Lenses for Nikon D800

 
(December 2012)
 

The Nikon D800 demands the use of top-notch lenses. In order to achieve the best possible results, I decided to systematically test all of my lenses (including DX) and some that I borrowed. I was interested only in their sharpness, since other optical errors like chromatic aberration, distortion and vignetting are relatively easy to fix with photo-editing programs, especially with RAW-converters (such as Lightroom, Capture NX etc.).

When I used to shoot with the D3, many of my lenses were very sharp, but with the D800 the situation changed considerably... Not to say that the lenses became unsharp, but they are sharp in different ways, whereas they were almost equally sharp while shooting with the D3.

The testing methodology was as follows: I took photos of the test card while using for illumination only two remote controlled flashlights (2x SB-800) and thus eliminating the effect of camera shake. I used Live view mode to focus. When analysing the photos on the computer screen, I took samples from nine points of the test card (in which the lines are oriented in different directions) and entered the results (at which point the lines subjectively merge together into greyness) into an Excel spread sheet. I gradually joined the acquired data. In the first phase I gathered three readings from every of the nine points: sharpness in the middle, on the border and in the corner. When I joined this information I received a grade for the entire photo at a certain f-stop value. The average of these grades from the biggest f-stop till f/11  gave me an average grade for the lens or for a certain focal length on a zoom lens. I did not consider more stopped-down apertures as the photo quality becomes more or less similar for all lenses due to diffraction. For zoom lenses I also calculated an average grade for all focal lengths, which represents the overall grade of the lens. All results were graphically presented in Excel charts, which enabled a clear reading.
 
 

Centre = (Ah + Av) / 2
Border = (Bd + Bv + Cv + Cd) / 4
Corner = (Dh + Dv + Dd) / 3

I am aware that this grading system for lens sharpness is quite subjective, but this is the only way of creating an overview of the mass of information available. I can confidently claim that my grades are quite realistic, since the sharpness of photos taken at a later point on my travels coincides well with the grades given to specific lenses.

The results I got were mostly in line with my expectations, yet I was sometimes surprised by them. What I've found most intriguing is that there were little correlations between my results and the official MTF curves published on the Nikon homepage. Nikon only shows results achieved with completely opened apertures, while I graded the lenses at a range of aperture values and calculated an average grade based on them. Different lenses achieve picture quality by stopping-down the aperture in different ways, so the information about sharpness at a fully opened aperture does not provide a realistic assessment of lens quality at aperture values, which are most commonly used.

My impression while testing was that the difference between two different lenses of the same type can sometimes be even greater than between lenses of different types. Therefore exact testing results will not be published, since these results are only valid for my specific lenses and I do not wish for them to become a general guideline.

It is also interesting to see how sharpness profiles for lenses can differ from one to another. Some lenses are incredibly sharp in the centre, while being weak in the corners (a typical example is the AF-S Nikkor 24-70mm f/2.8G ED), while others are equally sharp from the centre to the corners (a great example is the AF-S Micro Nikkor 60mm f/2.8G ED).
 
 
An example of a lens, that is very sharp in the middle and soft in the corners (AF-S Nikkor 24-70mm f/2.8G ED @ 35 mm)
 
 

An example of a lens, which is equally sharp in the centre as well as in the corners (AF-S Micro Nikkor 60mm f/2.8G ED)

 
In one case, with the AFS Micro-Nikkor 60mm f/2.8 lens, I've unfortunately discovered that my specific lens is an extremely bad one (very sharp in the centre and very unsharp in the corners!). Of course I couldn't sell such a lens, but instead gave it as a gift to by daughter, who is using the D7000, while I took a deep breath and bought myself a new one! The new lens is incomparably sharper, so to say "ideally" sharp. Considering that I've been using this lens almost exclusively for underwater photography with the D2X, I've never even noticed this unsharpness in the corners and I've always thought of it as one of my best lenses!
 
 
Results (FX)
 

My sharpest lens is the AF-S VR Micro-Nikkor 105mm f/2.8G IF-ED with a very broad range of most useful aperture values between f/4 and f/11, however I must note that the central area is noticeably sharper that the corners.

