal's eye would be unsharp in any case. Even if I tried very hard to stabilize the camera while changing the composition and held the camera housing on the sea bottom (thus eliminating the effects of shaking and changing the position of the camera in regard to the object), the eye would still not be perfectly focused.

Let's suppose we want to take a picture of a 40 mm long fish from a distance of 100 mm in a way that the fish will be exactly in the middle of the shot. The distance between the lens and the middle of the fish's body is exactly 100 mm. The eye of the fish is 20 mm away from the point where we set the centre of our composition, this means the distance between the eye and the lens is actually 102 mm (if the fish is perpendicular to the optical axis of the lens). If we turn the camera towards the eye and focus on that distance (102 mm), then turn the camera back so that the fish will be in the middle of the composition (and the eye 20 mm outside of the centre of the shot), the camera will still be focused on 102 mm, while the fish remains on a 100 mm distance from the lens. These two millimetres, believe it or not, matter and the eye is not perfectly focused!

 
 

Some might find my theories to be over-complex, but the photo of a goby protecting its eggs on a whip coral is a great example of what I was trying to say before. On the 100% crop of the goby's head we can see how shallow the depth of field is (using f-stop f/16). The eye is in perfect focus, but the mouth, which is less than a millimetre more distant from the lens than the eye is already unsharp (105/2.8, f/16, 1/250, ISO 200). To achieve this sharpness in the eye, I first composed the photo and only then focused (on the eye)!

 
 
 

If we are lazy (or from any other reason a fan of focusing with the central sensor), we can help our self with another technique: we can always position the eye of the animal in the exact centre of the photo and then later crop the picture to a desired composition in photo-editing program in the computer (with a 36 MP image this is obviously not a problem).

 
 

The bargibanti pygmy seahorse is a very popular theme in underwater photography. The seahorse from Lembeh strait, Indonesia in the photo above was shot with the 105 mm f/2.8 lens in combination with the TC14E teleconverter in DX crop (the size of the vertical side of the picture is about 17 mm). I've used f/22 – the biggest f-stop that I still use with the D800. At f/32 the photo would have been already too soft due to diffraction, that I couldn't have sharpened it to a satisfactory level in photo editing programs.

 
 

The crop displays a portrait of the seahorse as it was captured by the camera with the following settings (105/2.8 + TC1.4X, f/22, 1/250, ISO 100, DX crop).

If you slide the mouse over the photo, you see the photo that was later sharpened with photo editing software. Generally speaking, I believe it's better to stop down the aperture a bit more because of the increase in depth of field and then sharpen the photos at a later stage. If parts of the picture are "soft", we can recover that later, if they're unsharp due to insufficient depth of field we can't do anything about it. If we want the entire object or at least the majority of it to be more or less sharp, we have to stop down the aperture.

 
 
 

Portrait of a shrimp (105/2.8, f/22, 1/250, ISO 100). I photographed the shrimp (intentionally) from the front side, since I wanted that only a small part of its body will be focused. In such photos it's crucial that the most important elements of the face - the eyes - are in focus. The 100% crop below clearly shows that the eyes are really in focus, and that nothing else is in focus except the eyes... To achieve such results it's crucial that the camera used has a top-level autofocus system and that we focus with the AF sensor that is positioned exactly on the eye of the subject.

The photo of a young harlequin shrimp (size about 2 cm) on a blue starfish and its 100% crop shows again how small the depth of field at the pixel level is. (105/2.8, f/18, 1/250, ISO 100)
 
 
 

However, the small depth of field is just a perceived problem of the D800, for it is actually a characteristic of the high-resolution sensor. Depth of field is created by the lens, not the sensor and it's completely equal on all sensors, when shooting under the same conditions (same reproduction ratio and same f-stop value). The same shot taken with the D3 would seem to have a greater depth of field (on the pixel level), but if we printed pictures from both cameras in the same size, they would have been completely equal regarding the depth of field. We would also get the same depth of field (on the pixel level) if we reduced the size of the 36 MP D800 picture to 12 MP.

