"Best" 35mm focal length lens, for Nikon F mount

[dokko]

This might be an interesting article for those interested in the original question:

Some New Wide Angle Lens MTF Tests: Sony 24mm f1.4 GM; Sony 35mm f/1.8; Tamron 35mm f1.4 SP

Most of you MTF junkies are aware that I've published all of our prime lens MTF curves in a series of articles by brand, and that I said I would just add new lenses to those articles. I recently tested a series of new prime lenses, though, and thought they were also worth putting up together [...]

www.lensrentals.com

Wether or not it is worth it will depend heavily on a lot of personal factors, like intended use of the image, workflow, budget and muscle strength

Personally I normally use the Tamron 35 mm F1.8 SP Di VR because I like the image stabilization in low light, the good close up range, and on big scans/prints, it‘s clearly sharper than the 35mm F2 Ai-S lens that I used before.

 

[dynachrome]

There is another context in which the discussion of lenses can be thought about. When I started taking pictures more than half a century ago, my first good camera was a 35mm SLR. I was attracted to it for two reasons. First, it allowed me to see through the taking lens. Next, I liked the idea of how it was easy to carry. Medium format cameras were larger, heavier and more expensive. My first medium format camera was a Yashica Mat 124G. I remember taking pictures with the Yashica using Verichrome Pan and a tripod. The results were very satisfying, especially for someone so young. There is the old expression that the sharpest lens is a good tripod. Still, not every photographic opportunity lends itself to the use of a tripod. When I read about resolution numbers in the hundreds of liner per mm, I know that whatever the true exact numbers were, they were not accomplished by shooting hand held. The original purpose of small equipment was to allow shooting in a variety of circumstances. I think this is still very much the case. Some of the differences between lenses are less apparent when hand held shooting is involved.
It is true that even the earliest film SLR lenses with image stabilization, from Canon and Nikon were a big help when it came to hand held shooting. The technology lessens camera shake but does nothing to offset movement of the subject

 

[reddesert]

As a student in the mid-late 1980s, I only had a short tele for portraits, got a task to photograph an interior event, and bought a used Nikon 35/2.8 non-AI for IIRC $60. I still have it. For several years I just had 2 or 3 lenses. It may not be the best of all 35mm lenses or whatever, but the deficits in the pictures I took with it have more to do with my own mistakes than any shortcomings of the lens. It is possible to fuss too much over lens resolution; further, a truism is that the best lens is the one you have with you.

 

[George Mann]

It is possible to fuss too much over lens resolution; further, a truism is that the best lens is the one you have with you.

For years I search for wider lenses that could deliver a similar level of perfection that only my 50mm f2's appear to achieve.

But then I managed to find a few wide angles, and even a series E lense that could provide me with suitably pleasing images despite their shortcommings which I have since come to accept.

I own the particular lense you speak of, and while its 6-element replacement is technical superior, it nonetheless is a decent one as long as you use the proper lense hood, and avoid situations that exhibit its rather noticeable barrel distortion.

 

[beemermark]

OMG the OP asked for the "best" Nikon 35mm lens. What is the best 35mm lens who can take a picture that a normal person is "wowed" by the picture. All the comments are ridiculous. A good photograph is a good photograph no matter how sharp or how many pixels.

 

[koraks]

From a moderator's perspective, and not specifically addressing @George Mann here, but merely taking his remark as a point to latch onto: it also appears to me that this is the phase on a debate where the discussants will have to "agree to disagree". Of course, a respectful and constructive further exchange of insights is welcomed, as always. But please, keep it civil, respectful and overall nice. Thanks all!

FYI we've started culling posts about digital sensors etc. to keep the thread focused on OP's question. If you guys want to play a "I'm always one more right than you" kind of game, take it outside.

 

[brbo]

Personally I normally use the Tamron 35 mm F1.8 SP Di VR because I like the image stabilization in low light, the good close up range, and on big scans/prints, it‘s clearly sharper than the 35mm F2 Ai-S lens that I used before.

Which film body do you use it on?

 

[Angarian]

That is correct. We were determining if a specific invention might be feasible for manufacturing. Thanks for the reminder about European and United States usage of commas and periods. This is indeed an international forum.

Thanks Bill.
Which film was used?

