Lens Sharpness and Contrast
There is a reason why Nikon charges $18K for this lens – there is simply no equivalent to such a lens in Nikkor’s super-telephoto line in terms of sharpness and contrast. No matter what combination you try, whether you couple the Nikon 400mm f/2.8 with the 2x teleconverter, the 500mm with the 1.7x TC or the 600mm with the 1.4x, none of them are even remotely capable of resolving what the 800mm can, as demonstrated later on in the review. And this is pretty evident when looking at the MTF chart of the lens:
If you do not know how to interpret the above chart, I wrote a pretty extensive article on how to read MTF charts. With such a flat-like MTF curve, this lens basically suits the definition of a “perfect lens”. In comparison, even the superb Nikon 300mm f/2.8G VR II does not show the same type of optical performance. I don’t know if it is the design of the lens or the fluorite elements that are the reason for this, but according to the manufacturer, there is nothing from Nikon that performs close to what the 800mm f/5.6 is capable of. And I do not know of any other lens made to date by lens manufacturers that can make such a claim for lens performance. When I initially saw this chart, I knew that the Nikkor 800mm would be something very special, something we have never seen before. So when the lens finally arrived, I could not wait to test its MTF performance in my Imatest lab. Even before I started testing the lens, I had very high expectations, since the sharpness and contrast I saw in images from Yellowstone just blew me away. Whether I was shooting with the Nikon D4 or the high-resolution D800E, the lens delivered crisp details at the pixel level. Just take a look at the below shot of a great gray owl, captured by Tom on a cloudy day:
And then take a look at a 100% crop of the above shot:
The details speak for themselves – each individual feather is distinguishable, which is what we photographers commonly refer to as “tack sharp”. The image was captured at 1/320th of a second @ f/5.6. The owl was far enough to comfortably shoot at f/5.6, but note the slow shutter speed of 1/320 – the lens was mounted on a tripod, and VR was turned on to prevent camera shake.
However, the moment I put together my lab and started shooting, I realized that something was very wrong – I was not getting results anywhere close to my expectations. Results were coming out inconsistent and sharpness varied greatly from shot to shot. At best, the results I got were about twice as bad as I got from the 300mm f/2.8 lens. How could that be? When I opened images and viewed them at 100%, it did not take very long for me to realize that camera shake was the sole reason for such bad performance. And that’s despite the fact that I had the lens set up on a very sturdy ballhead mounted on a big Gitzo tripod. Believe it or not, but I later spent about two weeks trying to develop a methodology for testing such long lenses. It was a very challenging task and after I was done, I understood why most review sites never publish their results from testing long lenses, or do not even bother testing them in the first place. And I can tell you that anyone who claims to have been actually able to quantify the performance of such long lenses without using a monstrous optical bench (which does not exist as far as I know) probably has bad / wrong results. I do not want to make any accusations, but let me just state out a simple fact – at 300mm and higher, even the slightest micro-vibrations result in enough blur to invalidate the results. You might not see it in the image even when examining closely at 100%, but the software does. Those micro-vibrations multiply like crazy as the focal length of the lens is increased. At 400mm or more, it becomes an impossible task to measure sharpness with a typical continuous light setup in a lab environment. And it does not matter how stable your tripod and the tripod head are – even the most expensive setups will create too much vibration. One would need a custom rail system with a plate that spans across the lens extending to the camera, with special vibration-absorption mechanisms to reduce vibrations. And even then it would probably not be enough to accurately measure such long lenses, since the setup would still most likely not be stable enough. One possible solution to this would be to do measurements on a camera that has no mirrors and no shutter mechanism. However, Imatest scores are tied to both the camera and the lens, so if I were to swap the camera body, it would invalidate all other lens measurements. In short, it is simply impossible to quantify lens performance with a traditional continuous light setup, no matter how bright the lights are. The method I developed turned out to be quite complex and involves flash – the only method in my opinion that would yield more or less accurate results.
Nikon 800mm f/5.6E VR MTF Performance
Once I figured it all out with the proper testing methodology, I had to re-test everything from scratch. It took me an additional week of lab testing, which is why this review took me so long to write. But I am proud of my accomplishments, because I can confidently measure super telephoto lens performance and I have been able to prepare all the data I need for the 200-400mm, 300mm, 500mm and 600mm lenses – with reviews to be posted soon after this one.