 
Resolution
 

The very high resolution of the D800 is less relevant in underwater photography especially because of the optical imperfections of water, which I mentioned earlier. The greatest gain from a high resolution is in macro photography, where we can shoot from a short distance with very sharp macro lenses.

The photo below shows the surface of a rock, which the damselfish in the Red Sea was very determined to protect. The 100% crop shows exactly why that was the case (105/2.8, f/16, 1/125, ISO 100).

 
 
 

Details on the face of the small Indonesian glassfish in the middle of the school are very clear on the photo below, but only if the subjects face is exactly in the plane of focus (105/2.8, f/16, 1/250, ISO 100).

 
 
 
ISO Settings
 

Despite the high resolution the D800 works very well with high ISO speeds. The photo below of a goby on a whip coral was taken at ISO 800. That enabled me to capture the blue ambient tone of the surrounding water without a problem (105/2.8, f/16, 1/125, ISO 800). In the pre-digital era, when I was limited to using ISO 100, I had to put much more efforts into taking a shot like that, considering that I had to use an 8x longer shutter speed (hand-held of course!), which often led to photos being shaken or blurred. I somewhat solved the problem by shooting directly towards the sun, which shortened the shutter speed to some degree.

 
 

The 100% crop below clearly shows that the picture is not shaken at all, what is more the ISO 800 setting is low enough for the D800 that there is no disturbing noise visible in the photo.

 
 

Ultra-violet fluorescent photography was limited mostly to macro photography until recently. The UV light can trigger a weak fluorescent glow in some sea animals (especially corals). With the use of special filters that are mounted on the flashlight, one can block the visible part of the spectrum so that the flashlight emits mostly UV light and a small part of the blue spectrum. With another filter, which is yellow and has to be mounted on the lens, one can block UV and blue light, so that only fluorescent light, which is very weak, can reach the sensor. Therefore, most fluorescent subjects are usually limited to macro photography, as less light is being absorbed on such a short distance.

The Nikon D800 broke these limitations and for the first time I could carry out something I've been dreaming about for a long time – wide-angle UV fluorescent photography. It was in Indonesia that I was shooting for the first time at night with the 16-35 mm f/4 lens, because I cannot mount the yellow filter on my fisheye lens. The photo below shows a coral acropora, with a diameter of about 1 m. It was taken at ISO 3200.

 
 

The 100% crop below shows a slightly reduced sharpness due to the high ISO setting, however the photo is still sharp enough and practically without any disturbing noise (16-35/4, 16 mm, f/10, 1/250, ISO 3200).

 
 

Green tones are dominant in underwater UV fluorescent photography, because the green fluorescence is the strongest. It is actually so strong, that I was able to shoot relatively good macro shots already with the D2X at ISO 400. Other fluorescent tones, especially red, are much weaker. The picture and 100% crop below show a detail of a coral with red fluorescence taken at ISO 3200 (105/2.8, f/16, 1/250, ISO 3200).

 
 
 
Autofocus
 

The autofocus of the Nikon D800 (and D4) is truly fantastic. I've never experienced anything like it before. Before I continue describing my experience, I want to explain some things.

Until now I've been using the D3 for land photography only. My land photography was mostly limited to travel reporting work, portraits (mostly exotic), landscapes and wild-life animals. I was never interested in sports photography for which I would need a top-level autofocus system. It's reasonably easy for a good autofocus to follow a bird in the sky, because the bird is usually flying in a linear and predictable way. Therefore I was never in a situation where I'd need to use the D3's entire autofocus potential. In other words: I don't really know how much the D3 can actually do.

Underwater photography, however, is an entirely different story and maybe it presents the most demanding task possible for the autofocus system. It has to deal with small objects in a short distance, where the distance is changing quickly, the light conditions are usually very bad and most subjects (fish) are moving in a completely random and unpredictable way. Because of these reasons, I often found myself in situations that my D2X (and before that F5) just couldn't master. I simply cannot know how the D3 would handle such situations under water.

