Studio Tests - FX format
Switch the 16-35mm to its native FX format and it produces impressive results. However there's one distinct gotcha, which is extreme barrel distortion at 16mm, to a level beyond anything we've previously seen from a conventionally-corrected SLR lens.
|Sharpness||Sharpness on FX is generally very high. The extreme corners are somewhat soft wide open, especially at 16mm, but sharpen up nicely on stopping down with best results around F8 - F11. Again we see the least impressive results at 35mm, with the same pronounced asymmetry and softness towards the upper left corner as on DX. Overall optimum sharpness across the frame is generally obtained at F8.|
|Chromatic Aberration||Chromatic aberration is very low - there's a little green / magenta fringing in the corners at wideangle, but you have to look very, very closely to see it.|
|Falloff||We consider falloff to become perceptible when the corner illumination falls to more than 1 stop below the center. Falloff (or vignetting) reaches a maximum of 2 stops in the extreme corners at 16mm F4, but decreases rapidly on stopping down to F5.6.|
|Distortion||Distortion is strikingly high, especially at 16mm where it measures 4.5% barrel - the highest we've seen to date by some considerable margin. Fortunately though the distortion pattern is relatively simple, and therefore easily correctable in post-processing. It diminishes rapidly on zooming in , through neutral around 24mm to moderate pincushion at 35mm (-1.1%).|
Specific image quality issues
As always, our studio tests are backed up by taking hundreds of photographs with the lens across a range of subjects, and examining them in detail. This allows us to confirm our studio observations, and identify any other issues which don't show up in the tests.
Control of flare is a critical feature of a superwide lens; with such a broad view of the world, bright light sources will find themselves in the frame on a regular basis. Shooting extensively in bright summer sunshine, we saw few flare problems unless the sun was either in or just outside the frame.
Place the sun in the frame, though, and you'll get flare patterns which aren't especially attractive, and which become increasingly visible on stopping down. At F4 they're rather diffuse, but can still be obviously visible, and by F11 or F16 they're distinctly defined (as in the example below left). In strongly backlit situations the story is somewhat similar - there's little obvious flare at larger apertures, but stop down and unsightly patterns and streaking can become much more obvious. This isn't the best handling of flare we've ever seen from a wideangle zoom, but neither is it the worst.
|16mm F16, Nikon D3X||35mm F4, Nikon D3X|
Our studio tests show that lateral chromatic aberration is generally very low, and with most recent cameras any residual fringing will be removed automatically, both from the out-of-camera JPEGs and the RAW files processed through Nikon's software. However if you're using an older camera, or shooting in RAW then processing with third-party conversion software, then color fringing will start to become visible.
The samples below show the kind of fringing you can expect to see. Even at its worst, at 16mm on the D3X, it's scarcely objectionable, and you'd have to be making a huge print to see it. On the D300 the CA is somewhat less visible in these 100% crops, although this is mainly due to the lower pixel count. In any case it's readily correctable in raw conversion when required.
Nikon D3X, 16mm F8
Nikon D300, 16mm F9
RAW + ACR
100% crops, left of frame
100% crops, CA corrected in ACR (Parameters: -15 Red/Cyan, +10 Blue/Yellow)
Barrel Distortion at 16mm
As shown by our studio tests, the lens suffers from considerable barrel distortion at 16mm on full frame (indeed at 4.5% it's higher than we've seen from many Micro Four Thirds lenses, which rely on integrated software correction to give acceptable images). This means that if you shoot a composition with straight lines towards the edge of the frame, they'll end up disconcertingly bent, and you'll probably want to apply some correction in post-processing to get the picture to look right. Therefore if you're allergic to to both distortion and post-processing, this really isn't the lens for you.
The example below shows this in practice. The image on the left is a straight conversion in Adobe Camera Raw (with a little CA correction applied to remove fringing), and the barrel distortion is rendering the pillars rather the wrong shape. On the right is the same image, but now with distortion correction applied in Photoshop (+8 in CS4's Lens Correction filter, combined with a little perspective correction to remove converging verticals). This gives results that are more than acceptable in terms of geometry, although at the expense of a slight loss in angle of view.
|16mm, Nikon D3X||Distortion corrected in Photoshop|
One objection that's often made to this kind of processing is that it inevitably degrades corner detail, and while this is true, in reality the blurring effect isn't exactly huge. After distortion correction has been applied the corner is visibly a little softer, but you'd have to look at a large print very closely indeed to have much chance of noticing this (while the negative impact of leaving distortion uncorrected will be visible at any size).
|16mm F11, Nikon D3X, RAW + ACR||Barrel distortion corrected in Photoshop|
100% crops, bottom left corner
Aside from the obvious affect on your images, one further problem with distortion to this degree is the effect it can have on your image composition; it's all too easy to end up with shots that aren't at all straight, by subconsciously aligning using a dominant line towards the edge of the viewfinder. Personally I found the D3X's virtual horizon particularly usedful when shooting with this lens.
Wide-open softness at 35mm
Our studio tests show that our review sample of the 16-35mm suffers from softness when shot wide open at 35mm (but not at other focal lengths), and the samples below illustrate how this looks in practice. At first sight it would be easy to dismiss this as user error - perhaps a touch of shake, or slight misfocus - but the asymmetry of the effect, with the top left corner clearly worse than the rest of the frame, betrays it as a lens issue.
Focus was placed on the clock on the right side tower, and at F4 this is visibly just a little softer than at F8. Look at the wind direction indicator opposite it on the left side, though, and things are much worse, with clear double-image 'ghosting'; again, though, it cleans up at F8. In the crop from the lower center, contrast is visibly reduced at F4, while fine detail is still well-defined.
|35mm F4, 1/1000 sec, Nikon D3X||35mm F8, 1/320 sec, Nikon D3X|
100% crop, upper right
100% crop, upper left
100% crop, lower center of frame
Of course there's no reason to believe that every copy of this lens that rolls off the production line will show this same fault, but it demonstrates why our sharpness graphs at 35mm are relatively low.
Optical Image Stabilization
The 16-35mm features Nikon's second generation 'VR II' vibration reduction system, which claims to allow hand holding at shutter speeds up to four stops lower than usual before blur from camera shake becomes apparent. The mechanism is near-silent in use, with just a near-inaudible whirring noise when operational, and with distinct clicks when it activates and deactivates due to the VR group moving in and out of the 'at rest' position.
To quantify the effectiveness of Nikon's VR II system we subjected the 16-35mm to our studio stabilization test at the wide and long ends of the range, using the D3X as the test camera. We take 10 shots at each shutter speed and visually rate them for sharpness. Shots considered 'sharp' have no visible blur at the pixel level, and are therefore suitable for viewing or printing at the largest sizes, whereas files with 'mild blur' are only slightly soft, and perfectly usable for all but the most critical applications. The subject distance for these tests was approximately 1.5m at 16mm, and 3m at 35mm.
|16mm VR OFF||35mm VR OFF|
|16mm VR ON||35mm VR ON|
We've generally found modern in-lens stabilization systems to be highly effective, and it comes as no surprise to see the same again here. The VR II system is doing an excellent job, giving significant benefit at both the short and long ends of the zoom; and while shutter speeds as low as 0.8 sec appear to be a step too far, overall this is still an impressive result.