50/2 AI vs. 50/1.8 AI

[Airy]

Today I started making series of shots at various places, distances and with varying lighting conditions.
It is, as usual, very difficult to come to stable conclusions even though, on a particular shot, differences may spring into the eye. One does not expect two nifty fifties from the same generation (34-38 years ago) with similar designs from the same manufacturer to exhibit significant differences. Surprize: there are differences, and pixel peeping is not even needed.

The following shows the first shot, f/2 for both. I put all comments on top because I do not know how to intersperse shots and comments.

Left images are with the 50/1.8, right images with the 50/2. The side-by-side pics are screen captures, capturing the X/Y comparisons allowed by Lightroom, a very useful feature.

On the general shot, you'll see that there is more vignetting with the 50/2: not surprizing.
The center crop (100%) shows that the 50/1.8 is more contrasty: see for instance the patterns in the marble. Also, it exhibits less purple fringing on shining edges (background). Resolution seems however similar.

Looking at the bottom left reveals that the 50/2 is much sharper there. Is that a consequence of field curvature?
The bottom right details (tree, slightly in the foreground) and top right details (tower, in the background) tell that in both cases, the 50/2 is sharper - significantly so on the tree leaves. Bottom line: the 50/2 is a more homogeneous performer at f/2, by a big margin. The 50/1.8 has the better center in terms of contrast and fringing.

[Tristin]

Doesn't look like field curvature as the poor corners don't appear out of focus.   An easy way to check for fied curvature is to take shots of a flat detailed subject , like a poster, from a close distance.

[longzoom]

Unfortunately, those two lenses are incomparable - one from the left has clearly visible centering problem.  Sorry.  LZ

[Airy]

Could be. What makes you think of a centering problem ?
(NB there are more shots to come, so we'll see if the diagnosis is confirmed or not. I had so far two clearly decentered lenses - Oly 45/1.8 and Nikkor 50/1.8G - but the problem was much more obvious)

[Airy]

Doesn't look like field curvature as the poor corners don't appear out of focus.   An easy way to check for fied curvature is to take shots of a flat detailed subject , like a poster, from a close distance.

I agree it is not field curvature. However to diagnose it, I'd use a flat subject, but shot at an angle. And not necessarily at close distance, since field curvature can be present at any. A good example is the Zeiss 25/2. Admittedly, most lenses show an exacerbated field curvature at close distances.

[Airy]

Unfortunately, those two lenses are incomparable - one from the left has clearly visible centering problem.  Sorry.  LZ

OK, whatever the cause, you are right. I just checked the pics taken at the shop yesterday, and I definitely picked the poorer copy. Sh*t happens. I'll try and exchange it for the other one.

Since the low performance appears +/- equally in all four corners, the cause is more likely an axial misplacement of some element, rather than de-centering or tilting.

[Roland Vink]

To test for field curvature at infinity: take a picture while focusing in the center, then take the same shot and focus on the edge or corner. Using live view on a tripod is best. If the image focused on the edge or corner shows noticeable improved edge/corner sharpness (and loss of central sharpness), then field curvature is present.

You might run into problems with this test if the field curvature takes moves the focus plane "beyond infinity", in other words beyond the infinity hard-stop on the lens - it will be impossible to get the corners into focus. The only way around this is to stop down to increase the DOF.

[richardHaw]

http://www.nikkor.com/story/0002/

read the full story here

the 1.8 is regarded as in improvement

[Roland Vink]

The AI 50/1.8 has better bokeh:
- Smoother background rendition (gets worse towards the corners though)
- Nicer 7-blade aperture, compared to 6 blades for the 50/2

[longzoom]

OK, whatever the cause, you are right. I just checked the pics taken at the shop yesterday, and I definitely picked the poorer copy. Sh*t happens. I'll try and exchange it for the other one.

Since the low performance appears +/- equally in all four corners, the cause is more likely an axial misplacement of some element, rather than de-centering or tilting.

.      Yeah, as you said - things happen...  Could you imagine the situation when my new 200-400 VR1 appears show  the same problem, just 2 days before fly to Europe? 90 min of crazy driving to Melville, (do not want to disclosure my arguments!), got it back in 7 hours of the same day... In absolutely stunning condition! Went back as happy as possible. After this lens never have the  version 2, till that 80-400G. Good luck! THX!   LZ

[Airy]

Got the 50/1.8 replaced

[Harald]

Hi,

the 50mm f2 is extremly good regarding its Age: The Design is about 50 years Old. The 1.8 Ai-S is so good that it is Producer still today. As Far as i know it is identical to the 1.8D. I tested both against the 1.8G and there is not that much difference in IQ starting @f4 or so.

Harald

[Roland Vink]

the 50mm f2 is extremly good regarding its Age: The Design is about 50 years Old. The 1.8 Ai-S is so good that it is Producer still today. As Far as i know it is identical to the 1.8D. I tested both against the 1.8G and there is not that much difference in IQ starting @f4 or so.

The "long nose" AI 50/1.8 (serial no 1760801 - 2165418) and AIS 50/1.8 (serial no 3135001 - 3304551) have the same optics, but they are different from the optics found in the "pancake" 50/1.8 series-E  and AIS versions, and the AF and AF-D versions. Both are good designs but I think the older is a little better.