Camera Labs

Now, what about the behavior in strong contra-light and with specular highlights/reflections on water?

I've uploaded two shots which really test the capabilities of this lens and normally produce purple fringing (see a really bad example of this effect from a Tokina 80-400 lens here).
Let's go for the reflection example first.

This shot was done at 70mm, f/2.8, 1/8000 sec, ISO 200, -1EV. If you click on the images you get straight to the flickr-page of the large original.

I'm showing you two 100% crops, one from the center, one from the foreground. And each crop is displayed in two versions: The first is processed normally (standard settings in CaptureNX), the 2nd version with loCA-suppression dialed in at 50%. I'll come back to that in a second, just watch for the red/blue outlines of the specular highlights in both versions:



Now the second crop:


Well, what's "loCA-suppression"? loCAs are longitudinal color aberrations that are purple in the foreground and blue/green in the background (see a really bad example of this effect from a Nikon 180/2.8 lens here). CaptureNX has a switch (normally set to OFF) that makes the software try to eliminate this special form of CAs that are harder to kill than the "normal" (lateral) CAs. As you can see, it is quite effective in this case.
Btw.: The large original was saved w/o loCA-suppression.

Here's the second shot, trees against the winter sky. The sun was barely veiled by some thin clouds and made for the ultra-bright background that makes these shots so challenging*. Shown is a 100% crop from the twigs directly in front of the sun. Shot at 70mm, f/4.0, 1/8000 sec, ISO 200, -1EV. The first version developed from RAW at standard settings with CaptureNX2. The second version of same crop with loCA-suppression dialed in at 50% (the large original that you can reach through the images is w/o loCA-suppression):





Watch for the red cast around the twigs.

----
*Don't try this for yourself: It may be dangerous for your health and for your camera!

Now finally I found a real ugly example of non-curable loCAs with this new lens. See the following 100% crop:


Shot at 150mm, f/2.8, 1/2000 sec, ISO 200, +/-0EV. Standard development in CaptureNX2 plus loCA-suppression at 50%. In this case the CA removal tools in CaptureNX2 did nothing to improve the situation.

Do I care? Well, not really...

The effects seen in the last post should not normally be so clearly visible. But, who knows? I'll keep my eyes peeled for any occurrence of this optical nastiness!
---
Back from test-shots here's something without Siemens-stars, purple fringing or loCAs: Our dog Ira hunting for that elusive mouse under the wintry tree:

Shot at 116mm, f/2.8, 1/250 sec, handheld (VR=ON/normal), ISO 200, -1EV, standard CaptureNX2 development.

I've uploaded the 100% quality full-resolution image, which you can reach by clicking on the image and then choosing "all sizes". Look at her silky hair - beautiful

Two other images here uploaded in full resolution (click through) shot at 116mm f/5.6 (developed as "vivid") and 200mm f/2.8 (converted to b&w):

The Flying Bike:


The Ladder:

Back to Siemens-star test-targets and a comparison to the fabled Sigma 150/2.8 macro. First image is the New Nikon zoom, second row the Sigma fixed focal. I'm showing 100% crops here of a DX corner (around 12mm from center = 58% of FX diagonal).

Nikon AF-S VR 70-200mm 2.8G ED II:


Sigma AF 150mm 2.8 EX DG APO HSM IF macro:


Developed in CaptureNX with standard settings plus exposure compensation to bring all images to a comparable level of brightness (around +1..+1.7EV). These shots were taken with hand-optimized focus with Camera Control pro2.

Looking at those images shows that the new offspring from Nikon-Labs is a little behind the Sigma lens in sharpness wide open but comes pretty close at f/5.6 and beyond. This is quite an achievement for a zoom and also shows that this new Nikon zoom is as close to a dedicated fixed focal as you probably can get.
Very impressive indeed!

Now the comparison at 105mm focal length. First image is the new Nikon zoom, second row the micro-Nikkor. I'm showing 100% crops here of a DX corner (around 13mm from center = 62% of FX diagonal).

Nikon AF-S VR 70-200mm 2.8G ED II:


micro-Nikkor AF-S VR 105mm 2.8G IF-ED:


The same conditions apply as in the previous post.

Again the Nikon zoom shows lower contrast than the micro-Nikkor at f/2.8, improves at f/4.0 and is almost indistinguishable at f/5.6 and beyond.

Let's continue with the same sequence of shots at 70mm, the first row as is and the second row with the Nikon tele-converter TC-17EII attached and shot from a 1.7x larger distance to end up with equal magnification. This 1.7x TC converts the new lens into a stabilized 120-340mm f/4.8 tele-zoom - something you could compare to the Sigma AF 120-400mm 4.5-5.6 DG APO HSM OS which I reviewed here.

At 70mm w/o TC:


At 70mm with TC = 120mm:


All images developed in CaptureNX 2 at standard-settings, only exposure compensation was used to match brightness. Shot with remote SB800. Focus was achieved with hand-optimization in Camera Control pro 2 based on contrast based AF in live-view.

