Nikon 58 mm F/1.4

[138S]

And such claims here by "138S" that "sensors generally outresolves lenses" and that "lenses are only to resolve 17 to 25 effective MP" have the same truth level as the flat-earth claim:
No truth in it at all.

Perhaps no truth at all, this can be debated...

but another thing is you going directly to personal attack in a rude way. People sporting true knowledge and wisdom do not go that way. Instead post your lp/mm tests and evaluation methods, by insulting others you provide no technical evidence.

Just be polite, don't go to personal disqualification and provide the technical evidence instead insulting.

You may start by trying to understand the MTF graphs provided by nikon for the 28-105mm (https://imaging.nikon.com/lineup/lens/f-mount/zoom/normalzoom/af_zoom28-105mmf_35-45d_if/index.htm)

If you know what's an MTF chart you will see that by 30 lines/mm that is 15 LP/mm that zoom cannot even hold an average 40% MTF in all the frame width (wide setting). Perhaps you may be confusing LP/mm with L/mm... instead insulting, please show the CMS 20 micrography of your 200 LP/mm rating for the 28-105, with your procedure exactly described.

 

[Sal Santamaura]

"138S":
Your statements here over a very long time have been often so extremely far away from physics and the facts that a clear wording which took it to the point has finally been necessary.
Lots of other very experienced users have also tried to give you the facts. But you have unfortunately never really listened and continued with spreading huge misinformation.
We have lots of new members here, with no or very little knowledge in film photography. They need reliable information and proven facts. And not the same totally misleading myths which are spread on many other places on the internet...

Henning, "138S," under his real name (I was admonished for using it here, accused of "doxing" him, but PM me if you'd like to know it), earned his way onto my ignore lists after years of reading his blatant BS. He's created discontent at several other forums; I don't know whether he's been banned, but the bad feelings he engendered have resulted in most posters paying no attention to the nonsense he spews. I appreciate your dedication to counteracting misinformation, but, for your peace of mind, suggest you follow my lead and use the ignore function.

No surprise here. Henning is known for his passive-agressive style, maybe it has to do with the fact that English is not his first language...

That's hysterical. Denigration of his English language capability coming from Slovenia. If half of the U.S. had as good a command of English as Henning does, I wouldn't be so embarrassed to call myself an American. Would that more in my country could detect and identify BS as well as Henning does.

...no one has the right of accusing people of BS, spreading lies, being completely wrong on smth and to behave in an offensively condescending manner toward fellow forum members.

Pot, meet kettle.

 

[Joseph Bell]

I own and love the Voigtlander 58mm 1.4. It renders beautifully and feels great! I was considering the Nikon version but couldn't afford it. I am not at all disappointed with the Voigtlander and I recommend it highly.

 

[Henning Serger]

Instead post your lp/mm tests and evaluation methods,

I have done exactly that here on photrio several times, for years. If you have missed or ignored that, than it is your fault, not mine.
You could also have a look into the film photography print magazine 'PhotoKlassik" - edition IV.2016 - in which I have described my tests methods in even greater detail. And have published my resolution test results of 64 (!!) different film types. And test comparisons to digital sensors. And test comparisons of different imaging chains.
That edition can still be ordered via the PhotoKlassik webpage, by the way.

To quote you:
"Lenses are only to resolve 17 to 25 effective MP"

So, what you are saying by this is that:
1. All film manufacturers, which have offered higher resolution films like Kodak Technical Pan, Agfa Ortho, Agfa Copex Rapid, ADOX CMS 20 I/II, Agfa APX 25, Kodak Panatomic X, Kodak T-Max 100, T-Max 400, Ilford PanF+, Ilford Delta 100, Neopan Acros I/II, ADOX HR-50 / Scala 50, Kodachrome 25 / 64, Velvia 50 / 100, Provia 100F, Sensia 100 III, Astia 100F, Provia 400X, E100, E100G, Portra 160, all Superias and some more.......,
are idiots, because they have invested dozens of millions of R&D over the years in films which are useless because the lenses cannot use at all their resolution potential. Exactly that would be the consequence of your statement.
Of course the film manufacturers are not such idiots, because they have done thousands of tests and documented that the lens resolution is huge and often only limited by physical diffraction. Therefore you can get system resolution (film+lens) of much much more than 25 MP (85 lp/mm) with all the above mentioned films from low-medium object contrast on.