Surprisingly, my second sharpest lens is the "cheap Chinese" AF-S Nikkor 50mm f/1.4G, which is at its sharpest in the central area at f-stops between f/2.8 and f/5.6, and in the extreme corners between f/5.6 and f/8.

It is closely followed by my new AF-S Micro Nikkor 60mm f/2.8G ED, which gives the sharpest pictures between f/5.6 and f/11. The sharpness distribution is incredibly even from the corners to the central area. My old AF-S Micro Nikkor 60mm f/2.8G ED lens took the last place due to the before mentioned problem. It produced the sharpest pictures between f/5.6 and f/11, the corners however were only satisfactory sharp at f/16. At f/2.8 it was almost twice sharper in the central area than it was in the corners!

Due to the incredible sharpness of the AF-S Nikkor 50mm f/1.4G, I've recently deeded to buy two more "cheap Chinese" lenses: the AF-S Nikkor 85mm f/1.8G and the AF-S Nikkor 28mm f/1.8G. At first I borrowed the AF-S Nikkor 85mm f/1.8G from Nikon Slovenia and I was very impressed by the results. When I bought my own one, I noticed that it's sadly slightly worse than the one I tested, but it still ended up on place four with the greatest sharpness in the central area between aperture values f/4 and f/11 and the best sharpness in the corners at f/11. The lens I borrowed was slightly sharper in the corners, comparing to my lens. But unfortunately I have to report another problem about this lens, which I've only noticed later in everyday use: the lens has an extremely strong back focus, which I had to correct with value -15. I haven't noticed this while testing, since I only focused in the Live view mode.

The AF-S Nikkor 28mm f/1.8G has slightly worse results than the 50/1.4 and the 85/1.8, which was to be expected due to the sophisticated construction of wide-angle lenses, yet it is still sharper than most of the tested zoom lenses. It showed good sharpness in the central area in the broad range between f/2.8 and f/11, and it gives the sharpest results in the corners at f/11.

Even though Nikon does not list the before mentioned "plastic Chinese" lenses in their recommendations for the D800, I believe that every D800 user should consider them, since they cost little, yet their optical quality is very high and probably not lagging far behind the recommended "Made in Japan" Holy Trinity (24/1.4; 35/1.4 in 85/1.4).

Considering lenses made by other producers, the highest rank goes to my Sigma 150mm F2.8 EX APO DG HSM Macro (before the 28/1.8 and after the 105/2.8, 85/1.8, 60/2.8, and the 50/1.4). This lens achieves peak sharpness at f/11 in the central area as well as in the corners. However, I've noticed that the right border is slightly sharper than the left...

In my array of zoom lenses the sharpest by far is the AF-S Nikkor 70-200mm f/2.8G ED VR II. It is the sharpest at focal length 200 mm and slightly less at 70 mm. In the middle it's slightly weaker, but still extremely sharp. The greatest sharpness is achieved in the range between f/5.6 and f/11, depending on focal length, however at 70 mm great central sharpness can be achieved already at f/2.8. For portrait photography this lens is "indecently" sharp, as it shows every flaw on the skin, which is surely not in the interest of the model...

It is followed by the AF-S Nikkor 14-24mm f/2.8G ED and slight behind it the AF-S Nikkor 24-70mm f/2.8G ED. The 14-24 gives the sharpest results at 14 mm, then the sharpness is slightly decreasing and starts to increase again at 24 mm. The range of best results is between f/4 and f/11, again depending on the focal length.  With the 24-70 lens I was probably again unlucky, since my specific lens of this generally highly praised model is not something to brag about. The corner sharpness was quite weak at all focal lengths and aperture values, while the sharpness in the central area is extremely good. My lens has a very strong chromatic aberration, mostly at the wide end, which is unacceptable for a professional lens like this. The main difference between the 14-24 and the 24-70 lenses is (surprisingly) the fact that the sharpness distribution of the 14-24 is much more even from the central area towards the corners compared to the 24-70. Consequently, the 24-70 is sharper in the central area than the 14-24, but weaker in the corners. Thus the 14-24 is more suitable for landscape photography, while the 24-70 is better for reporting.