When I got the D800 I skipped using one generation of AF sensors under water and immediately started using the newest technical wonder. The number of sharp photos jumped instantly, however I also spent more time deleting unnecessary photos. Even after almost hundred dives with the D800 I couldn't really say that I have fully managed its AF system. The system is incredibly complex and offers many possibilities; consequently it's difficult to find out which featured option is truly the best for a specific situation in underwater photography.

I might have gotten used to the new AF system faster, had I been previously shooting with the D3 under water. But when I switched from the D2X I had to change my personal photographing philosophy from its roots. With the D2X I usually used the AF-S (single-servo) mode. I used it with most of my wide-angle and static macro shots and sometimes even when photographing fish, in which case I constantly pressed the shutter release half ways to correct the focus.

With the D800 this technique is out of question. Not only that it is highly unreliable on the 36 MP sensor, it is also unnecessary. The AF-C (continuous-servo) mode has been perfected to a very high degree (at least comparing to the D2X) and works excellently even in bad lighting conditions and with short distances. At this point I have to say that I chose "Focus priority" in "Custom settings" for both autofocus modes, so that the camera only captures a picture when it is sharp.

The AF-S mode is different with the D800 comparing to how it used to be with the D2X and the D3. In this mode there are only two options – focusing on only one sensor, which is chosen by the user or auto area focusing, where the camera chooses the most suitable sensor according to its priorities. This way the photographer has no control and in my opinion this mode is useful only for wide-angle shots with a great depth of field. The camera doesn't necessarily focus the closest object; it might be the brightest or the most contrasting one. Luckily the photographer can see in the viewfinder which sensor is active. I use AF-S almost exclusively for wide-angle photography and only very rarely for macro photography (when the macro object is static and I can lie on the sandy bottom).

In underwater photography the AF-C mode is the primary mode of autofocus with the D800. However we have to consider which of the six possible settings of this mode is the most suitable for a specific focusing situation (1-, 9-, 21- and 51-Point dynamic area, Auto area or 3D-tracking).

When shooting a static macro and I want the sharpness to be in a specific spot (i.e. the eye), but I am not very stable (i.e. when I'm floating in the water so as not to harm the organisms on the sea bottom), I focus with one AF sensor, which I chose with the multi selector on the camera's back. The AF-C mode on the D800 is so precise, that it compensates even the hand shaking and (almost) always captures a sharp shot. Sometimes, when the sharpness is not so critical (i.e. when the eye is bigger), I decide to focus with a group of nine AF sensors.

Small fish, which swim unpredictably in all directions (especially in bad lighting conditions) are the biggest challenge for the autofocus system in underwater photography. I think (even though I'm not sure...), that the 3D-tracking mode is most suitable for such situations, despite the fact that it takes the colour of the subject into account but underwater environment is monochromatic blue.

 
 

The clownfish on the photo above (60/2.8, f/16, 1/250, ISO 100) is a very restless fish, who swims randomly and unpredictably over its sea anemone. If we use the AF-S mode in combination with 3D-tracking, the AF sensors literally stick on the fish and the result is a very high percentage of sharp photos. Technically speaking all photos are sharp, but not all of them in the right place (the eye). It's truly fascinating to observe in the viewfinder how the camera chooses the most suitable AF sensor and focuses on it.

 
 

The pontohi pygmy seahorse from Lembeh strait, Indonesia is no bigger than 10 mm and it doesn't stand still even for a second, therefore it's technically one of the most challenging subjects to shoot in underwater macro photography (105/2.8, f/18, 1/250, ISO 100, DX crop). The seahorse on the photo gave me quite a chase and despite the top-level AF system it took quite a lot of time to capture some technically successful photos. Since the subject is very small and restless, I used the 9-point dynamic area mode.

