NikonGear'23
Gear Talk => Lens Talk => Topic started by: Akira on October 19, 2016, 04:21:33
-
Sometime ago, Digital Camera Watch published an interview to the Nikon designers of AF-S 105/1.4E in two parts:
http://dc.watch.impress.co.jp/docs/news/interview/1017554.html
http://dc.watch.impress.co.jp/docs/news/interview/1018916.html
I'm afraid it is too long to translate everything, so I extracted some interesting points. If you have any further questions, please feel free to ask.
Hope this would be of any help.
--105/1.4E was intended as the lens that celebrates the 100th anniversary (in 2017) of the foundation of Nikon company, and originally designed as a 100mm lens. For the coherency in the existing product line, however, it was redesigned as a 105mm one.
--Around the time 58/1.4G was designed, Nikon founded a special project team to analyze the optical characters of the historically renowned lenses of different makes using their original measuring machine called OPTIA. The results of the measurements have been taken into considerations in the design of 105/14E
--Zenji Wakimoto, the legendary Nikon optics designer, decided that the steps of focal lengths should be decided according to more natural transition of the angle-of-view and propounded the 50, 36 (nominal 35), 24, 20mm step instead of Contax’ 50, 35, 25, 21mm step. (At that time, neither made 28mm lenses.) To extrapolate the steps into the longer range, they decided on 85 and 105mm focal lengths.
--The basic concept for the aberration correction of 105/1.4E is similar to that of 58/1.4G and is based on Nikon’s original 3D Hi-Fi concept. However, 105/1.4E was designed to be a little sharper than 58/1.4G when it is set wide open and focus closer.
Nikon thinks that there is no absolute answer to the best balancing point of the aberration correction that satisfy all the users. They will offer lenses with different balances of aberration corrections to choose from.
--The conventional lenses have been designed for the infinity (except for older Micro lenses or special purpose lenses). The subject or the scene at infinity can be considered as 2-D object, so a lens can be designed for the flat subjects. On the other hand, the closer subjects are mostly 3-D (except for the reproduction of documents), and 2-D design doesn’t work well. That’s why they started to design the lenses based on the 3-D concept.
--The resulted optical design of 105/1.4E is the combination of the front converter plus the master lens of a modified Gaussian type which is rather similar to the current super-tele design concept.
--ED elements are employed to correct the chromatic aberration. The ED elements are most effective in the lenses longer than 135mm. So Nikon hasn’t used ED glasses even for the fast mid-tele lenses. Also, the conventional mid-tele lenses are of Gaussian (or its derivative) type, and its convex lenses should be made of high-refractive-index glasses. The refractive index of ED glasses is generally low, which makes it difficult to use in the Gaussian design.
--The aspherical element was not absolutely necessary thanks to its relatively narrower angle-of-view.
--The MTF chart of 105/1.4E tells that the mid-angle area is sharper than the central area, because the main focal point doesn’t always coincide with the center (portraiture, for example).
--That said, they don’t like their lenses to be judged only by the MTF chart.
--The cut out of the round bokeh will show up occasionally, which is inevitable in the combination of (D)SLR and f1.4 lenses.
--Ais 105/1.8 was difficult to convert to an AF lens because the total weight of the optics was too heavy to move them at enough speed. The power efficiency of a supersonic motor is not really high.
--VR function was discarded at an earlier stage of its design process, because they decided that the lens would be too long, fat and heavy to use at reasonable comfort.
-
Thanks for sharing this info with us! LZ
-
Many thanks Akira for translating the very interesting information for us.
Rarely such detailed information is given out to the public.
-
Thank you Akira!
-
Akira - Thanks for translating the important points. I'm glad to see that Nikon remains determined to design for more than MTF scores, which seem to generate so much internet fuss, hype and commentary. It's also interesting to see that they deliberately optimised sharpness away from the central field of the 105 mm, for off-axis detail. It sounds like a great deal of thought went into this lens!
Cheers, John
-
--105/1.4E was intended as the lens that celebrates the 100th anniversary (in 2017) of the foundation of Nikon company, and originally designed as a 100mm lens. For the coherency in the existing product line, however, it was redesigned as a 105mm one.