Here is the final MTF result from Imatest:
As you can see, the performance of the Nikkor 800mm f/5.6E VR is incredible wide open. There are some wide-angle and telephoto lenses that can reach this level of sharpness in the center, but none can truly match such high performance throughout the frame. The mid-frame is a tad worse than the center and the corners are outstanding.
One thing that you absolutely have to watch out for when using such long lenses is thermal distortion, also known as “atmospheric haze” or “heat distortion”. Those who shoot with 600mm lenses know this very well, since they see it quite a bit too, especially when using teleconverters. I forgot about thermal distortion during the first day of shooting and I was getting frustrated when my photos lacked details and looked blurry. At first, I thought it was my bad hand-holding technique. Then I tried everything from increasing my shutter speed to setting the lens on a tripod. Nothing changed. Then I thought that perhaps the lens needed to be tuned for my camera. I played with AF Fine Tune a bit and again, nothing improved my images. I had no issues with sharpness at close distances, so I knew that I could not blame it on a badly calibrated setup. And this was all happening on a hot sunny summer day in Yellowstone!
When we spotted an Osprey nest a mile away from the road later in the day, we stopped and I decided to experiment with the Nikon teleconverters (read more on teleconverters below). I attached the Nikon TC-17E II to the lens and I was able to actually see the effect of thermal distortion through the viewfinder. The heat was not coming from the ground, because there was a deep canyon between where I stood and where the osprey nest was – it was in the air! And when I tried the Nikon TC-20E III on the lens, the thermal distortion was even more obvious at 1600mm! Just take a look at what the image looked like when I captured it:
Every image that I captured varied in blur, because the heat waves were moving all over the place. Obviously, all images were a blurry mess, but it reminded me very well about what thermal distortion can do to images. So keep this in mind when shooting with such lens. At 800mm and 2° angle of view, the magnification is so high, that it literally fools you to think that you should be able to get sharp photos every time. When you look through the viewfinder, it looks perfect. And then you take a shot and view at 100%, only to see that it was not. After realizing this, I stopped blaming the gear and knew that it was totally my fault for not thinking about it in the first place.
One of the strengths of long telephoto lenses is the beautiful, creamy bokeh they are able to produce. On top of that, due to the shallow depth of field, subjects can be isolated very effectively, resulting in very smooth backgrounds. As with all other super telephoto lenses, bokeh on the Nikon 800mm f/5.6 VR is remarkable. Background highlights are rendered very smoothly and have no defined edges that can look distracting. The quality of blur is very similar to what one would get with exotic 300mm, 400mm, 500mm and 600mm lenses. When photographing with this lens, I only stopped it down in order to increase depth of field and even at f/11, the lens produced very smooth bokeh.
Vignetting / Light Falloff
There is a little bit of vignetting present at f/5.6, but it is not distracting as it can be on super wide angle lenses and most other telephotos. Once you stop down to f/8 and smaller, vignetting is practically absent. Take a look at the following results produced by Imatest:
And here is the visualization of the worst case scenario:
Ghosting and Flare
The large and long hood on the Nikon 800mm f/5.6 VR is there for a reason. Most long-range telephoto lenses, including the 800mm, do not perform well when shot against a bright light source. So, if you shoot against the sun, you might get some large, nasty flares and plenty of ghosting, which is quite normal. The integrated “Nano Crystal Coat” certainly helps in reducing ghosting and flare, but does not eliminate it in any way. In fact, Nano coating is mostly good for increasing contrast on telephoto lenses.
Distortion is practically non-existent on this lens. If you put up straight lines on the wall and photograph them, you might see a very minimal amount of pincushion distortion, which is normal. And it is even less noticeable on cropped-sensor cameras. Distortion is generally not a problem, because it can be easily fixed in Photoshop or Lightroom using the Lens Corrections module.
Imatest measured a very slight amount of pincushion distortion, at -0.57.