I borrowed the AF-S Nikkor 24-120mm f/4G ED VR lens from Nikon Slovenia and I was amazed by the test results. Even though the focal length range has a factor of five, it is extremely sharp in the range of 35 to 70 mm, even though the corners were noticeably weaker than the centre at all focal lengths with open apertures. Online reports about the sharpness of this lens are very mixed, so it seems that the lens I was testing was one of the better specimens.

My weakest zoom lens is the AF-S Nikkor 16-35mm f/4G ED VR, mostly due to the lowest focal lengths (16 to 20 mm), where the corners are very weak. The sharpness then improves and in the range between 28 and 35 mm it is very good, including corners. At 28 mm the corner sharpness is so good, that it's even slightly superior to the AF-S Nikkor 28mm f/1.8G in the corners! The best sharpness is achieved between f/5.6 and f/11, depending on focal length. I've noticed a certain amount of back focus with this lens while using it, but I could correct it with the value -7. Unfortunately, I use this lens much more often than the optically superior 14-24, due to the greater focal length range and the filter use possibilities, and I always carry it with me on my travels. Even though I marked it as "the worst", I want to emphasize that this is still a very sharp lens and definitely worth buying.

I've also tested the AF-S VR Nikkor 70-300mm f/4.5-5.6G IF-ED, which I borrowed, since I used to borrow it from time to time, when I was still shooting with the D3 because it was lighter than the 70-200/2.8. According to my expectations, it unfortunately turned out to be the worst of all tested zoom lenses and thus least suitable for the D800. The 70-200/2.8 with the teleconverters TC-14E, for example, is much sharper at 280 mm than the 70-300/4.5-5.6 at 300 mm!

In the array of FX lenses the 16mm f/2.8 Fisheye lens turned out to be the worst by far, especially due to the very weak borders and corners, where the sharpness is twice or even thrice worse than in the central area! Considering that all fisheye lenses have a similar sharpness profile, this is obviously a general feature of this kind of lenses.

 
 
Results (DX)
 

I've tested a smaller number of DX lenses comparing to FX lenses and most of them were borrowed. A direct comparison between FX and DX lenses is unfortunately impossible, since I've tested the FX lenses with all 36 MP, while I used a 15 MP crop on the DX lenses. Among the tested DX lenses the clear winner is the small and priceworthy AF-S DX Micro-Nikkor 40mm f/2.8G IF-ED, which shows the greatest sharpness in the range between f/4 and f/11. This lens, similarly to the 60mm f/2.8, distinguishes itself by the incredibly even sharpness distribution from the centre to the corners.

Among the DX zoom lenses, the AF-S DX NIKKOR 18-200mm f/3.5-5.6G ED VR II turned out to be incredibly sharp. It achieves the greatest sharpness between 18 and 24 mm, the sharpness then decreases slightly at 35 mm, but remains almost unchanged until 135 mm, then declines again around 200 mm. With shorter focal lengths the central area is sharp already at the most opened aperture, but the corners are noticeably weaker. The best compromise between central and corner sharpness can be achieved at almost all focal lengths in the range between f/8 and f/11. For comparison I've also decided to test the FX lens AF-S Nikkor 70-200mm f/2.8G ED VR II in the DX crop at 200 mm (in order to perform a direct comparison). The professional FX 70-200 is at 200 mm noticeably sharper than the DX 18-200 at any chosen focal length!

I've got almost equally good results with the borrowed AF-S DX NIKKOR 16-85mm f/3.5-5.6G ED VR, which is sharpest at 16 mm, then the sharpness decreases slightly, but remains steady almost till the end, where it declines sharply, similarly to the 18-200. In the range of 16 to 35 mm the sharpest f-stops are between f/8 to f/11, in the range of 50 to 85 mm only f/11 gives sharp results. The weakest corners can be seen at 16 mm and  85 mm, while the sharpness is quite evenly distributed from the centre to the borders in the central focal length range.

Next in line according to sharpness among DX lenses is the AF-S DX NIKKOR 10-24mm f/3.5-4.5G ED, which is surprisingly sharp already at 10 mm, however with very weak corners. Even at 12 mm the corners remain weak, but from 16 mm on the lens has very evenly distributed sharpness from the central area to the corners. The best f-stop at all focal lengths is f/11, except at 24 mm, where it's f/8.

