This article is the first in a “versus” series. Since I for the moment have two cameras with almost identical image quality, the D90 and the D300 (if you are not convinced, see the first part of my D90 review), it is relatively easy for me to do comparision tests with different lenses. The first contenders - the new Nikkor 18-105VR on a D90 compared to the Nikkor 18-200VR superzoom on the D300.
The 18-105VR is Nikons newest addition to the ever expanding 18-xxx line. I thought it might be interesting to compare it too the 18-200VR (which I have reviewed before, see here). I want to point out that this is not about the cameras (they are reviewed here and here), but I decided to keep the same lens/body combinations during the test if someone does chooses between these two particular sets.
So lets weigh in the contenders. The D90/18-105VR combo weighs 1,152 g.
So it is easy to do the math. If we switch lenses a D90/18-200VR weighs 1,324 g and a D300/18-105VR weighs 1,375 g.
The difference in size between the lenses is not as big as one would expect, given the weight difference. However the 18-200VR is more front heavy, giving it a slightly less optimal balance, at least on the small D90. In the photo below we see one of the signs that the 18-200VR is a more expensive lens - the steel bajonet. How much durability suffers with the plastic mount on the 18-105, I cant really tell. But surely a metal mount inspires more confidence.
But, the plastic mount aside, the build of the 18-105 feels rather good. The mechanics are smooth, especially the zoom action. The autofocus is the more advanced version with full time manual override, in contrast to the cheaper micro AF-S found on lenses like the 18-55 and the 55-200.
VR and AF
The 18-105VR has the simpler kind of stabilization, VR I, while the 18-200VR has a more advanced version, VR II. In practical use I could not find any consistent difference, the VR seems equally effective on both lenses. It is possible that the softer shutter/mirror action of the D90 cancelled out the VR II advantage of the D300/18-200 combo.
These samples are shot at 1/15 s, 105 mm, ISO 1600. D90/18-105 on top, D300/18-200 bottom.
Full size image here.
I have not made any special tests of the autofocus speed or accuracy. The impression is that they are about the same. I have made light action shots with both - see the D90 review (the dog shots) and the 18-200VR review, but they should not be a match for a lens like the 70-200VR 2.8.
OK - over to images. First one thing I would like to mention. All images were shot as RAW+jpeg medium. Once opened up in RAW with CS3/ACR one thing was clear. Nikons in-camera processing in both the D90 and the D300 is removing CA in jpegs and quite succesfully. There has been some debate over if the D90 really has the CA-removing function - now that is settled.
Here is sample that shows how CA is automatically removed in jpegs (D90/18-105VR, 18 mm).
100 percent crops, RAW file on top, jpeg below:
I imagine that the same processing is carried over in Nikon Capture NX, but for different reasons I prefer to work in ACR/Photoshop. Since many of the test shots were made in low winter sun in contrasty light, I prefer to take advantage of the CA-removing function, and the images shown in the article are jpegs. (It was also very hard to see any sharpness advantages of the RAW-files, so it should not influence the test much.) Sharpening is the same on all compared images, USM 100/0.5/0 in Photoshop. Picture style normal (default), ADL off. Exposure varies slightly on some images because of the difference in metering between the cameras. I realize that that can have some influence on the results, but what I write is based on more images than these shown here.
I have of course made more images at full wide angle besides these. My opinion is that the two lenses are very close in sharpness at 18 mm. So lets move a little further, to around 50 mm.
D90 18-105VR top image, D300 18-200VR below:
There is a difference in exposure, 1/250s, F 8 on the D90/18-105VR image, 1/200s F 7.1 on the D300/18-200VR (both at ISO 200). However, it is hard too see any difference in sharpness.
The 18-105VR has more resistance to CA and straylight, which plays out in high contrast images like the one below (also at 50 mm, D90/18-105VR top):
At the longer end the 18-105VR pulls ahead with a slight sharpness advantage that is consistent in all test images. D90/18-105VR top, D300/18-200VR below:
For a comparison of the NEF/RAW-files at 105 mm in larger size, go here. (D300/18-200 left, D90/18-105VR right.)
And this is how far you reach with the 18-200VR. A little softer than at 105 mm.
Colors, contrast, tonality
You can not judge resolution on images in 980 pixels size, but I think they have a value in judging more subjective image quality factors as colors, contrast and tonality. I have made a small test gallery with five images in 980×651 pixels size where you easily can switch between the renderings from the 18-105VR and the 18-200VR. See the gallery here. I think you will agree that the lenses are hard to tell apart in these regards.
To summarize my views - the 18-200 helds up well in sharpness in the wide and medium range, but is more prone to CA. At the long end the 18-105VR is sharper. The 18-200 loses a bit more at 200 mm, but this is after all a 11x zoom. In terms of colors and contrast/tonality the lenses are hard to tell apart. The 18-200 is better built (mostly manifested in the metal bajonet), weighs more and is more front heavy which makes it a little less well balanced. On the other hand the 11x zoom range is a big advantage for trying out different compositions and angles. The 5.8x zoom of the 18-105VR is versatile, but does not give you quite the same powerful feeling as the 18-200VR.
The 18-105VR must be considered a bargain at the price you can get it as a kit lens. A well corrected, sharp flexible lens, a good compromise in terms of versatility and portability. Which one should you choose? Well, personally I have long thought about the 18-200VR, but the deal on the 18-105VR was too good to pass up.