As I have already pointed out at the beginning of this review, the Nikon 800mm f/5.6E VR comes with the latest generation Vibration Reduction system that Nikon claims allows shooting with up to 4.5 times slower shutter speeds while retaining sharpness. This VR system is essentially the same as VR III used on the Nikon 70-200mm f/4G VR, except Nikon claims 4.5 stops versus 5. I believe the reason why the lens is rated half a stop worse than VR on the 70-200mm f/4, is because of the focal length of the lens. Again, at 800mm, the lens is very prone to a lot of camera shake, so VR can only do so much to compensate for it. Still, 4.5 stops of compensation is a lot! In my field tests, I got very good results at 1/250-1/500 of a second, with mostly keepers. So for up to two stops less than the focal length of the lens, you can shoot with confidence, even hand-held. Shooting at slower shutter speeds is quite difficult due to camera shake and the number of keepers obviously goes down quite a bit. Still, if you mount the lens on a Gimbal head and use a good lens holding technique, you could get sharp photos at much slower shutter speeds like 1/60. So in my experience, VR can be effective up to 4 stops max. Obviously, adding teleconverters will certainly affect VR’s effectiveness. As I have pointed out earlier, it is very difficult to control all the micro-vibrations at extreme focal lengths.
The lens will automatically detect if it is on a tripod and will compensate for the camera shake accordingly. I mostly shot with Vibration Reduction set to “Normal”, but also turned it off when using very fast shutter speeds. See this article to understand why you should first let the lens stabilize when shooting with VR on and why you should turn VR off when shooting at very fast shutter speeds.
The above image is as-is, without any cropping applied. The bear was so close, that I was able to fill the frame with just its head at 800mm. Hand-held, VR turned off, 1/1600, f/8.
Performance with the Nikon TC800-1.25E ED teleconverter
The Nikon 800mm f/5.6 is the first lens that is shipped together with a teleconverter. The Nikon TC800-1.25E ED is a 1.25x teleconverter build specifically for this lens. I tried to mount it on the 500mm f/4G VR lens as an experiment and it would not mount, so it will only work with the 800mm lens. At 1.25x, it will increase the focal length of the lens to 1000mm, which is insane! Obviously you do lose some light with the teleconverter, so the maximum aperture decreases from f/5.6 to f/7.1. Still, 1000mm is a lot of focal length! The TC800-1.25E is so good optically, that you lose very little sharpness with this combo. However, your shutter speed, lens handling technique and thermal distortion are even more important at 1000mm, so keep that in mind. I used the Nikon D3s and D4 with the 1.25x TC and while autofocus seemed to work well on both cameras in good lighting conditions, the Nikon D4 seemed to be better in AF accuracy. I was not surprised to see this, since the D4 comes with a superior “Advanced Multi-CAM 3500FX” AF system, which works with lenses up to f/8 maximum aperture.
Here is how Imatest measured the MTF performance of the lens, when the TC800-1.25E ED teleconverter is mounted on it:
There is definitely a drop in sharpness, but the resolution figures still stay above 3000, which is very good. Looks like the teleconverter hurts the performance of the lens in the corners the most.
Performance with the Nikon TC-14E II teleconverter
I also used the Nikon TC-14E II (1.4x) teleconverter quite a bit. The lens performed well optically, but AF was less reliable than with the 1.25x TC, especially on the D3s. With the 1.4x TC, it is a 1120mm f/8 lens, so I would only recommend to use this combo with the latest generation Nikon DSLRs that can handle AF with f/8 lenses. Let’s take a look at how the lens performed with the TC-14E II:
There is definitely a bigger drop in sharpness with the TC-14E II. The wide-open performance of the lens drops a bit, with f/11 producing the best overall sharpness.
Performance with the Nikon TC-17E II teleconverter
The Nikon TC-17E II (1.7x) went on the lens once, because it was difficult to work with. At 1360mm f/9.5, it was not only difficult to get AF to work reliably (even on the D4), but it was also too difficult to properly handle the setup. Even slight vibrations are visible at 1360mm and the sharpness impact seems to be too great. Here are the Imatest MTF measurement results:
I would not recommend to use this lens with the TC-17E II, unless you absolutely need the reach.
Performance with the Nikon TC-20E III teleconverter
At 1600mm f/11, this combo is just not practical for the same reasons. Autofocus will fail, even when shooting in daylight conditions. Sharpness is also impacted greatly, as can be seen from the below chart:
Still, if you are after the most reach and you are OK with trying to adjust the focus ring manually, you might be pleased with what this lens can achieve at 1600mm. At such long focal length, you will be battling more with vibrations than with the focus! Even touching the lens will shake everything really badly. I experimented with the TC-20E III for fun when photographing the full moon and here is what I was able to capture:
If you would like to see the moon in its full glory in wallpaper resolution, download this file and check it out! You can even see the stars behind the moon. Another shot of the moon in waning gibbous state can be seen at the end of the review.
Let’s now move on to the good stuff – Comparisons to other super-telephoto lenses.