--Zenji Wakimoto, the legendary Nikon optics designer, decided that the steps of focal lengths should be decided according to more natural transition of the angle-of-view and propounded the 50, 36 (nominal 35), 24, 20mm step instead of Contax’ 50, 35, 25, 21mm step. (At that time, neither made 28mm lenses.) To extrapolate the steps into the longer range, they decided on 85 and 105mm focal lengths.
Thank you Akira.
100mm lens for 100 years. Nice idea. If the lens had been 100mm, the entrance pupil would be reduced to 71mm, so the standard 77mm filter would have been used and the barrel would be smaller. 105mm would have been chosen to provide more equal spread of focal lengths between 85 and 135mm. This was the reason Nikon chose 10.5cm (105mm) for the original rangefinder Nikkor-P 10.5cm f/2.5 lens back in the 1950s.
If lenses had a genuine equal progression between focal lengths, we would see 25, 35, 50, 70, 100, 140, 200, 280, 400 ... (1.4x or sqrt(2) difference between each lens)
-
good job :o :o :o
-
LZ, Daniel, Tristin, John, Roland and Rick, thanks for kind words. I'm glad you enjoy reading the summary.
John, an engineer also said that the MTF of the old 105/2.5 from the rangefinder era showed characteristics similar to that of 104/1.4E (a little higher in the middle-angle area than the center), which suggests that Mr. Wakimoto already had that design concept decades ago!
Roland, even if the lens would have been 100mm, it would required the front element of similar diameter in order to avoid vignetting, because of the wider (even slightly) angle-of-view.
Correction: both Nikon and Contax made 28mm lenses in the rangifinder era.
-
surprised that wakimoto is still alive :o :o :o
-
Very interesting, Akira. Thanks for the translation !!
-
Thank you so much for the detailed translation, makes perfect sense! Very interesting reading the back ground - IMHO the performance of the 105mm AFS 1.4 E is much closer to 58mm AFS 1.4 G in real life use on a DSLR than may people 'think' - the whole series of 1.4 lenses are perfectly matched by design.
Again, thank you for the translation ;)
-
Most illuminating explanations. Thank so much for translating into this succinct synopsis :-)
I'm glad to see that Nikon remains determined to design for more than MTF scores, which seem to generate so much internet fuss, hype and commentary.
Hear!! Hear!! Bravo
-
Akira,
Thank you for the translation. I tried the Google translation and made sense of some of it but much didn't make sense.
Best,
Dave Hartman
---
My take on optimizing sharpness to the mid area is giving up a little in the center for better performance near the edge. The eyes in a portrait won't be in the center a way. That is a point made in the interview. Read it a few times and it will stick. :)
I wonder if the 105/2.5 AIS & AI were designed that way. I remember lens reviews in Modern Photography where the 75-150/3.5 E was sharper in the center than the 105/2.5 AIS but the latter was sharper at the edge.
-
This has more to do with field curvature, depth of focus distribution, I would think.
-
Roland, even if the lens would have been 100mm, it would required the front element of similar diameter in order to avoid vignetting, because of the wider (even slightly) angle-of-view.
Correction: both Nikon and Contax made 28mm lenses in the rangifinder era.
In the rangefinder era, Nikon made 21, 25, 28, 35mm wide angle lenses. In the SLR era, changed 25 to 24 to provide better separation from 28, and consequently changed 21 to 20 to provide better separation from the new 24, so we now have 20, 24, 28, 35. Most manufacturers followed the same logic, except Zeiss which retains the older focal lengths.
As for the proposed 100/1.4 lens, it would have an entrance pupil = 71mm which is 6mm smaller than 77mm filter.
This is identical to the Canon 85/1.2, a faster lens with wider angle of view, so 100/1.4 with 77mm filter should be less difficult. Of course it is likely the 85/1.2 has very high mechanical vignetting.
Nikon have made a number lenses with similar focal length which have very tight margin between entrance pupil and filter:
Lens Pupil Filter
Auto 85/1.8 47 52
Noct 58/1.2 48 52
AIS 105/1.8 58 62
AIS 135/2.8 48 52
-
John, Erik, chambeshi, Dave, I'm glad you enjoyed the article and my humble summary.
In the larger part of the second part, the interviewer asked nerdy questions about the aberration correction. If you are interested, I will be glad to summarize it.
-
As for the proposed 100/1.4 lens, it would have an entrance pupil = 71mm which is 6mm smaller than 77mm filter.
This is identical to the Canon 85/1.2, a faster lens with wider angle of view, so 100/1.4 with 77mm filter should be less difficult. Of course it is likely the 85/1.2 has very high mechanical vignetting.