2. All current digital camera manufacturers are idiots, because they offer 30MP, 36MP, 42MP, 45MP and 60MP 35mm sensor-size digital cameras. But all that is useless, as the lenses cannot resolve these megapixels at all. That is what you have said by your claim.
Well, the manufacturers will not only contradict you, they will also easily show you that you are totally wrong.

3. There are millions of photographers out there using these above mentioned cameras with much more than 25MP. You say they are idiots, because they cannot use that resolution at all because the lenses lack the needed performance.
Well, these photographers will also contradict you, because they have experienced in their daily photography life that they can get indeed higher resolution with these cameras compared to their former 12, 16 or 24 MP cameras.

As all the film, lens and camera manufacturers are contradicting you, as well as million of photographers, you should at least think about the possibility that you may be wrong.........

And here an excellent test report by our member Tim Parkin, who publishes the online photo magazin 'onlandscape', and is also running a professional drumscan service:
https://www.onlandscape.co.uk/2014/12/36-megapixels-vs-6x7-velvia/
Another very well reputed and reliable source that give evidence that your claims here are completely wrong and have nothing to do with reality.

Regards,
Henning

 

[138S]

And here an excellent test report by our member Tim Parkin, who publishes the online photo magazin 'onlandscape',

This is an excellent report.

But we cannot use the peak performance in the center to evaluate the effective resolution in all the frame.

if you check DXOmark performance of the 50mm f1/4G you have 23MPix effective of the 45 possible:



Instead the Pro 300 (one of the sharpests) do deliver 38MPix effective of the 45 possible:




From 38 to 23 there is an impressive performance jump, clearly the sensor yield is limited by the 50mm lens optical peformance, with the sensor clearly outresolving the glass.

With the 300mm probably the glass may be still outresolving the sensor, but it's a $5500 toy:




Peak ratings are mostly useless as usually we are not able to reach optic peak performance:

> We shot 3D scenes that are not in perfect focus but in the DOF

> Most 35MM and MF is shot handheld. Trepidation... (if not using a fast strobe)


So it's a good academic exercise to see what lenses do on the sensor/film in a Lab when shooting a flat target, real photography is another thing, in practical conditions for sure the modern sensors are outresolving common glass, and from that fact distortion correction has little impact in IQ, which is the context we were debating.

Still it's quite interesting to know what a lens resolves... mostly I'm rating enlarger lenses and machine vision lenses, in that case I do it in the field, to see the impact of motion of the product under camera, and the effect of the frame/suport camera vibrations in the industrial place.

To rate photo lenses you don't need CMS 20, you can do it "in the air" with an strong eyepiece, in that way you remove the camera alignment factor, anyway you lose the field curvature effect if checking focus again both in the center/mid/corner.

mentioned cameras with much more than 25MP. You say they are idiots, because...

I don't use the "idiots" wording, it's you needing to clean your mouth with some soap. Please, be respectful, and don't speak like a gangster. At least in my country people using "bull shit" etc wording usualy had (sadly) a suboptimal education, I don't know in germany... I was thinking most of you were quite polite... So please correct that, at least when you speak to me.

 

[138S]

2. All current digital camera manufacturers are xxxxx, because they offer 30MP, 36MP, 42MP, 45MP and 60MP 35mm sensor-size digital cameras. But all that is useless, as the lenses cannot resolve these megapixels at all. That is what you have said by your claim.
Well, the manufacturers will not only contradict you, they will also easily show you that you are totally wrong.

You perfectly know that film/sensor has to outresolve the optics by a far extent to get all the optics can deliver:

Also think that today the LPOF on high MPix sensors are not used, it was necessary when the optics were outresolving sensors to avoid Moiré, etc

Today the glass performance works like a Low Pass Optical Filter because usually the sensor outresolves the glass. The pix count increase provocated that industrial change.

