Just how good is the Nikkor 28-70mm f/2.8 "D" AFS?

The Nikkor 28-70 f/2.8 AFS has been plagued by various rumours on the Internet that it is not as sharp as it should be. Most of the rumours stemmed from the expectations that users had for the lens. At a price twice of that of the third-party manufacturers, users expected the Nikkor to deliver stellar performance. However, some users complained that the AFS lens gave soft results, and soon the rumours began circulating on the Internet.

Many other users came forth to defend the lens, giving testaments about the superb performance of their lens. It is thus interesting to perform a mini-test on two samples of the 28-70mm AFS, and let readers judge for themselves. This is by no means a subjective test, by rather a real-world test shoot to let users see the actual results, rather than listen to rumours.

The test was conducted on 2 Nikkor 28-70mm f/2.8 AFS lenses, on Fujichrome Velvia Professional (ISO 50) to test for the following attributes:

Because of the large amount of image files, I did not load them all onto this main page. Click on the individual links to check out the lens performance. Related images are grouped together on the main page for comparison purposes:
 


Distortion control at 28mm, 35mm, 50mm and 70mm

In this first test, we check the Nikkor 28-70mm AFS for distortion control at four focal lengths:
 

28mm f/8 (slight barrel) 
50mm f/8 (slight pincushion) 
35mm f/8 (slight barrel)  70mm f/8 (slight pincushion)

From the tests, we can see that there is visible barrel distortion at the 28mm and 35mm focal length, while the 50mm focal length exhibits slight pincushion distortion, becoming more visible at 70mm. So are those distortion acceptable to users? Well... it depends... all zoom lenses have such distortions, and it depends on the use of the lenses as to whether the amount of distortion is acceptable. During average uses, most photographers are unlikely to spot the distortion, but for more critical uses such as architecture, such lenses might be rejected.
 


Sharpness in the center at 28mm, 35mm and 70mm
 

       Test target:  for center sharpness 
at various focal length and aperature

yellow box indicate magnified portion

This test aims to check the center sharpness of the Nikkor 28-70mm f/2.8 AFS. The images were scanned at 300 dpi and middle portion was cropped to evaluate the sharpness. All the images were scanned at approximately the same size to enable easier judging of sharpness. As such, the 28mm and 35mm images look softer because they are smaller in size and enlarged to a greater degree for you to compare. The Nikkor was tested at the following settings:

                                        28mm f/2.8
                                        28mm f/8
                                        35mm f/2.8
                                        35mm f/8
                                        70mm f/2.8
                                        70mm f/8
 


Click here are the test results


 


What does it all mean?

The results show that the images at f/2.8 are indeed softer than those at f/8, but such results are expected of any lens due to optical properties, and not bad quality or a lousy lens. The two Nikkors 28-70 AFS tested showed similar results, and the so-called softness is definitely within specs and something to scream about.

See the similar test for the Tokina AT-X 28-70mm f/2.6-2.8 Pro II for a comparison.


Comparative performance of the Tokina 28-70mm f/2.6 -2.8 AT-X Pro II

What good is a test with you don't have somthing to compare it with? So I pitted the Nikkor against a tried and tested competitor - Tokina 28-70mm f/2.6-2.8 mm AT-X Pro II, which is based on a successful Angenieux formula:
 


Here are the test results.



The Tokina was tested in a similar fashion to the Nikkor 28-70 AFS. The results showed that the Nikkor is slightly sharper than the Tokina at all settings except at 70mm f/8, where the Tokina was exceptionally sharp. The Tokina is a well-respected lens, and it is not easy for the Nikkor to outdo it at almost all the settings. I suspect a lot of the unhappiness over the Nikkor is that users pay double the price, and they expect something magical to happen, such as twice the resolving power (which is impossible by the way).

The Nikkor 28-70mm f/2.8 AFS has been proven (in the above tests) to be a capable lens.
 


Corner sharpness of the Nikkor wide-open and stopped down

Here we will test the corner sharpness of the Nikkor 28-70mm f/2.8 AFS:
 

      Test image for
corner sharpness

yellow box indicate 
magnified portion


 
  
  
35mm at f/2.8
35mm at f/5.6

We tested the lens at the 35mm focal length. Clearly, the lens is soft at the edge at the f/2.8, as expected of any lens. Stopped down a couple of stops to f/5.6, the performance improves tremendously and it is much sharper.
 


Comparative sharpness of center sharpness between Tokina and Nikkor at 40mm f/8
 

Test image for
center sharpness

yellow box indicate 
magnified portion


 

Tokina 28-70mm AT-X Pro II

Nikkor 28-70mmm AFS

As you can see from the images above, the Tokina and the Nikkor both performed very well in the center at f/8. Both resolved the images very well, and offering neck-to-neck competition. In the previous tests, the Tokina actually gave a slightly sharper image than the Nikkor.


Final Verdict:

The Nikkor 28-70mm f/2.8 AFS is a good performer, and it actually outperformed the Tokina AT-X lens. I'm slightly surprised, because I always thought that the Tokina was already very good. There are traces of softness wide open at f/2.8, but that is expected of every optical lens. Stopped down a couple of stops, the Nikkor is as sharp as it can be. There is none of that critical unsharpness that some consumers have complained about. The lenses tested were from two different batches of lenses (one of them came fresh from the factory), and they were not prototypes.

This little test is not scientifically done in the lab, so you can argue your way around if you don't buy the results. However, this should be good evidence enough for most people, that the lens is a good performer. Let's hope that this will help to settle the argument and rumours floating around in the Internet.
 
 

- Tested by Eugene Phua and Nelson Tan -