The AF system of Nikon D800 is very precise. If we choose Focus priority in personal settings, the camera will really only focus when the subject is completely sharp. At the beginning when I didn't know the system well enough I was a little disappointed, because the camera wouldn't focus when I was shooting with a +10 wet dioptre (a dioptre lens that is mounted underwater in front of the macro lens from the outer side of the housing). When the camera finally managed to focus, I analysed the photos on the computer and I discovered that there is no real sharpness anywhere on the photo. The reason for this is the dioptre lens, which is simply not sharp enough for the 36 MP sensor! And since the camera didn't see any sharpness, it didn't want to release the shutter... Photos taken with a +5 wet dioptre are sharper and the D800 has much less trouble focusing with it.
At the same time I also realized that the D800 excellent focus with the 105 mm f/2.8 macro lens attached on the TC14E teleconverter (which didn't really work with the D2X...). Considering that the photos were much sharper, I decided to go with the teleconverters rather than the wet dioptres, which is a pity in a way, since the wet dioptres are much more practical (we can mount or unmount them while underwater, while the teleconverter is fixed on the lens through the entire dive).

Back when I was using the D2X I preferred the Sigma 150 mm f/2.8 lens to the Nikon 105 mm f/2.8. One of the main reasons was a slower autofocus system of the Sigma, which the D2X could handle very well around the reproduction ratio 1:1. The autofocus system using Nikon's 105 mm was quite unreliable around the reproduction ratio 1:1 and the camera had to refocus many times. However, the situation is completely different with the D800. I rarely use the sluggish Sigma now, while Nikon's 105 mm lens works excellently and reliably even around the reproduction ratio 1:1. Because of its speed it's a great lens for fish photography.

Because it can focus even at f/8, the D800 can focus well with the 1.4X teleconverters as well. On my travels I did not have my 2x teleconverter with me, but I found out later at home, that this combination focuses well around infinity, but it fails around the reproduction ratio 1:1. Just recently I've also bought the TC17E teleconverter, which is ranging somewhere in the middle between the TC14E and the TC20E according to my tests – regarding sharpness and reliability of the autofocus system. I haven't used it yet to shoot underwater.

When I got well acquainted with the features of the D800's autofocus system, I removed all manual focusing gears from my macro lenses, that I kept mounted until then just in case, if the autofocus should fail. With the D2X it happened sometimes that the autofocus failed, but with the D800 this almost never happens. I wrote "almost" - of course even the D800 sometimes fails, but in situations like that manual focusing would be of no help, since all manually focused shots around the reproduction ratio 1:1 are completely unsharp on the 36 MP sensor!

Last but not least, I want to mention another advantage of Nikon D800's AF system – its ability to focus even in very low-light conditions. Nikon really made a great leap forward compared to the D2X (and maybe even the D3). To demonstrate this, let's look at UV fluorescent photographs again.

 
 

When I shot the photo above (16-35/4, 30 mm, f/11, 1/250, ISO 3200), the surroundings were so dark due to the very weak fluorescent light that I basically couldn't see the subject in my viewfinder and had to anticipate how to compose the photo! The autofocus was searching and refocusing for some time, a second or two, sometimes even ten seconds, but in the end it did focus the subject and the camera captured a completely sharp photo! I would like to emphasize that the photo was taken with a relatively strong flashlight, while the camera only had the weak light available to focus, coming from pilots lamps (which were filtered with strong UV filters too).

 
Sunburst
 

The sunburst is probably the weakest spot of digital underwater photography, especially for older underwater photographers, who vividly recall how beautifully the sunburst was captured on film. On film we could capture the nuances from white to light blue to dark blue, but with most (if not all) digital sensors this transition is marked by an unnatural cyan ring around the sun's whiteness. Newer cameras with a greater dynamic range capture this transition better, but still not as beautiful as with the "good old" film.

The intensity of the cyan ring depends on many factors, first and foremost on the quality of water. The more greenish the water, the less intensive the ring and the more beautiful the blue tone of the water, the uglier the cyan ring. However, we can control it quite well by changing the colour temperature under white balance settings. The lower we set the colour temperature (i.e. 4000 K), the more beautiful the blueness of the water, but at the same time the cyan ring around the sun will be uglier and more intensive. At higher settings (i.e. 6000 K) the water will have a less attractive blue colour, yet the cyan ring will be less obvious.

The cyan ring is also less visible if we underexpose the photo. This way the sun turns out clearer, but other parts of the photo (which are not lighted with a flashlight) are unfortunately to dark.