 

[j.c.denton]

There is no such thing as "outresolving". You are argumenting with MTF terms, but do not seem to understand the concept. The whole "perceptual" or "effective" megapixel figures of dxo are just a try to compress tests results into just one figure with a technical unit that is familiar to every user of digital cameras. There is an excellent read about this by Roger Cicala in the appendix of the following article: https://www.lensrentals.com/blog/2019/10/more-ultra-high-resolution-mtf-experiments/

Your last statement shows your misunderstanding.

"You perfectly know that film/sensor has to outresolve the optics by a far extent to get all the optics can deliver:"

followed by a quote of a formula to estimate the final resolving power which shows, that mathematically, the same could be said vice versa:

"You perfectly know that the optics has to perform to the highest degree of contrast transfer to get all the film/sensor can deliver."

The situation is similar to impedance adaption of power sources. Maximum efficiency is obtained when the inner impedance of the power source is equal to the impedance of the driven load. In optics this means, that for a chosen MTF frequency, there the optics need to be as good as the sensor/film. Better performance of either is just delivering diminishing returns while at the same time, lesser performance influences the result also decreasingly.

If you do not believe in the test results of Henning, there are a few others who have performed similar tests and gotten similar results with respect to slide, negative and high resolution b/w film, some of them even higher resolution figures. A collection of a few of them (including Henning) can be found here, the results can be understood despite being in german:http://shops.venditio.com/rollei-p/showpage_php/Diafilm-Test-Aufloesung.html?upage=7

Regards,
Christian

 

[138S]

There is no such thing as "outresolving".

Ok, let's see the Lens Rentals tests, we have for example this:

Yes, in the center it peaks 30% contrast transfer at 200Lp/mm , but it's LOL thinking to get that in practice.

First this is in the center, as you move to the sides contrast transfer falls dramatically.

Second, subjects don't have 1000:1 microcontrast textures at 200lp/mm on sensor, you may have 2:1 perhaps, and the low contrast transfer at high lp/mm destroys what "microcontrast" may have (even in the center).

Third: Real photography is not flat tagets on a wall, scenes are in the 3D world and usually you have almost nothing in perfect focus, you have things in the DOF, as you stop to get DOF you add diffraction, so the Lab test has a relative meaning for real photography.

Fourth: Perhaps less than 1/5000 of the shots are made on tripod, and none on a technical Lab fixture that's isolated from ground vibrations.


So... concerning if today Top today's sensors outresolve the glass work there is an easy answer: Totally in real photography situations, with the exception of really scarce situations we can enumerate if you want.

Yes, the lens can outresolve the sensor in the image center at low contrast transfer in Lab conditions... but this is irrelevant two fold because we are speaking of a low transfer and only in the center, and because real photography is not a flat target in a Lab, in real situations the Lab performance won't be reached by far.


In the past this was different with lower pixel densities...

 

[Lachlan Young]

It's fascinating that someone who has done so much to try and deny the critical importance of optical MTF/ system MTF to scanner performance in the past suddenly appears to care about it when it applies to the performance of a taking lens he can't get his hands on...

 

[NortheastPhotographic]

If you are attempting to evaluate the 58/1.4G with charts and graphs you are missing the point.

The other thing that is an under rated feature of the 58/1.4G (or really any 58) is that because it's slightly longer than 'normal' you reduce distortion when doing close up portraits. It really should be thought of as a portrait lens, as it was a gift to the entire wedding and portrait industry that shoots Nikon. Especially since for so many year's Canon shooters had the wonderful 50/1.2. The 58/1.4 actually basically matches or exceeds the 50/1.2 in subject isolation, and renders a more pleasing close up portrait. Another commenter mentioned that this lens really is for isolating a relatively centralized subject, and I would agree. I shot weddings with one for a few years and it was like the 'make everything beautiful' lens. Last year I used the 50/1.4G a lot (long story), and it did an admirable job, however the images just didn't jump off the screen/page like they did with the 58mm.