Nikon have made a number lenses with similar focal length which have very tight margin between entrance pupil and filter:
Lens Pupil Filter
Auto 85/1.8 47 52
Noct 58/1.2 48 52
AIS 105/1.8 58 62
AIS 135/2.8 48 52
I would suspect that the vignetting was not considered as critically up to the Ais era as in later years. For example, the AF and AF-D 85/1.8 already had 62mm filter threads. As you know, the designers of 58/1.4G said that they decided on f1.4 instead of f1.2 to avoid the excessive vignetting.
-
I really like vignetting in fast lenses to keep the image together, is very nice and often supportive in slim dof images.
Also with todays amazing image quality from the cameras/sensors vignetting is so easy to fix/manipulate in PP.
After shooting a lot with the 105mm 1.4 AFS E and seeing others images from the same lens I notice more and more the swirly Bokeh when there are highlights in the background.
These are more noticeable than for instance compared to the Bokeh ball from Nikkor 135mm AF-D 2 DC or 125mm 2.5 APO Lanthar - the 105 1.4 Bokeh Balls have more of an sharp corner in the two corners where they are cut off,,,
-
Experience photographers would know how to live with or make use of the vignetting. As mentioned in another thread, the Yahica/Contax Distagon 25/2.8 showed excessive vignetting even well stopped down and was known as "favorite lens for connoisseurs" (at least in Japan) who liked to use it as the positive image effect.
Admitting that the vignetting is rather a favored effect for some, it would be an issue to address for the lens designers. Sigma also mentions (or even boasts about) the rich amount of light in the peripheral image area achieved by their Art 50/1.4.
-
--105/1.4E was intended as the lens that celebrates the 100th anniversary (in 2017) of the foundation of Nikon company, and originally designed as a 100mm lens. For the coherency in the existing product line, however, it was redesigned as a 105mm one.
Interestingly, Lensrentals measured the focal length to be 101mm, maybe it was designed as 100mm after all?
"... the numbers suggest since the Nikon actually is 101mm at infinity, not 105mm"
See: https://www.lensrentals.com/blog/2016/11/nikon-105mm-f1-4-e-mtf-bench-tests/
Maybe someone who has this lens can compare with other 105mm lenses to see if the field of view is wider or the same...
-
LOL! You would have to compare it with, say, the venerable 105/2.5!
I have once tested two different AF-S 35/1.8 lenses (DX and ED) on my D7000 side by side at the Nikon Showroom. Focused at a fairly far subject (at least tens of meters), the ED (FX lens) showed noticeably wider angle of view. So, I wouldn't be surprised by the deviation of the actual focal lengrh from the nominal one. It is still within the 5% tolerance.
-
Very interesting. Thank you for your translation.
-
You are welcome, Frank.
-
LOL! You would have to compare it with, say, the venerable 105/2.5!
At least we can be sure that lens really is 105mm:
(http://www.mir.com.my/rb/photography/companies/nikon/nikkoresources/preAI70/105mmf25preAIoptc.gif)
-
Maybe it should have stated "redesignated" instead of "redesigned".
-
At least we can be sure that lens really is 105mm:
Roger C tested the new 105mm AFS 1.4 to be 101mm at infinity :)
-
Roger C tested the new 105mm AFS 1.4 to be 101mm at infinity :)
Yes, we know. I was curious about the actual focal lengths of other Nikkor 105s. Ais Micro 105/2.8 is known to be 82mm with the built-in teleconverter. How about Ai/Ais Micro 105/4.0?
-
Yes, we know. I was curious about the actual focal lengths of other Nikkor 105s. Ais Micro 105/2.8 is known to be 82mm with the built-in teleconverter. How about Ai/Ais Micro 105/4.0?
Not obvious form the thread,,,
Anyway the old 105mm 4 Ai Ais is unit focusing so must be very close to 105mm
-
Thank you so much for the article and translation Akira. Really generous of you.
-
Not obvious form the thread,,,
Anyway the old 105mm 4 Ai Ais is unit focusing so must be very close to 105mm
Sorry to be unclear. I've forgot that a Micro lens should offer correct magnification along with the 52.5mm extension tube, so it should be 105m!
Aguinaldo, glad to be of any help.
