Opinions on lens sharpness

I have a few old manual lenses that I run on my Canon 5d with an adaptor. Nikkor mount. I stopped using the 50mm1.2 a while ago as it was feeling soft, but I never did a test. So today I did a little test exercise with the lenses I have atm. All shot on a tripod with ISO 6400 at f5.6 and 1/30s (except the Tamron 24-70). It seems like the 50mm1.2 is not any less sharp then the 1.4... I was considering getting the 1.2 serviced, but don't want to throw away money. Any feedback based on these images? Thanks in advance?

Nikkor 50mm f1.4

Any feedback based on these images?

Sorry, that would be close to impossible.

I don't know which 5D you are (ab)using here. - The old 1st one?

Looking at high ISO noise doesn't convey an impression of sharpness.

If I am supposed to judge a lens by pictures of print products, I'd either have to be able to see the individual screen dots of the writing or would need writing printed in solid black (instead of a screened gray). Please redo the testing with a solid printed newspaper at base ISO and maybe also let us know what we are seeing. - Are the pictures displayed here downsized or genuine 1:1 pixels?

Also maybe ask yourself the questions what you are demanding of which kind of lens.

I'd expect a Noctilux f0.95 to perform quite well on the few megapixels of a matching Leica, if somebody happens to be lucky enough to get it focused...

Heritage 50mms were probably more like horses for courses? - Anything fast is unlikely to be able to compete with sharpness optimized sane speed counterparts. - I'm trying to say: If I have light to shoot at f5.6, I'll pick my most promising slow lens to do so. Anything f1.4 & faster is a low light compromise optimized to shine in a different field. Figure out if you like it's (rather) wide open look for some kind of subject, use it accordingly or resell it.

If you are planning low light shots with small output: Why test for sharpness? - Isn't bokeh more interesting in such a case?

Hi Jochen, thanks for your thoughtful feedback, I learned some things! And I redid the test using a newspaper at ISO 200 (1/50s, f5.6). The main point of the test is actually to see if the 50mm1.2 is actually in OK shape, because I felt it was soft, but had no real testing. It seems OK, based on this second test, at least to my untrained eye.

I'm using the 5dMkII. These images in second test are JPGs straight out of camera, not resized. I used the 'unsmoothed' setting. (First test were also not resized but smaller JPGs).

And yes, I like a lot of bokeh, and often shoot in low light, which is why I like these prime 50s, especially the 1.2. And they are affordable for me.

 

Sorry, that would be close to impossible.

I don't know which 5D you are (ab)using here. - The old 1st one?

Looking at high ISO noise doesn't convey an impression of sharpness.

If I am supposed to judge a lens by pictures of print products, I'd either have to be able to see the individual screen dots of the writing or would need writing printed in solid black (instead of a screened gray). Please redo the testing with a solid printed newspaper at base ISO and maybe also let us know what we are seeing. - Are the pictures displayed here downsized or genuine 1:1 pixels?

Also maybe ask yourself the questions what you are demanding of which kind of lens.

I'd expect a Noctilux f0.95 to perform quite well on the few megapixels of a matching Leica, if somebody happens to be lucky enough to get it focused...

Heritage 50mms were probably more like horses for courses? - Anything fast is unlikely to be able to compete with sharpness optimized sane speed counterparts. - I'm trying to say: If I have light to shoot at f5.6, I'll pick my most promising slow lens to do so. Anything f1.4 & faster is a low light compromise optimized to shine in a different field. Figure out if you like it's (rather) wide open look for some kind of subject, use it accordingly or resell it.

If you are planning low light shots with small output: Why test for sharpness? - Isn't bokeh more interesting in such a case?

50mm1.2

 

50mm1.4

 

50mm1.4

 

[ATTACH=full]1224349[/ATTACH]

 

Both photos of the newspaper look sharp to me. Too bad there is nothing in the corners to see if sharpness falls off in the 1.2 lens.

 

By the way, do you live in the New Orleans area? I lived in New Orleans most of my life (I'm 67), but my wife and I got flooded out by Katrina. I live 200+ miles away now, but I want to move to Baton Rouge, which seems to be safe enough from hurricanes LOL.

You need to be more specific on the lenses. Nikon made many versions of their manual lenses. IMO, the common 50 mm f/1.2 AiS is a somewhat soft and flare prone thing. You'd use it for that effect, not for razor sharp high contrast shots. Never used one, but the NOCT version is supposed to be way better. It's also big $$. The Micro Nikkor 55 mm f/2.8 is for those shots where crisp is essential, and probably the sharpest lens I own. IMO, the old 50 mm f/1.4 Ai with the chrome filter ring soft compared to the black 50 mm f/1.4 AiS version. No idea about bokeh on any of them. I've always considered that a bit of a distraction, not because it isn't an issue, but because it's so scene dependent.

Both look usable to me. - I know close to nothing about Nikkors.

