Wed 07-Nov-12 02:38 PM | edited Thu 08-Nov-12 02:36 AM by nrothschild
I just happened to have purchased a 16-35 VR yesterday, and just happened to have taken a few quick snaps to check for vignetting and corner smearing and a few other things.
This image was shot on a D700 at 16mm, wide open at f/4. That should be the worst case vignetting for this lens but I have not gotten any further yet, testing it stopped down, as it would be used for a landscape image like this.
I use this scene because the uniform grass at the bottom, plus the sky in upper left, usually makes for a somewhat worst case. In a more typically complex scene I don't think the vignetting is as obvious and that is consistent with some other casual shots I've taken.
This first image is straight out of camera, saved by Photomechanic as a JPG from the embedded JPG in the raw file.
(edit: I had the in camera Vignette Control set to "High" so the vignetting could be worse than illustrated here, depending on your camera settings and workflow)
This second image is from the same image file but was processed in Capture NX 2.3.4, with the vignette control slid all the way to 200%. I think this should make any but the most obsessive pixel peeper fairly happy .
I personally find that the vignette control in CNX2 is not strong enough but that is what I use and don't have any other raw rendering apps for comparison. It can not, for example, fully unwind the vignetting from my 70-200 VR (original I version) when the scene accentuates it. I fault Capture for that and why this is so is one of those annoying mysteries.
Anyway, since I do use Capture I use this as an acid test to see how easy it will be to eradicate any vignetting in those images that bring it out. As far as I'm concerned the 16-35 passes this test easily. I would not worry about it.
The vignetting here is probably about comparable to my 20mm f/2.8 Ai-S shot at f/4, on the same scene but in very different light. I have not yet tried to do a better comparison.
I agree with Brian - vignetting is an issue, to some extent, with most lenses on FX - it is the dirty secret of FX for those that obsess over this but as I illustrate it is easily corrected even with a not very good tool for this job like CNX2. I don't obsess over it, but I am curious about these things and thought that an actual sample would help reinforce his comments.
Please don't try to divine sharpness issues from this image. I was focusing rather close in order to try to keep the lower corners reasonably within the DOF, which is not easy, nor is it really very meaningful at f/4 or sensible in the real world . I may not have picked the best focus point for my objective; it was just a quick snap to help me get my bearings.