28mm winners on f-mount

[davela]

I recall those older aftermarket primes being decent. The zooms, not so much.

Yeah I agree with you about most of the Zooms, as they seem a bit soft to me, although some are still acceptable. They don't hold up as well mechanically and tend to be prone to haze. There are some gems however among the fixed focus lenses.

 

[Chuck1]

Kiron has a good reputation

 

[flavio81]

sharpness is more important to me

Then the answer is easy: 28/3.5 in AI version. Tested to be really, really good.

 

[flavio81]

PC-NIKKOR 28mm f/3.5 The Nikon 28mm f/3.5 PC is a purely mechanical manual-focus perspective correction (PC) lens for film and FX digital cameras. It works on DX cameras, too.

I own this one, and it's a good lens, but I would guess a regular 28/3.5 AI or AIs will be slightly better.

I have or have had:
28/3.5 H
28/3.5 PC nikkor
28/2.0 Nikkor-N-C
28/3.5 AI

I kept the 28/2.0 for versatility and the PC-Nikkor because I absolutely love shift lenses.

My next observation is that if I know I will need to make a large print, I will use the 6X4.5 or 6X6 or 6X7 format. Knowing that one 28mm lens or another, made for the 35mm format, is 2% sharper in the far corners is just not that useful.

Well, this is a very important observation.

If the OP is really obsessed for sharpness, he/she should move to medium format.

 

[Paul Howell]

The Vivitar S1 (early versions) and Soligor CD 70 to 210 3.5s were as good as Nikon or Canon made until the Canon L version in FD mount. But there we the exceptions, most of the 70 and 80s 3rd party zooms lens were mediocre at best.

 

[skahde]

If the OP is really obsessed for sharpness, he/she should move to medium format.

Welcome to the slippery slope, close your eyes and let go! Next stop: 4x5".

 

[George Mann]

If the OP is really obsessed for sharpness, he/she should move to medium format.

Or shoot slide film with the sharpest primes, and view it directly with the best loupe on a lightbox.

 

[MartinCrabtree]

Kiron has a good reputation

I have a Kiron made Vivitar 28 f2.8. Hasn't spent much time mounted to a camera, but so far seems like a good lens.

 

[chuckroast]

Then the answer is easy: 28/3.5 in AI version. Tested to be really, really good.

Yes, it has, and it's an excellent choice especially if you want a smaller lens.

I chose the 28mm f/2.8 AIS Nikkor. Although slightly larger, it is reputed to be somewhat sharper than the f/3.5. More importantly, the f/2.8 is a newer lens with better coatings. I suspect that in any practical situation, both of these would be outstanding performers. Certainly, in the month or so I've owned the f/2.8, I have found it delightful to use on my Nikon film bodies.

 

[chuckroast]

Yeah I agree with you about most of the Zooms, as they seem a bit soft to me, although some are still acceptable. They don't hold up as well mechanically and tend to be prone to haze. There are some gems however among the fixed focus lenses.

It took the advent of cheap computing to develop high quality zooms. The optical calculations required to optimize a zoom are painfully complex and needed a computer to get done right/in any reasonable amount of time. Since prime lens elements do not get moved other than to focus, the calculations are much more tractable.

Even Nikon themselves had some fairly not-great zooms back in the day. The 43-86mm is a legendary piece of trash and the 80-200mm was only "OK" on the best of days. (I've owned and used both).

Today's modern zooms are leagues better. I've used a Nikkor 28-300mm on a digibody and it's tack sharp throughout. But I am mostly a film shooter and you're hard pressed to find something compatible with the older Nikon legacy film bodies. I do a have a 28-70 AF-D lens that I use both on digital and film bodies that serves very nicely. It's dirt cheap used and a really good performer. I took it to Italy last year and was very pleased with the results.

 

[davela]

It took the advent of cheap computing to develop high quality zooms. The optical calculations required to optimize a zoom are painfully complex and needed a computer to get done right/in any reasonable amount of time. Since prime lens elements do not get moved other than to focus, the calculations are much more tractable.

Even Nikon themselves had some fairly not-great zooms back in the day. The 43-86mm is a legendary piece of trash and the 80-200mm was only "OK" on the best of days. (I've owned and used both).