Curiosity: How are you focusing them? - Just according to the camera's default screen? with green dot assistance or with a dedicated screen for manually focusing fast glass? - I'm not sure if the latter is available for my Mk IV too. (Let's hope aftermarket caterers need just more time!) but I'd love to hear feedback from users.

I'm using the AiS 50mm f/1.2 on Nikon bodies, and at f/2, mine is very sharp. I don't really concur with what Conrad said - it's not soft when stopped down, nor particularly flare-prone. At f/1.2 and f/1.4, however, it is a completely different beast: a very soft-looking type of 'glow-in-the-dark' rendering (actual details are there, but the blooming effect "hides" it). Useful and nice trick, but not if you like optical perfection. This photo was with a 50mm f/1.2 wide open (on a FM2n with iso100 slide film).

 

Of course, using it with an adapter on a different mount introduces another number of variables, so that makes it hard to say a lens is at fault or not. Testing at very high ISO and very slow shutterspeeds of course also adds variables - better do some test outdoors at a reasonable distance (none of these lenses are going to be best at their shortest focus distance), in good, contrasty light.

I'd agree with that, a sort of ethereal effect wide open, but that's where I judge the lens. It's big and heavy, so if I were going to shoot at f/2, I'd probably pick the f/1.4 or the f/1.8 just for ease of use and less risk to the rarer f/1.2 :cool:

Conrad, I can understand, but comparing my Nikon 50mm lenses, the 50mm f/1.2 is really the sharpest at f/2, with good contrast too (the others are the f/1.8D and G AF versions, and a Nikkor-H f/2 pre-AI). The 50mm f/1.2 is not much heavier, and still shares the 52mm filter too. In my view, there are very good reasons for using the f/1.2, even if not used at f/1.2 - and it's not that rare, though clearly more expensive.

And I happen to like the f/1.2-effect, it's limited use but nice (though the 35mm f/1.4 AiS does it better).

 

Given the OP already owns the f/1.2, no reason to not use it. But when testing it, just make sure you do so under good conditions; I don't believe that any circumstance where you shoot 1/30th second at ISO6400, there is actually enough light for critical manual focussing, and for sure the 50mm f/1.2 can be a bit fiddly.

... Anything fast is unlikely to be able to compete with sharpness optimized sane speed counterparts. - I'm trying to say: If I have light to shoot at f5.6, I'll pick my most promising slow lens to do so. Anything f1.4 & faster is a low light compromise optimized to shine in a different field. Figure out if you like it's (rather) wide open look for some kind of subject, use it accordingly or resell it.

If you are planning low light shots with small output: Why test for sharpness? - Isn't bokeh more interesting in such a case?

 

I agree that if you're designing a lens to be exceptionally sharp at f5.6 that you'd probably not end up with an f1.4 or faster lens. But all lenses are a compromise of some sort: speed, size, weight, cost, etc. and they are designed to work well at multiple apertures. A lot of the time you're paying $$$$ for a fast lens so the designers made sure that the IQ was good. I wouldn't assume that a decent f/2.8 is sharper at f/5.6 than a decent f/1.4 is though it certainly could be. Most lenses are less sharp wide open so often a good quality f1.4 lens would be sharper at f2.8 than a f2.8 lens is.

 

My fastest lens is 50mm 1.4 and I use it all the time even though I have a 1.8. I'm in low light situations enough that it helps.

Edited December 18, 2017 by tomspielman

OK, now you got me curious about what I think, what I "know" and what I can prove. As usual, less than I thought. I did a very informal lens test- couple USAF targets at center and corner, about 15 feet away. Artificial light and a 'pod. The lens arsenal consisted of my favorite 55 mm f/2.8 AiS Micro Nikkor, a 50 mm f/1.4 AiS, a 50 mm f/1.8 AiS, a 50 mm f/1.2 AiS and my kit zoom. Camera is my antique D200 and I can say for this sort of thing that a better camera is in order. I took a couple shots, focusing by the camera sensor, and than my best by-eye effort. Some lenses did better one way, some the other. With the f/1.4 I could beat the sensor. With the f/1.2, the sensor beat me every time. The rest were a toss up. This was a highly flawed test, but better than no data at all. ISO was 100.

 

Micro-Nikkor 55 mm f/2.8 @ f/4- at least a couple elements better than everything, both center and corner, but remember it's at f/4, not f/2.

Nikkor 50 mm f/1.8 @ f/2- close to the Micro on center, loses a couple elements at the corners.

Nikkor 50 mm f/1.2 @ f/2- about an element worse on center than the f/1.8 but it beats the f/1.4 & f/1.8 by one element at the corner!

Nikkor 50 mm f/1.4 @ f/2- same as above, but one element worse in the corner. Smaller and weighs less though.

Nikkor 18-70 mm AF-S 3.5-4.5 G ED kit zoom @ f/5- almost as good as the Micro, save for what I think is field curvature. Pick the center or pick the edge.

