The deserved praise that has been given this lens is somewhat marred by quality control problems. I think it is worth the effort to find a good copy if you need a sharp, fast, wide angle prime for stills and video. I returned four of these to Amazon. The serial numbers were fairly close, so I assumed that Amazon had a bad production run. I ordered a fifth copy from Adorama, and it meets my expectations.
Nikon seems to have thrown quite a bit from their optimization arsenal at this lens, including two aspherical elements, Super Integrated and Nano-Crystal coatings and Rear Focus. See Nikon’s Web Site for more information.
The first place I look for a lens assessment is DxO Labs. Their review called the 28mm f1.8 “a stellar performer”, which is some cause for excitement. See the full DxO Mark review to see how it compares with other makers, including Zeiss.
My initial performance observations, based on using the lens with a Nikon D800 are in fair agreement with other reviewers:
Auto Focus speed fairly good, similar to the Nikon AF-S 50mm f/1.4G
The lens is very light weight
Compared to its f1.8 siblings, the build quality seems lower
How do you know if you have a good copy? Here are the problems I found.
Crunchy or gritty feel when the lens is focused manually. This is an indicator of AF problems.
Auto Focus sticks in one place. Every so often, three copies I tested would just not focus until I switched to manual mode, moved the focus collar a bit, then switched back to AF.
One sample was just flat out, obviously fuzzy.
Pronounced lateral chromatic aberration. Some samples were worse than others, but the quality was definitely poor. See the LCA comparison below.
The Nikon gold lettering has a brassy orange matte look compared to any other Nikon lens I’ve seen.
Extreme focus shift when stopping down. Some focus shift can be expected with a fast prime, but it is usually within the increased depth of field when stopped down. Some tests shooting a brick wall looked less sharp at f5.6 than f1.8 because the focus shifted so much beyond the wall.
I shot the same target in the studio using a Nikon D800 on a tripod, tethered to a laptop using Nikon Camera Control Pro 2, triggered remotely. Both exposures were ISO 100 @ f/1.8 @ 1/80s. The lenses were fine tuned for auto focus before shooting the target using the technique recommended by Nikon. Both copies needed to be adjusted for back focus.
The first image clearly show a problem with LCA for a lens I returned last week.
Here is the same test shot with the sample I’m keeping. LCA is still present, but looks more like the 10µm measured be DxO labs. Each 5µm LCA results in 1 pixel of false color for high contrast subjects on the D800. At 200%, it looks like 1 or 2 pixels.
At f1.8, vignette is pronounced…
Vignette is completely gone by f/5.6.
I need to run quite a few more tests, but it looks like the sample I have has it under control. Shots like those above, where focus is on the brick wall, had better sharpness on the building behind the wall at f/8.0.