Today's modern zooms are leagues better. I've used a Nikkor 28-300mm on a digibody and it's tack sharp throughout. But I am mostly a film shooter and you're hard pressed to find something compatible with the older Nikon legacy film bodies. I do a have a 28-70 AF-D lens that I use both on digital and film bodies that serves very nicely. It's dirt cheap used and a really good performer. I took it to Italy last year and was very pleased with the results.

I've found the Nikon AI an AIS manual focus zooms the be notably superior to most similar aftermarket copies, in accordance with your observations.

Yes of course, while computer aided lens design has been around since the 1950's, access to good computers and sophisticated optical design software was challenging for all institutions except the largest and best funded (and I'm sure Nikon was one of those fortunate large ones). Even today, options for optical design software capable of designing sophisticated multi-element lenses with relative ease, at reasonable cost, are limited. Zemax, for instance, was once a viable choice for smaller users, but through corporate acquisitions it has now moved to a subscription-based and very expensive pricing "model" - sort of a shame really. I remember purchasing the DOS Zemax version (for my employer) back in the early 90's for something like $900 and it was quite powerful even then.

 

[chuckroast]

I've found the Nikon AI an AIS manual focus zooms the be notably superior to most similar aftermarket copies, in accordance with your observations.

Yes of course, while computer aided lens design has been around since the 1950's, access to good computers and sophisticated optical design software was challenging for all institutions except the largest and best funded (and I'm sure Nikon was one of those fortunate large ones). Even today, options for optical design software capable of designing sophisticated multi-element lenses with relative ease, at reasonable cost, are limited. Zemax, for instance, was once a viable choice for smaller users, but through corporate acquisitions it has now moved to a subscription-based and very expensive pricing "model" - sort of a shame really. I remember purchasing the DOS Zemax version (for my employer) back in the early 90's for something like $900 and it was quite powerful even then.

Whereas the opposite happened in the electronics space. The pay-by-the-minute iSpice program for non-linear circuit analysis can now equivalently (more-or-less) be found in the open source Spice program. One wonders why electronics designers opted for open source but optical designers seem not to have done so.

 

[reddesert]

I was thinking about the question of optical design programs a little while ago when coming up with recommendations for this thread: https://www.photrio.com/forum/threa...und-with-optics-software.215954/#post-2932795

So far as I know, there isn't any open source or mess-around package that is anything like Zemax or Code V. However, I think it's also true that without a substantial optics background, a neophyte will find it hard to make progress with even a copy of Zemax or Code V (I am not an optical designer and haven't used either, but I know and have worked with people who do). The basic answer of why there are alternatives in EE might be that there are a lot more electronics/electrical engineers than there are optical engineers/optical scientists. Look at how many high-quality EE university departments there are in the US, vs how many optical sciences university departments.

Optical design leads you to non-linear equations without closed form solutions, more quickly than electronic circuit design does - see the books I recommended in the other thread, Warren Smith's "Modern Optical Engineering" or Kingslake's "Lens Design Fundamentals." So optical design, even with computer assistance, remains something of a black art.

The flipside is that a lot of design work is contracted out to optical design specialists. Nikon clearly had its own substantial in-house expertise - see the Nikkor Thousand and One Nights stories for examples. But it is possible or likely that the design of say many 3rd party zooms from the 1970s or 1980s was contracted out. I believe we know that this was true for many of the original Vivitar Series 1 lenses, as there are various patents and other traces of the expert contract designers, see https://camera-wiki.org/wiki/OPCON_Associates

 

[chuckroast]

Optical design leads you to non-linear equations without closed form solutions, more quickly than electronic circuit design does

With respect, I disagree. Circuit design that considers transient effects is inherently non-linear and requires successive approximation for a solution since the differential equations (if they are known at all) have no closed form solutions. It is exactly that fact that drove the use of iSpice and later Spice to solve such problems.

In my early career I did primary but applied research into how people hear. What I learned along the way is that the exact same equations exist in electrical, mechanical, structural, and I presume, optical design. There is the steady state analysis, and then there is the dynamic analysis which is always hard. They just have different units of measure. For example, electrical reactance is mathematically identical to mechanical reluctance. I've not done done optical calculations, nor do I wish to, but I suspect that electromagnetic arithmetic is what is is whether we are talking about and audio system, a radar, or a zoom lens. Maxwell didn't distinguish among them.

Oh, and I probably couldn't even do the linear differential equations that do have closed form solutions any more if my life depended on it

Math is hard. Let's go shopping...