 

Judging elements isn't exact and the sensor might be a limiting factor in some cases. MTF probably tells you more than a resolution test. You might rate any one of these differently. Any of these lenses are capable of great performance. That said, with a DX camera I can't always get it. I've proven in the field I can't focus the f/1.2 by eye to save my life. It should be easy because of the speed, yet I have no trouble with the f/1.4. I miss my F3HP finder. The Micro is definitely a cut above the others (OK, it was at f/4) and it shows noticeably when I do commercial product shots. None of the 50s really shine in the corners, but the f/1.2 seems best, something I never would have guessed. The kit zoom is quite remarkable, but it's slow. The results explain my usually shooting pattern- either the Micro, or the kit zoom are on the camera 99% of the time.

Edited December 19, 2017 by conrad_hoffman

Thanks everyone! I verified my lens is OK, at least for my use. And I learned a few extra things ftw!

The 50mm f/1.2 Ai-S Nikkor is soft wide open. Mine is. They all are. It's an f/1.2 lens with no aspheric elements!

 

Actually the impression of softness is caused by residual spherical aberration and coma. It'll resolve a test chart to quite a respectable figure. It's just that the contrast is low and everything has a kind of 'halo' around it.

 

The halo effect is the main reason to use this lens. It was never designed to be pin-sharp wide open. However at f/2.8 and smaller it should be sharp and contrasty.

That agrees with my subjective feeling. Any idea why I have so much trouble focusing the f/1.2? I'd think it would be the easiest lens to focus, not the hardest. It's mechanically and optically good.

I believe the 50mm f/1.2 suffers quite a bit from focus shift, which at the very shallow DoF it's capable of, creates quite an issue. As I wrote above, it's a fiddly lens, which was a clunky way of saying that focussing it critically well can be a pain, especially at very short distances.

 

Somehow I suspect - but I've got no means to prove it - the wide open glowing effect is also partial contributor, since focussing is done wide-open. Yet, the AiS 35mm f/1.4 is easier to focus, while it's a bigger "offender" in the glowing rendering, so not sure whether my suspicion makes any sense.

And realistically, I just enjoy these lenses for being capable of these silly effects (and yet be properly sharp at other apertures) so I never felt a need to test it.

These old "superfast" lenses are speciality tools, not daily drivers; you use them because they're not perfect, not despite their imperfections.

Edited December 19, 2017 by Wouter Willemse

I've found among the other members of my former photo collective that the perceived sharpness of a given lens is directly proportional to its cost.

 

Unless, of course, it is made by the wrong producer.

I don't own anything really expensive, so no worries there! I think hitting perfect focus is the trick with any lens, and as I get older I'm more and more thankful for good autofocus. Back to the lens tests, if one can live with a bit less crispness off center, and I'm not talking very much, the inexpensive Nikkor 50 mm f/1.8 is extremely good and it's always had a reputation for such. I also notice that the f/1.2 shows some color issues off center that none of the others did. IMO, these days the difference between f/1.2 and even f/2.0 doesn't mean as much as with film, since high ISO has become so good.

I don't have the 50mm 1.2, but I have an AI-converted 55mm 1.2. As other folks are saying, it's a quirky lens although it's decently sharp by f/4 or so. The problem with that, though, is that even at f/4 it's not as sharp as an AI-s or even AI 50mm 1.4, and I know that I'd rather carry the 1.4 unless I just NEED the extra half stop or want to embrace the 1.2, quirks and all.

 

Another fun one is the Nikkor-S 5.8cm 1.4. This is actually not a super common lens-it was introduced with the Nikon F in 1959, and was only made until 1961 or 1962 when it was replaced by the cheap, plentiful, and optically much better Nikkor-S 50mm 1.4. I've found the 5.8cm to have overall very low contrast but very interesting "swirly" bokeh wide open, although there again it sharpens up pretty quickly. Also, I'm NOT telling folks to do this, but the aperture ring on the 5.8cm is made in such a way that it will clear the AI tab of every modern camera I've tried, so it can be used unconverted higher end modern SLRs and DSLRs. It will NOT clear the EE tab used on many bodies without an aperture follower tab. I'll say this with the "do so at your own risk" caveats.

Maybe I've got an unusual good copy - it's not a very old copy, and when I got it, it actually seemed brand new. But apart from the f/1.2 and f/1.4 performance (quirky indeed) and being more tricky than other lenses to focus, my 50mm f/1.2 is a stellar performer. I have the 50mm f/1.8G as well (which is very good, and much better than the f/1.8D), and for all I see, the f/1.2 is just the better lens. Not by much, but it just renders nicer.

Then again, I simply use it - can't be bothered with test charts, brick walls or reading MTF graphs. I just judge the photos I get with it, and photos made with the f/1.2 from f/2 till about f/8 simply look better in terms of contrast, micro-contrast and colour rendering to me. Prints have more punch. Clearly, YMMV.