28mm comparison

[David Paterson]

nfoto started a very popular thread about the Nikon 28mm f2 AIS which is itself a popular classic lens, and for very good reasons.

I have owned several copies of the 28/2 and still have one. In an idle moment I decided to do a small comparison between that lens and the 18-35mm AFS G at 28mm, both lenses at f5.6 which should be optimum or near-optimum for both. This was far from an exhaustive, thorough examination of the two lenses, just a quick feel at how they compare at this often-used aperture.

The tests were made using my D600 and the two files were given absolutely minimal processing - exposures equalised and converted to tif in Capture NX2; given a single modest sharpening, converted to 8-bit, and 100% crops saved as a level-10 jpegs in CS6.

The results are interesting. The 18-35 at 28mm gives 7% wider coverage than the 28/2but was 1/2 stop darker; the zoom is distinctly cooler than the prime but both lenses give very acceptable well-saturated colour; contrast-levels are almost identical. Looking at sharpness, the centre and edges are excellent  with both lenses and I can detect no differences; in the extreme corners, however the modern zoom easily beats the classic prime; only in the *extreme* corners , it must be emphasised. Both lenses performed this little test admirably and the soft extreme corners produced by the 28mm prime would normally be of very little consequence.

1. The whole file.
2. 100% crop, centre, 18-35mm afs g zoom
3. 100% crop, centre, 28mm ais

 

[David H. Hartman]

To my eyes the 28/2.0 AIS is a touch sharper in the center not enough to make a difference is any size print I'd make.

Dave who will now drink a cup of coffee and see if his eyesight improves.

[Jack Dahlgren]

How are they at f/2.8? I'm often shooting the 28mm at that aperture.

[David Paterson]

To my eyes the 28/2.0 AIS is a touch sharper in the center not enough to make a difference is any size print I'd make.
Dave who will now drink a cup of coffee and see if his eyesight improves.

Well, one of them is the sharper, but I wouldn't take or make any bets on which one.   

Jack - the 18-35 AFS G doesn't have f2.8 - it's an f3.5 lens.   

[Jack Dahlgren]

Well, one of them is the sharper, but I wouldn't take or make any bets on which one.   

Jack - the 18-35 AFS G doesn't have f2.8 - it's an f3.5 lens.   

Indeed.

[Akira]

Hmm, to my eye, the wall surrounding the poster frame looks a little sharper with the zoom set at 28m.

The difference of the coverage for the same pixel areas could be caused by the difference of the distortion in addition to the difference of the actual focal length.

[Roland Vink]

the 18-35 AFS G doesn't have f2.8 - it's an f3.5 lens.   

Closer to f4.5 when zoomed to 28mm

The center crops look very similar, the 18-35 looks a little sharper to my eye but there is hardly anything in it. If focus is even a tiny bit out it could affect the results here so I wouldn't place too much weight on one test.

Differences in coverage could be explained by several things. Distortion will affect the field of view, the lens with more barrel distortion will see a little wider, even if the focal length/magnification at the center is identical. However, the crops don't show the same field of view - the zoom is a little wider, so the focal length is not the same. It's likely the zoom wasn't set to exactly 28mm, maybe where it is labelled "28mm" on the barrel is actually 27.5mm or similar. Also, the focal length of the AIS 28/2 is really 28.6mm.

The difference in brightness could be due to vignetting, the 18-35 @ 28mm is closed down less than one stop at f/5.6 so mechanical vignetting into the corners will be significant, and the image will appear darker. The 28/2 is closed down three stops at f/5.6, so will be free of mechanical vignetting and have much more even illumination across the frame. The same tests at f/8 or f/11 should yield more comparable results.

[David Paterson]

Thanks, Akira and Roland, for your interesting comments - you are both more knowledgeable about lenses and optics than me, and I learned a lot from your posts. My little "test" was merely intended to find out if there was any substantial difference between the two lenses - sufficient to influence my choice of lens in certain situations. I think my amateurish procedure just about established that there was only a very minor difference at the chosen aperture of f5.6.

I was aware that certain variables - true aperture on both lenses, true point of focus on the 28/2 and true focal length on the zoom - could not be guaranteed; I even accidentally nudged the tripod while changing lenses; there was never any chance of scientific accuracy! But I think I can stop wondering which lens is better at the common focal length, and that was the point of the exercise. 

[David H. Hartman]

One of the most important items to control is focus accuracy. If one misses the focus by just a touch the test is in valid.

[Akira]

If I remember correctly, there is +/-5% of allowance for the nominal focal length of a lens.  So, the actual focal length of a 28mm lens can be from 26.6 to 29.4mm.

What makes the matter more complicated is that the close focus compensation and the internal focus mechanism will influence the actual focal length according to the focusing distance.  The nominal focal length indicated on the catalogs is the one when the lens is focused at infinity.  That said, in the case of wideangles focused at this (seemingly) medium distance would not be influenced by the working of the mechanism.