Honestly I am very astonished now. Because:
The highest resolution capability in microfilm I know is the 800 lp/mm of Agfa Copex HDP (Adox CMS 20 II). And that value is measured without a lens, in direct surface-to-surface process. And at high object contrast.
AFAIK this film was also used for classic microfilm use like microfiche (e.g. in the car industry for spare part lists).

Now your value is almost double (!) of that, and even with a lens involved.
As I am doing resolution tests by myself for many years, I know that one possible mistake one can make lies in the mathematical calculation:
You have your test chart with a certain number of lines per millimeter in the original. And then you photograph from a certain distance.
By evaluating the results on the film you have to calculate the lp/number in the orginal, the distance and the focal length with the correct mathematical formula.
Could it be that in that process a mistake was done?

 

[Angarian]

Some of the differences between lenses are less apparent when hand held shooting is involved.

I can ensure you based on thousands of photos that the quality improvements of the modern lenses are very clearly visible also in hand held shooting.
If that wouldn't be the case, I would have never done this importing upgrading step.
But the improvement is so big that I can definitely say that the upgrade to modern lenses has been worth really every cent.

The success on the market of these new lenses also demonstrates that the users are very satiesfied.

 

[JParker]

I know "best" is subjective, and can mean different things, to different people. In this case, I'm looking for the 35mm focal length lens with the greatest sharpness, best color fidelity, and least distortion, that will work on Nikon film cameras (i.e. not an electromagnetic aperture); G-type lenses are fine.

Thanks, for any insight.

The Sigma Art 1.4/35 will fulfill all of your criteria.
And the current Zeiss Milvus 1.4/35 as well (important: this latest Milvus version is much better than its forerunner, the non-Milvus ZF.2 version).
In comparison of the Sigma and the Zeiss the Sigma offers less distortion, but the Zeiss has this outstanding Zeiss-typical color fidelity. The Sigma is also smaller, lighter and cheaper. But the Zeiss offers outstanding build quality.

On a bit lower optical quality level there are the Zeiss Milvus 2/35 and the current Nikkor AF-S 1.8/35G ED. But they are also cheaper.

As you are going for optimal performance, I recommend to avoid the Nikkor AI-S 1.4/35, AI-S 2/35, AI-S 2.8/35 and AF-D 2/35.
They cannot compete with the above at first mentioned lenses. Time has moved on, and lenses have become much better in the last decades.

My recommendations are based on own experiences and tests with all these lenses. I have grown up in a "photographer household" in Europe. My father, mother and all my siblings are enthusiast photographers, and all of them are even much more photo-crazy than I am . Some of my friends are also passionate photographers.
Because of that I have access to dozens of different Nikkor, Zeiss, Sigma, Tamron etc. lenses. Lots of possibilities to compare different lenses .

 

[JParker]

I can ensure you based on thousands of photos that the quality improvements of the modern lenses are very clearly visible also in hand held shooting.
If that wouldn't be the case, I would have never done this importing upgrading step.
But the improvement is so big that I can definitely say that the upgrade to modern lenses has been worth really every cent.

The success on the market of these new lenses also demonstrates that the users are very satiesfied.

Exactly.
I can completely agree.

 

[MultiFormat Shooter]

The Sigma Art 1.4/35 will fulfill all of your criteria.
And the current Zeiss Milvus 1.4/35 as well (important: this latest Milvus version is much better than its forerunner, the non-Milvus ZF.2 version).
In comparison of the Sigma and the Zeiss the Sigma offers less distortion, but the Zeiss has this outstanding Zeiss-typical color fidelity. The Sigma is also smaller, lighter and cheaper. But the Zeiss offers outstanding build quality.

On a bit lower optical quality level there are the Zeiss Milvus 2/35 and the current Nikkor AF-S 1.8/35G ED. But they are also cheaper.

As you are going for optimal performance, I recommend to avoid the Nikkor AI-S 1.4/35, AI-S 2/35, AI-S 2.8/35 and AF-D 2/35.
They cannot compete with the above at first mentioned lenses. Time has moved on, and lenses have become much better in the last decades.

My recommendations are based on own experiences and tests with all these lenses. I have grown up in a "photographer household" in Europe. My father, mother and all my siblings are enthusiast photographers, and all of them are even much more photo-crazy than I am . Some of my friends are also passionate photographers.
Because of that I have access to dozens of different Nikkor, Zeiss, Sigma, Tamron etc. lenses. Lots of possibilities to compare different lenses .