The results of my old AF-S DX NIKKOR 12-24mm f/4G ED lens (which I gave as a gift to my daughter, after buying the 10-24) were surprisingly bad. Even though the 12-24 lens is considered as "a professional DX lens" (and the price is considerably high too!), its results are substantially worse than those of the 10-24 lens in my case, mostly due to very bad corner sharpness at all focal lengths. The corners reach satisfactory sharpness only with f-stops between f/11 and f/16, while sharpness in the central area is achieved already at f/5.6 to f/11, depending on focal length. The sharpness is similar at all focal lengths with a small decline at 12 and 20 mm.

Just like in the FX series, fisheye lenses take the last place also in the DX series, due to the distinctively lower sharpness in the borders and corners compared to the central area of the picture. I tested three fisheye lenses: Nikon 10.5mm f/2.8, Tokina 10-17mm f/3.6-4.5 and Sigma 10mm f/2.8.

The best one of them was the AF DX Fisheye-Nikkor 10.5mm f/2.8G ED. The central area was sharp in a broad range between f/2.8 and f/11, the borders were sharp between f/4 and f/11, while the corners were unsharp throughout the range of different aperture values. Generally, I've found this lens to be very good performer and it seemed much better than its 16 mm FX counterpart. The main problem of this lens is very strong chromatic aberration, which can be removed with the use of photo editing software.

I've tested the Tokina lens only at 10 mm. Despite the fact that this lens is greatly praised by underwater photographers, my Tokina gave noticeably worse results than my Nikon lens. The best sharpness is achieved at 10 mm in the central area with f-stop values between f/5.6 and f/11, but it never reaches results similar to those of the 10.5 mm Nikon.

The borrowed 10 mm Sigma gave the worst results. The best sharpness in the centre can be seen between f/8 and f/11, however the borders are disastrous, incomparably worse than what I've seen with the Nikon and Tokina lenses. Before starting these tests, I had borrowed a 15 mm f/2.8 Sigma (full frame). Already back then I remember being greatly disappointed by the quality of corner sharpness, since it was incomparable to that of the 16 mm Nikon. I wonder about it, since I've noticed that many professional underwater photographers are using the same Sigma 15 mm! Sigma's fisheye lenses most definitely didn't leave a good impression with me and I will avoid them in the future. Sigma's macro lenses however, are a whole different story, since their quality is comparable to Nikon lenses.

 
 
Teleconverters
 

I own all three Nikon teleconverters: TC-14E II, TC-17E II and TC-20E III. I've bought the TC-14E II and TC-20E II years ago. Even while shooting on film with the AF-S 80-200mm f/2.8 lens I noticed that the TC-20E II gives much worse results than the TC-14E II teleconverter. As soon as the TC-20E III was launched, I bought it as a replacement for the old one and indeed the new teleconverter is slightly sharper than the old one (when shooting with the D3 and a 70-200/2.8 lens it was decently sharp). I only bought the TC-17E II recently. I'm using all mentioned teleconverters only in combination with the AF-S VR Micro-Nikkor 105mm f/2.8G IF-ED and the AF-S Nikkor 70-200mm f/2.8G ED VR II lenses, since they are not compatible with my other lenses.

I've achieved fabulous test results with the TC-14E II teleconverter and I recommend it to any owner of compatible AF-S lenses. The 105mm f/2.8 lens has only a very slight loss of quality in combination with this teleconverter (effective focal length 150 mm). One can barely notice that the photo is slightly less sharp than with the 50/1.4, 85/1.8, and 60/2.8 lenses alone, whereas it's basically as sharp as the Sigma 150/2.8 (without teleconverter!). The greatest sharpness was achieved at f-stop values between f/5.6 and f/11.

I've seen even more astonishing results with the 70-200/2.8 lens at 200 mm (effective focal length 280 mm). The photo had practically no loss in sharpness in the central area, only the corners were slightly softer. Best f-stop values are between f/5.6 and f/11, however f/4 is already extremely sharp!

The 105/2.8 combined with the teleconverter TC-17E II (effective focal length 180 mm) is still quite sharp. However a strong chromatic aberration in the corners appears with this combination, which couldn't be detected while using the 105 + TC14. The greatest sharpness is achieved between f/8 and f/11.

