28mm f/2.8 AI-S vs. the 28mm f/2 AI Lens Test

[Corran]

I've posted a few times about these two lenses and it seems today it came up several times so I wanted to post these test images for reference.

It is my finding that at any equivalent aperture, the Nikkor 28mm f/2 AI lens is equal or better than the 28mm f/2.8 AIS. This is of course one test with one lens each. I should note the 28mm f/2 lens actually has a bunch of pinpricks on the front of the lens element though! I got it cheap because of this. It of course opens up one more stop as well.

So here is the proof. Here are crops from the extreme upper-right corner of the image.

I shot them both on a tripod with my D800E. The first image is at f/2.8, so wide-open on the AI-S and one stop closed on the f/2. The second shot is at f/8. This was carefully focused via Live-View. The image is of course a landscape and shot at distance - not a close-up or macro, which the f/2.8 excels at - so if that's your thing, ignore this.

The 28mm f/2.8 AI-S also has a touch more fall-off at the wider stops, as one would expect.

This isn't a test chart in a lab so feel free to disagree, or maybe your f/2.8 model is just stellar or mine sucks. I just thought it might be appreciated to show these results.

By the way, the center images were a toss-up. At f/2.8, the 28mm f/2 AI is very slightly blurrier from spherical aberrations and has a touch more flare, but at f/8 it has pulled ahead and is sharper than the 28mm f/2.8 AI-S.

 

[johannielscom]

Mine is a 2.8/28mm AI and I think it gives my 3.5/55mm Micro-Nikkor a run for its money when it comes to sharpness, the 55mm barely staying ahead.

Hard to find something not to like with these 28mm Nikkors.


Just for fun, the lens on this camera is a 2.0/28mm pre-AI, that's been all over the world as a professional photographer's lens:

it's down to the bare metal but the glass is clean and it's still sharp from 4.0 onwards!

 

[walnuts]

Would like to see a test with the subject much closer, and filling the frame. Close up is where the 28mm F2.8 AIs shines, with its addition of the CRC element.

 

[Paulbe]

IMHO your particular copy of the 20/2.8 may have a problem. The one I have seems just fine--thanks for posting this test.
Paul

 

[Paulbe]

Whoops! That should be 28/2.8--please excuse!

 

[Corran]

That very well be true as I bought it used on eBay. But it's dead mint and works flawlessly. Generally abused equipment looks abused, but of course not always.

I'd be happy to see someone duplicate my test with the opposite result. That would sure be interesting.

As for the close-up performance, I wouldn't be surprised if the AI-S pulled ahead but I don't use it much for that and therefore didn't test it. I'm not much for lens tests but I was curious enough to do this.

 

[mdarnton]

I have the 28/2 and have never been disappointed by it. What's interesting to me is that I also have a 28/2.8 AF-D, and it's not nearly as horrible as I've read people saying.

Are these film or digital tests? And what are the specifics? If my D300 would offer a parallel experience (I never shoot color on film), I'd be willing to offer up examples from my two 28s.

However, my favorite 28, by a long shot, is my 28/1.9 Voigtlander on Leica. 🙂

 

[Corran]

As I mentioned I used a D800e.
The difference on the corners is cropped on an aps-c sensor.