The only thing I found frustrating about it was that it was unreliable on film. It doesn't have that tack sharpness wide open, and it definitely benefits from focus calibration. More often than not my film images just looked out of focus, because they probably were.

I left Nikon after I stopped doing weddings for a while. But to paraphrase, just when I think I got out....they pull me back in!!!! So I've purchased a D4S and the 28mm 1.4D for a song from KEH. (Thank you 28/1.4E for bringing prices down on the previous version) I'm currently seeking a used 58/1.4G. On the other shoulder, I'll use my Panasonic S1R with the Sigma 50 ART.

Another tip. If you're concerned with sharpness above all and don't mind the weight most (Sigma has gone to E lenses for some designs) of the ART lenses are an excellent choice for the F6 because, using their USB dongle mount, you can do some focus calibration. I would start by calibrating one so that it's sharp at "0" on a Nikon digital body, and assume it also will work well on a film body. Personally I find that the ART lenses are a bit heavy for 35mm film. If I'm going to carry an F6 with a 50mm ART, why not a Hy6 Mod 2? If I'm concerned about detail, why not my GF670? If I still shot film weddings though, oh man the F6 would be a great choice. I'd probably have an F6 on one shoulder and the Hy6 on the other.

 

[flavio81]

The 58mm is technical marvel because it avoid coma at F/1.4, not many glasses do that.

Have you heard about the Canon FD 55/1.2 AL, released in 1971?

 

[flavio81]

Even with my old Nikkor AI-S 1.8/50mm (long barrel version) I've reached the diffraction limit at f5.6 with 240 lp/mm on ADOX CMS 20 II.

On the other hand this is probably Nikon's best 50/1.8 or 50/2 lens, ever...

Keep this secret closely guarded my Nikonians...

 

[flavio81]

Perhaps no truth at all, this can be debated...

but another thing is you going directly to personal attack in a rude way. People sporting true knowledge and wisdom do not go that way. Instead post your lp/mm tests and evaluation methods, by insulting others you provide no technical evidence.

Just be polite, don't go to personal disqualification and provide the technical evidence instead insulting.

You may start by trying to understand the MTF graphs provided by nikon for the 28-105mm (https://imaging.nikon.com/lineup/lens/f-mount/zoom/normalzoom/af_zoom28-105mmf_35-45d_if/index.htm)

View attachment 251611

If you know what's an MTF chart you will see that by 30 lines/mm that is 15 LP/mm that zoom cannot even hold an average 40% MTF in all the frame width (wide setting). Perhaps you may be confusing LP/mm with L/mm... instead insulting, please show the CMS 20 micrography of your 200 LP/mm rating for the 28-105, with your procedure exactly described.

You need to go beyond reading simple manufacturer-provider MTF charts to evaluate lens performance.

 

[Sal Santamaura]

You need to go beyond reading simple manufacturer-provider MTF charts to evaluate lens performance.

And anyone who's grown weary of expending effort dealing with nonsense from "138S" needs to do what I long ago did, namely, add it to your ignore list.

 

[flavio81]

if you check DXOmark performance of the 50mm f1/4G you have 23MPix effective of the 45 possible:

This is an analog forum. Why are we even quoting DxOMark? Which, IMO, is an useless website.

Well, that's it. Adding 138S to my ignore list. Bye!

 

[flavio81]

And anyone who's grown weary of expending effort dealing with nonsense from "138S" needs to do what I long ago did, namely, add it to your ignore list.

Done!

 

[flavio81]

Hello Flavio,

I don't either, but in this special case with the 1.4/58 G it is not much of an issue. If you use the lens as intended. This lens is not designed as a "general all-round" lens, and it is especially not designed as a lens for architecture, where distortion really is an issue (and where also the field curvature of this lens is clearly visible).
This lens is designed to isolate the important subject in the middle / centre of the frame, and to separate it from the backgound in a three-dimensional way. Subjects like people, animals, flowers, trees etc. And it is doing that job very well, look at the example thread on the FM forum.
And in theses cases for which this lens is designed for, its level of distortion is no real problem at all.

