One lens characteristic that jumps out at you when comparing the same scene at different f-stops is the degree to which a lens vignettes. In my sample images of the D800E with the Zeiss 21mm I made a version at f/2.8, f/8 and at f/16. Flipping back and forth between the f/2.8 version and the f/8 versions on real images offers a more striking display of the extent of vignetting than any more technical test or single image display.
To use the slideshow feature to compare f/2.8 vs f/8 images click on the first image and then use the “Prev” and “Next” buttons to flip back and forth.
You’ll see a pronounced degree of vignetting with this lens at f/2.8, so much so that it may at first glance appear that I’ve adjusted the contrast of the image (I haven’t). The light falloff, of course, will darken the image around the periphery and, I presume, the camera’s light meter will compensate a bit by increasing the exposure.
In the sample images the effect shows up as greater overall image contrast, with edge shadows sometimes falling below the sensor’s range more than would be expected and in center areas high values falling higher than the sensor’s range (again, the camera was on Aperture Priority and would have no doubt increased exposure to compensate for the darkened edges). While vignette and de-vingette tools are available in software they won’t do any good if the highlights are blown and the shadow values are lost.
On the other hand I’d bet that non-technical viewers would generally prefer the look of the f/2.8 images. The have more zing to them and in some cases have a sort of 3-D look. Vignetting is, of course, part of the nature of lenses and is just one characteristic of images, inherently neither good nor bad.
I’m curious how well the Nikon D800E’s color matrix metering will do with this lens. Will the center highlights be preserved?