-
Yes, we know. I was curious about the actual focal lengths of other Nikkor 105s. Ais Micro 105/2.8 is known to be 82mm with the built-in teleconverter. How about Ai/Ais Micro 105/4.0?
Other 105mm lenses including 105/2.5 (sonnar and gauss) and 105/4 micro are exactly 105mm - you can check the lens schemas on the mir site. I'm pretty sure the AIS 105/1.8 is also 105mm. These are all unit focusing lenses so the focal length does not change when focusing.
The other 105mm lenses - AIS 105/2.8 micro, AF 105/2.8 micro, AFS 105 VR micro, 105 DC - all have CRC, IF or RF, so have focal length shortening at close range. I assume all are 105mm (or very close) at infinity ...
-
it's nice to see that Nikon is recognising that people are using class lenses for a reason and they are doing something about it :o :o :o
-
Sometimes the angle of view given in the lens specifications gives us a clue about the real focal length. For example the newer "35mm" lenses have a slightly wider angle of view than the older lenses, which suggests they have true 35mm focal length instead of 36mm as with most older models.
For the 105/1.4 Nikon gives the following:
Nikon FX-format D-SLR cameras: 23°10'
Nikon DX-format D-SLR cameras: 15°20'
This is almost identical to the angle of view for all other "105mm" lenses. Actually, other 105mm lenses give 23°20' for FX format, which suggests the focal length is very slightly longer than 105mm :o :o
-
Other 105mm lenses including 105/2.5 (sonnar and gauss) and 105/4 micro are exactly 105mm - you can check the lens schemas on the mir site. I'm pretty sure the AIS 105/1.8 is also 105mm. These are all unit focusing lenses so the focal length does not change when focusing.
The other 105mm lenses - AIS 105/2.8 micro, AF 105/2.8 micro, AFS 105 VR micro, 105 DC - all have CRC, IF or RF, so have focal length shortening at close range. I assume all are 105mm (or very close) at infinity ...
Ais Micro 105/2.8 goes up to 1/0.88 magnification with PN-11. All AF Micro 105s are intended to be used without PN-11, so a little deviation from the nominal focal length could be tolerated.
-
it's nice to see that Nikon is recognising that people are using class lenses for a reason and they are doing something about it :o :o :o
According to LensRental review, the sample variation of 105E is a bit wider than that of 85/1.4G. The variation may become smaller over time...
-
Sometimes the angle of view given in the lens specifications gives us a clue about the real focal length. For example the newer "35mm" lenses have a slightly wider angle of view than the older lenses, which suggests they have true 35mm focal length instead of 36mm as with most older models.
For the 105/1.4 Nikon gives the following:
Nikon FX-format D-SLR cameras: 23°10'
Nikon DX-format D-SLR cameras: 15°20'
This is almost identical to the angle of view for all other "105mm" lenses. Actually, other 105mm lenses give 23°20' for FX format, which suggests the focal length is very slightly longer than 105mm :o :o
Hmm...apparently things have become complicated. :o
The 105E uses three ED glasses that are known to have larger expansion factor than the normal glasses. I wonder if the difference of the tempertarure of the testing room should be counted in? Although it is hard to believe that the expansion factor affects the focal length by the order of several millimeters...
-
Ais Micro 105/2.8 goes up to 1/0.88 magnification with PN-11. All AF Micro 105s are intended to be used without PN-11, so a little deviation from the nominal focal length could be tolerated.
https://richardhaw.com/2016/10/30/repair-nikon-pn-11-extension-tube/
Here is my PN-11 teardown ::)
i opened it up from the wrong side :o :o :o
longer lenses doesn't have as much difference when compared to wide lenses so a little bit of difference in FL should not be too noticeable.
-
Good job, Rick, as always! Yes, the deviation of the focal length will be less influential on a longer lenses.
-
Lensrentals did a teardown of the 105/1.4E:
https://www.lensrentals.com/blog/2016/12/taking-apart-the-new-nikon-105mm-f1-4e-ed-af-s/
No ring type SWM; early information seems to have been incorrect regarding this aspect. (Not that it matters to me; I don't care about the implementation when in fact the lens focuses well.)
-
thank you Akira!
-
Thanks, Ilkka, for the link to the teardown article. I wonder if they could have used a stepping motor, if it is a USM micro motor.
The reason for the second element group is sealed should be because of the nano-crystal coating.
thank you Akira!
You are welcome, Bob!