With the 70-200/2.8 lens at 200 mm (effective focal length 340 mm) a decrease in sharpness is noticeable, however this combination is still sharper than the AF-S VR Nikkor 70-300mm f/4.5-5.6G IF-ED at any chosen focal length! The most disturbing by far are the soft corners with a very strong chromatic aberration. The best f-stop value is f/11.

In line with my expectations the TC-20E III turned out to be the worst teleconverter of the array, despite the fact that the newer version includes an aspheric element. With the 105/2.8 lens (effective focal length 210 mm) a substantial decline in sharpness of the central area as well as soft corners are clearly visible, however there is less chromatic aberration present than with the TC-17E II. The best f-stop values are between f/8 and f/16.

Even worse are the results of the TC-20E III with the 70-200/2.8 lens at 200 mm (effective focal length 400 mm) as this combination is comparable in quality with the worst zoom lens 70-300/4.5-5.6. The sharpness of the central area and especially in the corners is deteriorated strongly; the best f-stop values are between f/11 and f/16.

 
 
Macro lenses
 

Because of my focus on underwater photography, macro lenses are of extreme importance to me. They were also the reason why I bought all three teleconverters, since they enable me to increase the reproduction ratio by the factor of the teleconverter while working at the same shooting distance. Unfortunately there is a very high price to pay: the general sharpness of the system and the speed of the lens are reduced, which slows down (or even disables) the autofocus option.

The table below shows an overview of my macro lenses in combination with the mentioned teleconverters and their measured results at reproduction ratio 1:1.  While using the Nikon D800, one should consider its 15MP DX crop as a relevant factor for macro photography.
 
 

E.f.      = effective focal length of the system as noted in the EXIF file
W.d.     = working distance between the front border of the lens and the subject (in mm)
f ∞       = maximum f-stop of the system, when the lens is set to infinity
f 1:1     = maximum f-stop of the system, when the lens is set to the shortest focusing distance
FX cr.   = object size (measured in mm), when the lens is set to the shortest focusing distance using FX format
DX cr. = object size (measured in mm), when the lens is set to the shortest focusing distance using DX format

The main problem of the 60mm f/2.8 is the short working distance, therefore it is more suitable for shooting larger objects from a slightly bigger distance. As we can see, the lens is focusing slightly beyond the reproduction ratio 1:1 (portraying 35 mm instead of 36 mm!).

Just like in the good old film days, the 105mm f/2.8 is now again my "working horse" for underwater macro photography (in "DX times" it used to be the 60mm f/2.8). This lens is also focusing slightly beyond the reproduction ratio 1:1 (again portraying 35 mm instead of 36 mm!). The autofocus is extremely fast and the D800 controls it without problems (the D2X often had problems with this lens around the reproduction ratio 1:1 and had to refocus many times).

The 105/2.8 & TC14 combination is a great fit and should definitely be considered for underwater macro photography. The sharpness is on a good level and the autofocus is still quite fast and reliable. The user manual for the 105/2.8 lens states, that the autofocus is not available while using teleconverters, however that was written before the D800 and D4 were launched, when cameras were only able to focus till the f-stop value of f/5.6.

The 105/2.8 & TC17 combination is noticeably worse. Not so much because of a decline in sharpness, but rather due to the autofocus, which still works around the reproduction ratio 1:1, alas is not reliable and needs to refocus many times.

In my opinion, the 105/2.8 & TC20 combination is not suitable for underwater macro photography. The autofocus does not work around the reproduction ratio 1:1 (even though it works reasonably well on longer distances). Since I was forced to focus manually, I did not shoot a single photo with a satisfactory sharpness (on a 36 MP sensor)!

The Sigma 150/2.8 is brilliantly sharp, has a broad working distance but a slow autofocus and because of this slowness, it was much more reliable on the Nikon D2X, than it was the 105/28. With my Nikon D800 I now prefer to use the Nikon 105/2.8 & TC14 combination rather than the Sigma 150/2.8, considering that the sharpness is almost equally good, but I gain a 1.4x magnification. The autofocus works fast and is reliable in the normal range, only around the reproduction ratio 1:1 it becomes slightly less reliable.

 
Photos of a ruler with a millimetre scale at the shortest focusing distance with full FX and DX format crop.
 