Digital sensors show a cyan ring because the strong light around the sun turns not only blue but also green pixels into a saturated state (however not the red ones), and the result of blue and green is cyan. Sensors with a greater dynamic range get less saturated green pixels; therefore the cyan ring is less distinctive. The Nikon D800 has a very big dynamic range comparing to other digital cameras of our time, so one would expect that the cyan ring around the sun will be minimal.

So how does the D800 capture the sun in reality? On my trips to Egypt and Indonesia I unfortunately didn't have the opportunity to compare it to my old D2X. My personal impression is, that it does the job better, yet not "revolutionary" better.

Only after returning to Slovenia I had the opportunity to test and compare the D800 with the Nikon D200, whose sensor is of the same generation as that of the D2X. Unfortunately the Slovenian sea is rather greenish, so the cyan ring problem was less visible. I enhanced it intentionally by setting the colour temperature to a very low value.

I used exactly the same settings with both cameras and even used the same dome port (Seacam Fisheye port) and same lens (Nikon 10.5 mm f/2.8). The settings I chose: f/16, 1/250, ISO 200, WB 4000K. I was shooting with both cameras at the same time, since I had them both with me under water. Then I processed the photos (RAW files) in the Lightroom 4 with the same colour temperature setting I used (4000K), a lower one (2500K) and a higher one (5000K).

 
 
 
 

As we can see, there is a difference, yet it is not as great as one would wish. We can notice some cyan colour even with the D800 (especially at the 2500K setting), but the transition is much more beautiful (and less sharp) than with the D200.

In the photos below we can see 100% crops of the sun. The crops shown were taken with the D800 and when you slide your mouse over them, you can see the crop from the D200.
 
 
 
 

To conclude I would like to show younger readers, who never took underwater photos on film and therefore cannot understand what I mean (and why I am so disgruntled), a photo taken on film in the Red Sea in 1999. The Red Sea is (or at least used to be) very clean and blue, so the cyan ring should be even more expressed. The high graininess of the film (despite low ISO speeds) and the relatively low sharpness compared to digital photos are a different story...

 
 
 
The Battery
 

When I was testing the completely new D800 in my home lab test, I was afraid that the battery life would be too short and I'd have to change battery often, which means opening the underwater housing, often in very unfriendly conditions (high waves, hitting a small boat, spraying water all around...). Luckily my fears were unnecessary. The EN-EL15 battery's capacity is just enough to last four long and demanding dives, where as I had to change the battery of my underwater flashlights after every one of those dives. The camera was securely locked in the underwater housing all day long and I only changed the battery in my hotel room in the evening.

 
File Size...
 

...is a problem. And not a small one. I've only noticed this when I returned home from my travels and brought home some thousands of technically and composition-wise successful photos (in RAW format). A typical 14-bit RAW file in FX format, losslessly compressed is about 38 to 42 MB and a JPG file in the highest quality is about 15 to 25 MB.

When I bought the D800, I thought I will always shoot macro photos in FX format and then crop them at  home in the computer to the desired composition. Today I'm not doing that any longer. We can only crop JPG photos, but the NEF files remain unchanged and each of them takes about 40 MB on our hard disk. That is the reason why I now switch to DX crop (15 to 18 MB in NEF format), every time I can't come close enough to a subject or the subject itself is too small, and so my files are smaller already before I step out of the water.

 
Finnal Thoughts
 

Despite some weaknesses, most noticeable among them still insufficiently solved problem of the cyan ring around the sun, the D800 is in my opinion at the time of writing also in reality (not just on paper) the best Nikon camera for underwater photography.

Its enormous resolution opens many cropping possibilities in macro photography and enables the 15 MP DX crop, which allows for capturing photos that are sharper in the corners, when shooting with a dome port.

Despite the high resolution its ISO features are more than satisfactory for most tasks in underwater photography (blue background in macro photography, deep wrecks, UV fluorescent photography).

The camera's greatest advantage, at least for me, is its autofocus system, which in my opinion is the greatest advance in quality when switching from the D2X to the D800.