Thank you! I appreciate you insights. It sounds like the Sigma is the way to go. I didn't realize there was there was such a difference between the Milvus and non-Milvus Zeiss lenses.

 

[brbo]

Because of that I have access to dozens of different Nikkor, Zeiss, Sigma, Tamron etc. lenses. Lots of possibilities to compare different lenses .

Any chance you have or could do a side-by-side comparison of an optical print from a negative shot by the best and the worst lens you have access to?

 

[JParker]

Thank you! I appreciate you insights. It sounds like the Sigma is the way to go.

You're welcome.
You will not regret your decision for the Sigma. It's definitely also the 35mm lens with the best price-performance ratio.

I didn't realize there was there was such a difference between the Milvus and non-Milvus Zeiss lenses.

It depends:
Some Milvus lenses are completely new designed and significantly improved lenses. Like the Milvus 2.8/18, Milvus 1.4/25, Milvus 1.4/35, Milvus 1.4/50 and Milvus 1.4/85.
In other cases the Milvus lens is based on its ZF.2 forerunner, with identical optics, but the significantly improved mechanics of the Milvus line. That is e.g. the case with the Milvus 2.8/21, 2/35 and both Makro-Planars (50 and 100).

 

[JParker]

Any chance you have or could do a side-by-side comparison of an optical print from a negative shot by the best and the worst lens you have access to?

Yes, I could do that. But you have to be patient, as the worst lenses are owned by my family members, and therefore in Europe. I can do the test when I am back in my hometown meeting my family.

 

[brbo]

Yes, I could do that. But you have to be patient, as the worst lenses are owned by my family members, and therefore in Europe. I can do the test when I am back in my hometown meeting my family.

That would be great!

You don't have to go out of your way to get the absolute worst 35mm there is. I'd be interested in seeing what can be gained in prints that most of us can do at home between something like the good old cheap Nikkor 35/2D and the new Sigma or Tamron 35...

 

[JParker]

That would be great!

You don't have to go out of your way to get the absolute worst 35mm there is. I'd be interested in seeing what can be gained in prints that most of us can do at home between something like the good old cheap Nikkor 35/2D and the new Sigma or Tamron 35...

O.k., a comparison test with the AF-D 2/35. That's possible, my father owns it. He bought it right after its introduction.
I've used his, too. But I've not been very impressed by it, especially because it has significant field curvature (same problem has the latest 2.8/35 AI-S). I can do the test when I am back home, but that will take some months.

In the meantime you may have a look here:

Nikkor AF 35mm f/2 D (FX) - Review / Test Report

OpticalLimits - Lens reviews and beyond!

photozone.de

and

Sigma AF 35mm f/1.4 DG HSM | A ('Art') (Nikon FX) - Review / Test Report

Sigma Art 35mm f/1.4 DG Review (FX)

photozone.de

My experience is that the tests of opticallimits are very good and precise. Every time when I have tested a lens for myself, which they have tested, too, I found that my results are to 85-100% concordant with their results.
Their absolute resolution results (the absolute numbers) are of course based on their test camera and on 50% contrast. With lots of films you can get higher absolute resolution values (of course dependent on the used object contrast).

 

[George Mann]

But I've not been very impressed by it, especially because it has significant field curvature.

The 6-element longnose Ai is distortion free, while providing competitive levels of performance between f5.6 and f11.

 

[JParker]

The 6-element longnose Ai is distortion free, while providing competitive levels of performance between f5.6 and f11.

That's not quite what I've got when I used it: Distortion was visible. And that assessment is shared by other photographers as well (including online test reports).
Performance between f5.6 and 11: Good but not excellent. And not as good as the performance of current lens designs.
And much more important for me:
When I buy a f2 lens, then I want very good to excellent performance also at the f2-f5.6 range. Otherwiese it would not make much sense to pay for an f2 lens, if I need to stop it down at least two stops for a reasonable performance.

Using the f1.4/f2 to f5.6 range is very important for me because of creative reasons. I hate to be limited in my creative options by a lens which I have to significantly stop down for a reasonable performance.
Modern lenses with their very good - often excellent - performance at wider apertures give me exactly that creative freedom I need.

 

[brbo]

My experience is that the tests of opticallimits are very good and precise. Every time when I have tested a lens for myself, which they have tested, too, I found that my results are to 85-100% concordant with their results.