Best regards,
Henning

Yes, you are completely correct.

It's just a shame that the cited Nikkor lens is a G-lens. I hate G lenses, because they cannot be used in my classic Nikon cameras (F2/F3). Also, they look really really ugly.

But on the other hand it's great that Nikon is looking outside of the box and away from the silly resolution-race.

 

[138S]

The other thing that is an under rated feature of the 58/1.4G (or really any 58) is that because it's slightly longer than 'normal' you reduce distortion when doing close up portraits. It really should be thought of as a portrait lens, as it was a gift to the entire wedding and portrait industry that shoots Nikon. Especially since for so many year's Canon shooters had the wonderful 50/1.2. The 58/1.4 actually basically matches or exceeds the 50/1.2 in subject isolation, and renders a more pleasing close up portrait. Another commenter mentioned that this lens really is for isolating a relatively centralized subject, and I would agree. I shot weddings with one for a few years and it was like the 'make everything beautiful' lens. Last year I used the 50/1.4G a lot (long story), and it did an admirable job, however the images just didn't jump off the screen/page like they did with the 58mm.

IMO it is a NOCTurnal lens, you may use it for many other things.

Of course, the 58mm can be used as a portrait lens because it has a nice Bokeh, if you want to shot portraits with a 58mm... and not with a DC 105mm...

The 58mm vs 50mm decreases the "nose job" (selfies from phones) effect that happens when distance to subject is close (say under 3m), for the same framing you shot farther with the 58 than than with the 50, but the 58mm 1.4G has more distortion than the 50mm 1.4G

The expensive 58mm 1.4G was made to not have coma when wide open. Its direct predecessor is the Nikon 58mm f/1.2 NOCT-NIKKOR, which share with the 1.4G this very unique and uncommon feature: Very low coma when wide open.

So that amount of money is useful when shooting at night wide open and wanting bright spots rendered as spots and not as blobs.

Of course you may use it for portraits, but all that cost (2 aespheric elements, etc) is not to excel in portraiture, it's for low coma wide open in the corners that for portraits would be in the OOF anyway. For portaiture you have other more suitable choices at lower cost, but of course no problem if wanting an expensive NOCTurnal lens for that.


For portraits we may find some "Duftige Schärfe" effect... but this can also be also obtained with other glasses.

 

[Lachlan Young]

The other thing that is an under rated feature of the 58/1.4G (or really any 58) is that because it's slightly longer than 'normal' you reduce distortion when doing close up portraits. It really should be thought of as a portrait lens, as it was a gift to the entire wedding and portrait industry that shoots Nikon. Especially since for so many year's Canon shooters had the wonderful 50/1.2. The 58/1.4 actually basically matches or exceeds the 50/1.2 in subject isolation, and renders a more pleasing close up portrait. Another commenter mentioned that this lens really is for isolating a relatively centralized subject, and I would agree.

Fast 'long normal' lenses have long been quietly sought after - as you indicate, they can offer a better balance of perspective, speed and fall-off that creates a more pleasing image than the mug-shot/ surveillance effect of a longer focal length - I quite actively dislike lenses over 50-60mm in 135 (or equivalent) - I find their perspective really uncomfortable to work with.

the 28mm 1.4D for a song from KEH. (Thank you 28/1.4E for bringing prices down on the previous version)

The 28/1.4 D is a stunning lens on 35mm - if you want a wonderfully crisp yet beautiful 40" across optical darkroom print off Tri-X, it'll do that (if a little barrel distortion doesn't bother you). Great environmental portraiture lens - extraordinarily high performance compared to so many SLR lenses.

 

[MattKing]

I quite actively dislike lenses over 50-60mm in 135 (or equivalent) - I find their perspective really uncomfortable to work with.