 
Some examples
 

In the next paragraphs I will show some crops of the test card (in all cases the central area and the upper right corner), which show how much the lenses can differ between each other when it comes to sharpness. Since I only wanted to present the lens sharpness (and chromatic aberration, when it's so strong that the cameras own algorithms did not manage to remove it), I published jpg pictures as they were recorded by the camera (without additional sharpening), but with corrected levels, so that all photos are equally bright. Had I not corrected the brightness, the corner crops would have been much darker due to the vignetting effect of the lenses.

First here are the results of NIKKOR 14-24mm f/2.8G ED, NIKKOR 24-70mm f/2.8G ED, NIKKOR 24-120mm f/4G ED and NIKKOR 16-35mm f/4G ED lenses at focal length 24mm and at f-stop value f/5.6:

 
 

As we can see, the 14-24/2.8 lens at 24 mm shows incredibly even sharpness distribution from the central area to the most extreme corners. The 16-35/4 lens at 24 mm is slightly worse than the 14-24/2.8 in the central area, and much worse in the corners. The 24-70/2.8 lens at 24 mm is by far the sharpest of all tested lenses in the central area, but considerably less sharp in the corners. In addition to that it also displays strong chromatic aberration, which is not visible with the previous two lenses. The 24-120/4 has a similar sharpness profile as the 24-70/2.8 at 24 mm, except for being slightly less sharp in the central area, while the corners are very similar.

Unfortunately I do not possess my former 24/2.8 and 17-35/2.8 lenses anymore, since I would be very interested in their results in this test with the Nikon D800 (since they were noticeably less sharp than 14-24/2.8 and 24-70/2.8 at 24 mm on the 12 MP D3 camera...).

The next comparison is at focal length 28 mm at f-stop f/5.6 between the following lenses: AF-S NIKKOR 28mm f/1.8G, AF-S NIKKOR 16-35mm f/4G ED VR, at AF-S NIKKOR 24-70mm f/2.8G ED.

 
 

As we can see, the 28/1.8 lens is actually not much superior to the zoom lenses. At f-stop f/5.6 it's reasonably sharp in the central area and quite soft in the corners. The 16-35/4 lens is slightly less sharp in the central area compared to the 28/1.8, however it is surprisingly sharper in the corners! With the 24-70/2.8 lens a similar sharpness profile can be seen at 28 mm as with 24 mm. The lens is extremely sharp in the centre, even sharper than the 28/1.8 lens, however it is quite weak in the corners and displays strong chromatic aberration.

I have compared the following lenses at 50 mm and f-stop f/5.6: AF-S NIKKOR 50mm f/1.4G, AF-S NIKKOR 24-70mm f/2.8G ED, and AF-S NIKKOR 24-120mm f/4G ED VR.
 
 

Clearly the 50/1.4 lens is the absolute winner in this category. Even though there is practically no difference between all three lenses in the central area, the difference in corner sharpness is enormous. Regarding corner sharpness, the 50/1.4 lens is one of the sharpest lenses I've tested and consequently much sharper than both zoom lenses at 50 mm. The worst results were, surprisingly, made by the 24-70/2.8 lens, which displayed a worse corner sharpness at this focal length than the 24-120/4 and included strong chromatic aberration.

At 85 mm and f-stop f/5.6 I've only compared the AF-S NIKKOR 85mm f/1.8G and the AF-S NIKKOR 70-200mm f/2.8G ED VR II.
 
 

Despite the fact that the 70-200/2.8 is my sharpest zoom lens, the 85/1.8 is still a clear winner. In the central area it's slightly sharper, but the difference in the corners is obvious.

In the end I compared the AF-S NIKKOR 70-200mm f/2.8G ED VR II at 200 mm, the same lens in combination with the teleconverter TC14E II at 200 mm (effective focal length 280 mm) and the AF-S VR NIKKOR 70-300mm f/4.5-5.6G IF-ED at 300 mm, always using f-stop f/5.6:

 
 
As we can see, the 70-200/2.8 with the teleconverter TC14 does not lose even a single bit of sharpness in the central area! There is some loss of sharpness in the corners, however it is very low. An entirely different story is the 70-300/4.5-5.6 at 300 mm, which gave visibly worse results in the central area and even more so in the corners!