Interesting. I haven't done very thorough tests and don't even own a digital camera anymore, but I found out that the test done on digital sensors were pretty much useless. For all the lenses that I have or had they basically said that lens if pretty good in the centre even wide open and edge and corner performance (resolution and light falloff, even ca) they measured never really aligned with what I was seeing in scans or prints. I decided that with the the biggest size I can print (30x40cm) the most important parameter for me is distortion and maybe flare (if I want to keep the thing really compact so I don't need to use the hood). New lenses have mostly better coatings, but they are increasingly ignoring the distortion aspect. I know why, but I'm not interested in that since I'm using the lens on film and can't use software distortion correction.

 

[JParker]

Interesting. I haven't done very thorough tests and don't even own a digital camera anymore, but I found out that the test done on digital sensors were pretty much useless.

When I compare the results with lenses of my Nikon FF DSLRs to the results on film, I always have results of the same order.
Example:
When I find that lens X is sharper at a certain f-stop than lens Y on my DSLRs, I will get the same result on film.
The only difference is that with some high-quality films like tabular-grain ISO 100 film or color reversal film, the differences are (a bit) better seen.
But in general if a lens performs exellent on my DSLRs, it also performs excellent on film.

New lenses have mostly better coatings, but they are increasingly ignoring the distortion aspect. I know why, but I'm not interested in that since I'm using the lens on film and can't use software distortion correction.

That new SLR/DSLR lenses generally ignore the distortion aspect is not true.
I have several new modern lenses which have significantly less distortion than their older forerunners. The Sigma Art 1.4/35 has negligible distortion, and less than all former 35mm Nikkors. The current AF-S 1.8/24G ED has less distortion than the AF-D 2.8/24, AI-S 2.8/24 and AI-S 2/24.
The current Nikkor AF-S 1.8/20G also has significantly less distortion compared to its forerunner AF-D 2.8/20.
And the current, modern AF-S 1.8/28G has also significantly less distortion than its forerunner AF-D 2.8/28.
Just some examples, the list is even longer.

Distortion correction by software is something we as film SLR / DSLR users can mostly ignore, because it has not played a role in the design of the modern (D)SLR lenses we use.
It is mainly used in design for mirrorless camera lenses, and here especially used by Sony.

 

[George Mann]

Performance between f5.6 and 11: Good but not excellent. And not as good as the performance of current lens designs.

Then stick with modern lenses. I prefer the build quality and accurate rendering of vintage Nikkor lenses, so I stick with them.

And much more important for me:
When I buy a f2 lens, then I want very good to excellent performance also at the f2-f5.6 range. Otherwiese it would not make much sense to pay for an f2 lens, if I need to stop it down at least two stops for a reasonable performance.

Leitz lenses are very expensive.

 

[JParker]

Then stick with modern lenses. I prefer the build quality and accurate rendering of vintage Nikkor lenses, so I stick with them.

And I like the build quality of the Sigma Art and especially the outstanding built quality of the Zeiss Milvus lenses.
The Zeiss quality is much much better than that of the vintage Nikkor lenses. Completely different league.

Leitz lenses are very expensive.

I don't need Leitz / Leica to get excellent quality. Time has moved on, there are meanwhile much more manufacturers today also offering premium quality. Which is of course very good for us photographers, we have much more excellent options today compared to the 60ies, 70ies, 80ies.

 

[brbo]

Leitz lenses are very expensive.

But with Leica lenses I at least noticed the difference. Not in prints, but in size. They are almost without an exception smaller than lenses from other manufacturers that perform on similar level. Of course, I decided I'm not wealthy enough and use non-Leica lenses on M cameras.

 

[GregY]

And I like the build quality of the Sigma Art and especially the outstanding built quality of the Zeiss Milvus lenses.
The Zeiss quality is much much better than that of the vintage Nikkor lenses. Completely different league.



I don't need Leitz / Leica to get excellent quality. Time has moved on, there are meanwhile much more manufacturers today also offering premium quality. Which is of course very good for us photographers, we have much more excellent options today compared to the 60ies, 70ies, 80ies.

Nikkor lenses are much more than adequate on Nikons with modern films like Kodak Tmax 100/400 or Ilford Delta films. If you're printing and are concerned about lens quality...it's time to consider medium format options. Film size has much more effect on image quality than does lens quality IMO. The nikkor lens on the Plaubel Makina 67/670 is brilliant.