Which goes to show that YMMV applies to this issue.
My favourite lens for people pictures on 135 is my 85mm lens. For 6x6, I really like using my 135mm lens for the same purpose. Strangely, for 6x4.5 I am comfortable with the results from my 110mm and 210mm lenses, and I sold my 150 mm lens.

 

[StepheKoontz]

At all it won't separate the subject from the background better or worse than the regular 50mm f/1.4G (beyond focal difference), but it will deliver coma free corners wide open. If you defocus background then usually corners will be OOF, and not a single $1of this overcost will be much worth to spend. Also if you stop a bit the lens (2.8 or 4) you have no single benefit over cheaper designs at same aperture.

The 58mm is technical marvel because it avoid coma at F/1.4, not many glasses do that. For the rest it's like the 50mm 1.4G but with the drawbacks of slightly larger distortion and delivering focus breathing.

They aren't even similar for the intended use of the 58G, which IMHO is environmental portraits. What the 58G shines at is how it transitions to OOF on both sides of the focus plane, as can been seen in this "angled ruler" shot. So many optics, especially really sharp ones, are supper choppy in this critical transition zone. The 58G is ultra smooth, which is what gives it the look it has.

The other thing you have to understand is under about 3 meters, it shouldn't be used wide open as it's clearly not sharp used this way. At f2 up close it's fine and the DOF is so narrow under 3 meters wide open, that too was never it's intended use. Past 3 meters, it's sharp wide open. This fact is likely why it tests so poorly on some sites, folks are testing it at close distances.

 

[StepheKoontz]

Also it's quite interesting to see here: https://www.dxomark.com/Lenses/Nikk...4G-mounted-on-Nikon-D800E---Measurements__814

Sorry I didn't catch you read DXOmark reviews, nevermind...

 

[StepheKoontz]

Flavio, lens design is not only extremely complicated and a "science of its own", it is also a kind of an "art". Because as a lens designer you always have to make compromises: If you improve the parameter "A", you get worse performance on parameter "C" at the same time and vice versa. And you have dozens of different parameters to consider, and all are interacting with each other.
I am happy that I am not a lens designer, it would drive me nuts.......

And I think many of these traits are influenced by the lens designer themselves. You can look at the 35mm f1.4G and the 58mm f1.4G, both designed by the same guy, both have similar optical traits, smooth rendering over sharpness.

 

[Lachlan Young]

They aren't even similar for the intended use of the 58G, which IMHO is environmental portraits. What the 58G shines at is how it transitions to OOF on both sides of the focus plane, as can been seen in this "angled ruler" shot. So many optics, especially really sharp ones, are supper choppy in this critical transition zone. The 58G is ultra smooth, which is what gives it the look it has.

The other thing you have to understand is under about 3 meters, it shouldn't be used wide open as it's clearly not sharp used this way. At f2 up close it's fine and the DOF is so narrow under 3 meters wide open, that too was never it's intended use. Past 3 meters, it's sharp wide open. This fact is likely why it tests so poorly on some sites, folks are testing it at close distances.

In other words it's two quite clever lenses in one that allows them to service two premium markets (astrophotography & the whole shallow DoF social/ portrait market - neither of whom tend to have shallow pockets).

On the DoF matter, I'm still persistently amused by how people complain about the performance/ difficulty of accurately focusing their 105/ 2.4 on the Pentax 67 - not having quite grasped that in a 16x20ish print taken with the lens wide open, your effective DoF at a meter is rather under a centimetre...

 

[j.c.denton]

Another tip. If you're concerned with sharpness above all and don't mind the weight most (Sigma has gone to E lenses for some designs) of the ART lenses are an excellent choice for the F6 because, using their USB dongle mount, you can do some focus calibration.

The Sigma Art lenses are simply awesome. However, be warned that many of the newer ones won't fully work on the Nikon F6. In essence, all lenses using electromagnetic apertures can only be used wide open. This is the whole reason I started to use Canon EF mount SLR film bodies as they are compatible to even the newest lenses. I don't know if Canon's user interface will ever grow on me as I vastly prefer to use Nikon and Pentax bodies. But the outstanding lens performance weights in heavy.