Without TC (1st row) it looks like the performance at 105mm is repeated here at 70mm: fully open there is some slight "bleeding" of light from the left into dark areas. Thus contrast on fine structures is a bit reduced although sharpness looks quite ok (again this is at 12mm from the sensor-center).
With the TC attached (2nd row) this deficiency is magnified and produces an even stronger loss of contrast (see the letters at the bottom). Interestingly the situation can be improved markedly by stopping down only 1/2 stop to f/5.6. Still, even stopped down the images with TC cannot match those w/o TC. I guess the loss of resolution is around 20-30% here.

The big Q now is: How do images from the lens w/o TC compare to images with TC that were shot from the same distance and uprezzed in post-processing by 1.7x?

But this is a story for another post...

Further testing with the TC-17EII has to be postponed as my current setup only allows for max 150mm focal length testing at around 5m distance
But I can give you some real photos taken outside. This time at 340mm f/4.8 handheld at 1/250 sec and distances below 2m.

The first shot is a crop (presented here at 37%) I called Snow Green:

Click through to the 100% crop.

The other I called Fade to White:

Click through to the large original.

When judging these images remember that the test-shots at 70mm revealed that even stopping down 1/2 stop might already improve IQ visibly.

Now the final focal length to test: 200mm. This required some modification of the test setup as the distance needed to remain at the 1:33 magnification of the other test-shots was considerably greater than was possible in the other room: 6.8m.
I've added the other test-shots at 150mm, 105mm and 70mm for convenient viewing below. Here is the complete collection:

200mm:


150mm:


105mm:


70mm:


All in all I'd say: An impressive performance across the whole zoom range. Stepping down to f/4.0 helps just a bit.

At 200mm the only comparison I could do is against the venerable Nikon 18-200 VR. Below you find the results shot and developed under the same conditions as with the 70-200 VRII. Remember though that this is the only DX lens used for comparison. On a DX lens 12mm off center is equivalent to the border of the longer side, only 2mm from the corners while on FX it is only 60% of the diagonal. So don't be surprised, how bad the results look.

Nikon AF-S 18-200mm f/3.5-5.6 VR at 200mm:

After showing you the zoom in combination with a TC-17EII at 70mm focal length (=120mm with TC) I can now show you the results at the long end: 200mm x 1.7 = 340mm. It was extremely challenging to get decent results at these settings, but here they are:

Nikon AF-S VR 70-200mm 2.8G ED II at 200mm plus TC-17EII (=340mm):


Compared to the Sigma 300mm 4.0D Apo macro:


You can see that fully open the Nikon zoom+TC combo produced disappointing results - and believe me I tried many times to get the best shots. But even stopping down 1/2 stop makes the outcome very usable.
Although the fixed focal is ahead in IQ from f/5.6 onwards.

Now the final series with a TC-17EII at 120mm focal length (=200mm with TC)

Nikon AF-S VR 70-200mm 2.8G ED II at 120mm plus TC-17EII = 200mm:


To put that performance into perspective I've the results from the
Nikon AF-S VR 18-200mm 3.5-5.6G ED at 200mm:


So the combo performs roughly on a par with Nikon's super-zoom.
Stop down to f/5.6 to get usable results, better even to f/8.0.

Next up is a collection of all test-shots with the Nikon TC-17EII tele-converter.

Why do I show focal lengths here with TC (120mm, 200mm) that you could easily reach without TC - and with obviously better image quality? Well, changing a TC is no quick task. So there's a good chance that you leave the TC mounted when a photographic opportunity presents itself.
And it's always better to know the limits of your tools.
So here's the overview:

Nikon AF-S VR 70-200mm 2.8G ED II at 70mm plus TC-17EII = 120mm:


Nikon AF-S VR 70-200mm 2.8G ED II at 120mm plus TC-17EII = 200mm):


Nikon AF-S VR 70-200mm 2.8G ED II at 200mm plus TC-17EII =340mm:


My recommendation for using the TC is: stop down to f/5.6 when shooting at the long or short end and to f/8.0 for focal lengths in between.
But if you don't really need the extra reach, leave the TC in the bag: Nikon's new midrange tele-zoom is so sharp and contrasty (w/o TC) even wide open that you could easily crop if you need a closer framing. And don't forget: losing two or three stops between f/2.8 and f/5.6 or f/8.0 will require shorter exposure times if you don't want to risk shake and reduce subject isolation due to larger dof.

Coming back to the quality of combining the new zoom with a tele-converter I have one last comparisons for you to put the performance of the combo 70-200 VRII + TC1.7 into perspective:
I've uprezzed a series of shots with the Nikon 70-200 VRII at 200mm from the same distance as the shots with the TC at 340mm by a factor 1.7x in post-processing. Compare it to the last row in the previous post and it is perfectly clear, that uprezzing is the inferior alternative to using the TC.

Nikon AF-S VR 70-200mm 2.8G ED II at 200mm uprezzed 1.7x in post-processing: