Which one is the best 50 mm AI (or AIS)?

[ben_hutcherson]

You can tell of the lens has the compensating aperture- it does not have the "-P" suffix on the name ring. Same optical formula.

I ended up with one of each, a Nikon Ftn given to me with the compensating version of the lens.

 

Ooops, I'll have to look at mine again.

 

I have one of each also. I went looking for the compensating version not so much because of the compensating feature, but because it's supposed to be optimized for 1:10, which is around where I shoot.

 

I use it almost exclusively on my D600 or D700 under studio strobes, where the compensating aperture function actually retains its usefulness-since I'm exposing completely manually(with an external flash meter at best) having the exposure stay consistent regardless of the focus ring position is a small convenience.

 

On the other hand, the non-compensating one is what I generally stick on bellows or use in other situations where the focus ring MIGHT provide a small fine-focus tweak or might do nothing(if the lens is reversed).

 

I have two P-C non-compensating ones that are fairly late with ribbed focusing rings(not fluted). One has a factory AI conversion, while the other doesn't. I can't see any practical difference in IQ from them, and these are both outstanding lenses.

 

Regardless, though, I much prefer these simple unit-focusing lenses over even the floating element f/2.8 AI-S(although I also do a lot of macro work with the 105mm f/2.8 AF-D-I like that lens better than the newer VR version).

 

I wish I could find an AFFORDABLE example of the first 55mm pre-set lens, especially since it's actually supposed to be optimized for even higher magnifications than 1:10. The lenses themselves are not that common(and I don't want to spend a few hundred on one) and information on them seems even more scarce.

[ben_hutcherson]

Allowing 55mm you definitely have to consider the Otus. Or possibly a used car.

 

Hey, the last car I bought-a fine British Leyland product-cost me less than even a used D800 :)

[roland_vink]

I went looking for the compensating version not so much because of the compensating feature, but because it's supposed to be optimized for 1:10, which is around where I shoot.

 

I use it almost exclusively on my D600 or D700 under studio strobes, where the compensating aperture function actually retains its usefulness-since I'm exposing completely manually(with an external flash meter at best) having the exposure stay consistent regardless of the focus ring position is a small convenience.

I have instruction sheets for both the compensating and non-compensating 55/3.5 micros. Both say the lens is optimised for 1:10, which suggests their macro performance should be the same. Maybe the compensating is better but the brochure copy-writers weren't told and re-used the same material for the non-compensating version? Or maybe well-known lens reviewers happened to test compensating lenses which worked better at close range??

 

If you are mostly shooting around 1:10, then any of the 55 micros, including the AI-S 55/2.8 should be just fine. The floating elements of the 55/2.8 are only problematic if used on extension tubes or bellows - in these cases the lens should always be focused all the way out so the lens is optimised for macro shooting.

 

If you are using TTL metering, the non-compensating version is much easier to use, as you need to counter-compensate to get the correct exposure. With external metering, the compensating version is better as it maintains a constant aperture regardless of magnification (provided the lens is stopped down to f5.6 at infinity)

[Albins images]

Haha! :) I KNEW IT! ..we just HAD to talk about the micro's as well.

And again in 2 years time.

 

Myself, I liked my 55/3.5 AI a lot, but it is literally gathering dust since the 105/2.8 AIS .. and nowadays my 'go everywhere lens' (on D800) is actually an old 60/2.8 AF. yes, 60mm. That's still in the 50mm sort-of range as well, isn't it? ;) Autofocus can obviously be a great help in many situations and optically it is good enough for my use.

 

I somewhere have an E-type 50/1.8 'pancake' as well (the metal, later version? gathering the same dust as the 55), but I don't like it exactly for the reason that I DO like the micro's: it doesn't focus close .. at all. Even less close than the standard 50's, because of the short design.. UGH..

[Andrew Garrard]

Actually, if we're including 60mm lenses, the Coastal Optics f/4 APO?

[orsetto]

The performance difference between the compensating and non-compensating versions of 55mm f/3.5 Micro Nikkor is relatively small in the scheme of things, but significant to those who can repeatably exploit the slight edge the compensating version has for very close work. Hasselblad made similar modifications to their 120mm "Macro" design: the first C version was definitely optimized for closeup at the expense of infinity, the CF and later iterations slightly tweaked to boost infinity a bit at some minimal theoretical expense to the close range.

 

In my own experience, the close range performance differences aren't readily apparent except in select circumstances: exactly the right light, exactly the right distance, exactly the right subject texture will show subtle differences in the two otherwise-identical designs. This became jarringly obvious when I acquired a 'blad CF 120 Makro after using the older S-Planer version for years. At first, the new version gave all evidence to support the notion it was slightly less perfect at close range: the pics were just not quite as excellent as the older lens. Then I switched away from my usual subject matter and lighting, exploring other avenues. Suddenly, the newer lens began to slightly outperform the older "close-optimized" lens, while retaining its advantage of a more credible infinity. This prompted me to re-evaluate the Micro Nikkors, with similar results: with my changed shooting circumstances, the faint close-in advantage the compensating formula had over the later diamond-grip non-compensating vanished.

 

So yeah, there's some hair-splitting involved in these Micro-Nikkor conversations. At least both versions are fairly cheap for "legendary classics", so its easy enough to try both for yourself and see if you can detect any advantage to the older model in your own work. If you don't notice much (if any) difference, the later non-compensating version is a LOT easier to use with TTL film camera meters. Digital bodies seem to auto-magically counter-compensate metering for the older lens, esp mirrorless cams, but for film work the old lens can be a pain unless you strictly use external incident metering.

Edited September 11, 2019 by orsetto

[Brian1664876441]

The (non-compensating) 55/3.5 Nikkor-P is optimized for 1:10, I need to dig out an old Nikon/Nikkormat manual, or the old Modern Photo test that I read it in.

 

I should try both on the Df, put it to the test.

 

I did a test last year of the AF-D 105/2.8 vs the new VR lens. The latter is not as good, much more CA, not as sharp. I did this test for a friend that told me she was disappointed in the lens. Told her it was not her imagination, the images were noticeable softer compared with my lens.

 

Brian

[ben_hutcherson]

I did a test last year of the AF-D 105/2.8 vs the new VR lens. The latter is not as good, much more CA, not as sharp. I did this test for a friend that told me she was disappointed in the lens. Told her it was not her imagination, the images were noticeable softer compared with my lens.

 

I had and sold the VR lens because I was so disappointed with it, especially compared with my AF-D that seems to give me great results at pretty much any aperture and distance.

 

As for the compensating aperture:

 

As I mentioned, my most common use case for the lens is under studio strobes, which of course have a fixed output(and my old Normans are not even easy to change). Electronic lenses-like my 105mm AF-D-DO compensate for close focusing by keeping the apparent aperture constant regardless of the focus distance(provided that the lens is on the camera). It's handy on the 55mm compensating to know that as long as I'm using it at f/5.6 or smaller(which I almost always am under strobes-usually more like f/16 or smaller) it will keep the exposure the same.

[orsetto]

Just so we're all on the same page: all the various iterations of manual focus 55mm f/3.5 Micro Nikkor were "optimized for 1:10". The only difference between the compensating and non-compensating, like the difference between the Hasselblad S-Planar and Makro-Planar, lies in just how much "optimized for 1:10" the lens actually is. In the case of the older compensating Nikkor and S-Planar, the answer is "100% completely optimized for 1:10".

 

The later non-compensating Nikkor and Makro-Planar employ exactly the same optical formula as their forbears, but very very slightly offset to bring infinity performance more in line with standard lenses (lets say "80% optimized for 1:10" vs 100%). This answered complaints of the majority who wanted to use these lenses for all-round general-purpose photography: the original design assumed most buyers would use them primarily for close work, switching them out for standard lenses for distance and portraits.

 

The revised general-purpose tweak proved extremely popular, but a small subset of dedicated macro shooters were annoyed that peak performance at 1:10 and closer was ever so slightly compromised to improve infinity. Their dismay was ignored: as far as the camera mfrs were concerned, pleasing the most buyers was more important than courting the few who would notice the small difference. This subgroup was not catered to again until floating element macros (which could tweak themselves for both 1:10 and infinity) were introduced much later.

 

Of course, some people didn't like that either, claiming the now-ancient compensating f/3.5 Nikkor is still slightly better at 1:10 than even the floating element f/2.8. For their work, it could well be true: who is to say. And then you have the debacle of recent VR AFS Nikkors actually being infamously worse performers than the older AF-D versions. Lots of similar macro lenses, lots of testing to find one that works best for us. Sometimes newer is better, sometimes not.

 

FWIW, the finest "macro" lens ever offered (aside from specialty bellows glass) was the 120mm Zeiss Apo Makro Planar AF made for the late lamented Contax 645 system. Unfortunately the system relied completely on computer control of aperture, so the lens is difficult to adapt to digital cameras. Hence the still-thriving market in rare, sought-after Phase/Leaf digital backs in Contax mount.

Edited September 12, 2019 by orsetto

[chulster]

Hey, the last car I bought-a fine British Leyland product-cost me less than even a used D800

 

Use em-dashes instead of hyphens in the above context. On a Mac, they are easy to type: Shift-Option-minus.

[Andrew Garrard]

Use em-dashes instead of hyphens in the above context. On a Mac, they are easy to type: Shift-Option-minus.

 

I doubt I'll ever criticise someone for pedantry*, but if you're going to bother with that, modern style guides tend to encourage spaces either side of the dash, and often discourage the use of dashes to separate a sub-clause where the original sentence will be resumed; we should also be distinguishing en-dashes from hyphens. :-) This isn't a formal forum, and there's a long precedent for approximating typography with what's easy to type on all platforms. Arguably there's even a portability benefit in restricting the character set to ISO Latin 1, which doesn't include dashes at all. I try to stick to camera pedantry, and even then I'm not going to fight to hard about the definition of a telephoto, prime, zoom or normal lens. Besides, I'll make too many mistakes.

 

* One of the two most pleasing set-ups I've ever been given was when I was on a conference call, and a caller said "we don't want this pedanticness"; I got to say "I think you mean pedantry", and it made my week. The other was when a family member saw a zebra skin rug and said she wanted to own one so she could take it outside and hit it with a carpet beater to confuse her neighbours; I told her she was flogging a dead horse with that one.

 

As I mentioned, my most common use case for the lens is under studio strobes, which of course have a fixed output(and my old Normans are not even easy to change). Electronic lenses-like my 105mm AF-D-DO compensate for close focusing by keeping the apparent aperture constant regardless of the focus distance(provided that the lens is on the camera).

 

IIRC Canon bodies report the set aperture at all times, whereas Nikon reports the effective aperture at the macro distance (allowing, if I'm not using the thin lens equation inappropriately, for the effective position of the rear nodal point). Nikon will indeed try to match this aperture to the requested aperture when the shot is taken. I believe it's also one of the things that the better aperture control mechanism on higher-end bodies supports: if you have live view (I think) or DoF preview active, the D3/D4/D5 and D8x0 series will try to maintain the requested aperture as you focus, whereas other bodies have to set once and leave the aperture lever in place (which means leaving live view/DoF preview and entering it again). E aperture lenses, IIRC, can do the right thing irrespective of the body's aperture mechanism.

[Brian1664876441]

The compensating mechanism of the 55/3.5 is also useful for non-TTL meters, such as hand-held and the Photomic F "Bullseye".

 

The optimization talk has be very curious, good reason to do a tripod test on the Df. I also have the AF-D Micro-Nikkor 200/4, it is very sharp. But far from fifty!

[rodeo_joe1]

The talk of compensating apertures and Zebra stripes reminds me of old Zeiss Jena Praktica lenses. Many of those had compensating apertures, which I thought were very clever until I first took one apart.

 

It turns out that the mechanism is extremely simple: The connecting slide between aperture actuator and iris mechanism is sloped from perpendicular - i.e not parallel to the lens axis. Thus, as the lens is extended by focusing, the iris actuator is slightly rotated - et voila - the iris is compensated. Clever in its total simplicity, and easy to reproduce on Nikon unit-focussing primes that use a similar connecting slide.

 

As an aside (even more aside) Gene Nocon's book on darkroom printing went on at great length about focus-related aperture change; to the length of giving a table supposedly accurate to a quarter-stop..... except it wasn't actually accurately computed!

 

FWIW, the random post about em dashes made me smile. Like anyone would notice the difference on the average LCD screen, and I certainly won't be taking a micrometer to the screen of my smartphone! :p

Edited September 12, 2019 by rodeo_joe|1

[orsetto]

The compensating mechanism of the 55/3.5 is also useful for non-TTL meters, such as hand-held and the Photomic F "Bullseye".

 

And the original clip-on coupled selenium meter, which prompted the Nikkor pre-AI lens aperture fork (and Micro compensating feature) to begin with. After the CdS bullseye gave way to the first TTL Photomic and Nikkormat FT, the compensating feature in the 55mm was dropped (tho why this coincided almost exactly with tweaking the optics slightly more toward infinity is a mystery).

 

Not mine, but illustrative:

 

Edited September 13, 2019 by orsetto

[Brian1664876441]

Looks like the Selenium Meter model 2? I have the Model 3, which is still accurate.

[orsetto]

Good catch, Brian! Yeah, looks like that one is a Model 2. Before you asked, I never even knew there were three versions: I assumed just the one. Now that you prompted me to dig deeper, I wonder if the 1 and 2 were more sensitive than the 3 (most common, but has half the selenium cell area shown here)?

[chulster]

I doubt I'll ever criticise someone for pedantry*, but if you're going to bother with that, modern style guides tend to encourage spaces either side of the dash, and often discourage the use of dashes to separate a sub-clause where the original sentence will be resumed; we should also be distinguishing en-dashes from hyphens. :) This isn't a formal forum, and there's a long precedent for approximating typography with what's easy to type on all platforms. Arguably there's even a portability benefit in restricting the character set to ISO Latin 1, which doesn't include dashes at all. I try to stick to camera pedantry, and even then I'm not going to fight to hard about the definition of a telephoto, prime, zoom or normal lens. Besides, I'll make too many mistakes.

 

Pedantiception?

[mrjallen]

Pedantiception?

Chulster, just ignore the bloviations of that loquacious sesquipedalianist.

 

I have a copy of the 50mm f/2 AI that, to my amateur eyes, produces beautiful. images.

[Brian1664876441]

Good catch, Brian! Yeah, looks like that one is a Model 2. Before you asked, I never even knew there were three versions: I assumed just the one. Now that you prompted me to dig deeper, I wonder if the 1 and 2 were more sensitive than the 3 (most common, but has half the selenium cell area shown here)?

The Selenium Meter Model 3 takes an external cell to boost sensitivity- much like the Leica selenium Meter. I do not have the Nikon booster, just the Model 3.

[rodeo_joe1]

Back to 'Best 50mm AI or AI-s'.

I think I have an answer - sort of, and to my own satisfaction.

 

I amazed (or dismayed) myself by digging out my collection of old Nikon 50mm primes. I have:

50mm f/1.2 AI-S

50mm f/1.4 AI-S

50mm f/1.4 AI

50mm f/1.4 AI'd

50mm f/1.8 AI-S

50mm AF f/1.8

50mm f/2 AI'd

50mm f/2 pre-AI

I don't think I've missed any lurking in the corner of a gadget bag or fitted to a long-disused camera, but I can't be absolutely sure! It's enough to be getting on with anyway.

 

I spent, or wasted, a sunny afternoon photographing part of my dilapidated shed. Focussing on a rusty latch and flaking paint with plenty of fine detail.

 

To cut to the chase. My ranking on central sharpness at f/2 alone would be:

1. AI-S f/1.4

2. AI f1.4

3. AI-S f/1.2

4. AI-S (or AF) f/1.8

5. Pre AI f/2

I'm afraid the 'legendary' 50mm f/2 Nikkor didn't fare too well, and this was based on two samples. Even just beaten by a Zeiss Jena Praktica f/1.8 Pancolar that I threw into the test for giggles.

 

However, other factors would change my ranking a lot. All the above lenses suffer from change of focus with aperture and this alone would push the f/1.2 lens to bottom of the list; since getting accurate focus visually is near impossible wide open. The f/2 would definitely get near top place for lack of aperture-related focus creep. Closely followed by the AI-S f/1.8.

 

The amount of LOCA shown by all the lenses was quite off-putting, and again the f/2 faired pretty well. I also liked its bokeh, and hated that of the AI-S f/1.4. I wasn't that swayed by the bokeh of the f/1.2 lens either!

 

In conclusion. Is there actually an all-round 'best' 50mm out of that bunch? No. Not in my humble opinion.

 

More importantly; is it possible to focus any of those lenses accurately enough with any purely visual viewfinder for the tiny, tiny differences that can be seen by pixel-peeping to make any difference on film? Again, no. Not IMHO.

 

Here are my tests.

 

Overall image (on DX format).

AI-S f/1.4 @ f/2

AI-S f/1.2 @ f/2

AI f/1.4 @f/2

F/2 @ f/2

Feel free to disagree or post your own tests. But be warned - even using magnified LiveView, the slightest twitch of the focussing ring makes a huge difference to the critical focus.

 

Oh, and if you take their performance at f/4, f/5.6 etc. into consideration, that opens up a whole new can of worms!

[chulster]

All the above lenses suffer from change of focus with aperture and this alone would push the f/1.2 lens to bottom of the list; since getting accurate focus visually is near impossible wide open. The f/2 would definitely get near top place for lack of aperture-related focus creep. Closely followed by the AI-S f/1.8.

 

+1 for acknowledging the much worse focus shift of the faster 50s. I've never used an f/1.2, but the focus shift on the f/1.4 lens going from wide open to two stops down is so bad that I rejected the first copy I tried, mistakenly thinking that the lens was simply soft—when in fact focus error was to blame. I learned finally how to adjust focus on the f/1.4 to compensate for the focus shift. But the same issue afflicts the screw-drive AF versions, unfortunately. My copy of the AF 50mm f/1.4 focuses correctly without AF Fine Tune at f/2.8 and above. When shooting at f/1.4, I'll either dial in +10 AFFT (+5 for f/2) or just switch to manual focus, which is less of a hassle.

[orsetto]

If limited strictly to manual focus Nikkors, even our skeptical rodeo-joe demonstrated the 50/2 remains the best compromise of performance factors (and that sample variation will kill or thrill you with any individual Nikkor). The 50/2 is much less prone to sample variation than its faster cousins, but does show detectable differences based on age. The single coated H and early multicoated HC tend to be less snappy wide open than the later rubber-focus-ring K and AI updates, and all versions of the 50/2 suck in dust as much as a modern plastic kit zoom (dust dulls contrast wide open). Trying several examples of each desired Nikkor focal length is the only way to be sure of getting a "good" one.

 

Throwing in other vintage classics like the Zeiss Pancolar makes for interesting additional data points but doesn't directly apply to the original question. If we're gonna allow Zeiss and Leica to join the fray, all bets are off. The unassuming Yashica-made Contax RTS-era 50mm f/1.7 Zeiss Planar (with Leitax bayonet adapter for F mount) will blow away most any 50mm Nikkor of the same period, as will a similarly adapted Leica 50/2 Summicron-R. And of course the ne plus ultra F-mount standard lens is the mega-buck, mega-weight Zeiss Otus 55mm f/1.4 Distagon (nothing can touch it at f/1.4 or f/2.0).

[rodeo_joe1]

Skeptical? Moi?

The two samples of f/2 Nikkor I have are both multicoated (greenish reflection). One's been AI'd and the other not. I've not noticed a tendency to suck dust. In fact they both seem extremely well sealed against dust, or at least as much so as any other Nikon prime of the film era.

 

If you discount the extra 1/3rd stop, I suspect the formulation of the later 50mm f/1.8 Nikkors is almost identical to the f/2. Their performance is actually remarkably similar at f/2 and below.

 

The f/1.4 primes seem to show a little 'jump' in focus between f/2.8 and f/4. There's certainly a big gain in sharpness to be had by critically focussing at the working aperture down to f/4. If you do that, you're rewarded with absolutely stunning detail.

 

The f/1.8 Jena Pancolar isn't ZEISS Zeiss, if you get my meaning. OTOH it's not rubbish either. The corners are what let it down on full frame. But on the plus side it doesn't run you four figures to buy one!

Zeiss Otus 55mm f/1.4 Distagon

Que?

How did 'Distagon' creep in there? I thought Distagon was reserved for wideangle designs.

Edited September 17, 2019 by rodeo_joe|1

[Andrew Garrard]

Que?

How did 'Distagon' creep in there? I thought Distagon was reserved for wideangle designs.

 

That's been one of the advances in recent prime lens designs, as I understand it: Zeiss and Sigma have both used designs that would traditionally have been retrofocal wide-angles but for normal lenses; they've certainly been able to use a lot more lens elements to correct aberrations, although I don't know whether that's a requirement of the approach or whether the ability to insert more elements into the optical path was a goal.

 

I really need to read up on optics at some point so I learn some basics of what I'm talking about.

[rodeo_joe1]

That's been one of the advances in recent prime lens designs, as I understand it: Zeiss and Sigma have both used designs that would traditionally have been retrofocal wide-angles but for normal lenses; they've certainly been able to use a lot more lens elements to correct aberrations, although I don't know whether that's a requirement of the approach or whether the ability to insert more elements into the optical path was a goal.

 

I really need to read up on optics at some point so I learn some basics of what I'm talking about.

Or more cynically. Maybe Zeiss's suits just thought they'd trade on the reputation the 55mm Hasselblad Distagon had built for itself.