NickTrop
Mentor
So, I've always rather liked the 28-80 3.3-5.6G plastic Nikon kit lens produced between 2001-2005. Reading up on this thing on the internet to validate my opinion (nodding my head in approval of those who share my opinion that this is an undervalued little gem of a lens, and shaking my fist at the screen when someone disagrees...) I stumbled upon an article on the Nikon site by the actual designer of the lens,Haruo Sato, who gives an interesting (to a lens geek) account of the rationale for his design choices along with a completely honest assessment of his "baby", best aperatures/focal lengths, sample photos etc. (And his samples are using a D800 shot in 2017, so my guess is this article is fairly recent.)
The plastic fantastic 28-80 f3.3-5.6G is only a 6 element, 6 group design. On the internet in photo blogs, Youtube vlogs etc. there has been some debate over opting for older lens designs -- and primes over zooms, because they have fewer elements. Fewer elements giving more pleasing images with a better "3 dimensional quality". So it's interesting to hear an actual lens designer for 30 years "weigh in", albeit likely inadvertantly, on this raging "Coke v Pepsi"-ish debate. Some highlights:
IV. More lens elements are a "necessary evil"
-- As a young designer, a number of outstanding predecessors and mentors shared with me their knowledge of optical design. Among the tidbits passed along, I often remember hearing that "more lens elements are a necessary evil".
-- I have spent my entire career designing lenses to be as simple as possible and to utilize the fewest number of lens elements possible.
-- the less experience a designer has, the more they may think that simply increasing the number of lens elements will improve the design, but they then discover that aberrations are not compensated as well as they expected and find themselves caught in a dilemma.
-- Have you never used a lens, such as one with a six-elements-in-two-groups Dagor (Doppel-Anastigmat Görz) structure or a three-elements-in-three-groups triplet structure, that would seem to offer poor aberration performance, but actually rendered images better than did a seven-elements-in-six-groups Gauss structure?
-- MTF cannot be used to determine whether or not increasing the number of lens elements with this lens would be beneficial or harmful. That is why I often use cemented lens elements and always try to achieve designs that use the fewest number of lens elements possible.
Although Mr Sato-san doesn't tie "3D quality" to fewer elements, it's interesting he does make mention to this phenomenon at the end of the article:
-- These days, I find myself most interested in the optimization of image formation characteristics...I am of the opinion that current evaluation methods for imaging optics are not sufficient to judge the performance of lenses for the video age sure to come. As we all know, the subjects of both photos and video are three dimensional. That means that image formation should also be evaluated in three dimensions. The performance of imaging optics must be evaluated based on their three-dimensional characteristics. As for optical design as well, I had thought that a time when we are able to completely control three-dimensional optical characteristics would naturally come.
https://imaging.nikon.com/history/story/0063/index.htm
The plastic fantastic 28-80 f3.3-5.6G is only a 6 element, 6 group design. On the internet in photo blogs, Youtube vlogs etc. there has been some debate over opting for older lens designs -- and primes over zooms, because they have fewer elements. Fewer elements giving more pleasing images with a better "3 dimensional quality". So it's interesting to hear an actual lens designer for 30 years "weigh in", albeit likely inadvertantly, on this raging "Coke v Pepsi"-ish debate. Some highlights:
IV. More lens elements are a "necessary evil"
-- As a young designer, a number of outstanding predecessors and mentors shared with me their knowledge of optical design. Among the tidbits passed along, I often remember hearing that "more lens elements are a necessary evil".
-- I have spent my entire career designing lenses to be as simple as possible and to utilize the fewest number of lens elements possible.
-- the less experience a designer has, the more they may think that simply increasing the number of lens elements will improve the design, but they then discover that aberrations are not compensated as well as they expected and find themselves caught in a dilemma.
-- Have you never used a lens, such as one with a six-elements-in-two-groups Dagor (Doppel-Anastigmat Görz) structure or a three-elements-in-three-groups triplet structure, that would seem to offer poor aberration performance, but actually rendered images better than did a seven-elements-in-six-groups Gauss structure?
-- MTF cannot be used to determine whether or not increasing the number of lens elements with this lens would be beneficial or harmful. That is why I often use cemented lens elements and always try to achieve designs that use the fewest number of lens elements possible.
Although Mr Sato-san doesn't tie "3D quality" to fewer elements, it's interesting he does make mention to this phenomenon at the end of the article:
-- These days, I find myself most interested in the optimization of image formation characteristics...I am of the opinion that current evaluation methods for imaging optics are not sufficient to judge the performance of lenses for the video age sure to come. As we all know, the subjects of both photos and video are three dimensional. That means that image formation should also be evaluated in three dimensions. The performance of imaging optics must be evaluated based on their three-dimensional characteristics. As for optical design as well, I had thought that a time when we are able to completely control three-dimensional optical characteristics would naturally come.
https://imaging.nikon.com/history/story/0063/index.htm
Nitroplait
Well-known
Haven't followed the link yet, but looking forward to read it.
Although not a fan of triplets (not because of rendition, but because you have to stop down a bit to achieve acceptable sharpness and thus they become impractical), my favourite lenses are the 4 element Tessar and Elmar types.
The Nikon history site is interesting in many ways because, besides historical facts, it adds the human element to camera and lens design. Somewhere they talk about the modest and overlooked F-301/N2000 which despite the the plasticky appearance got quite a bit of attention and love from the designers, and several people within Nikon has it as their favourite manual focus camera - go figure.
Sometimes as much human ingenuity went into creating the modest marvels as in the prestige projects.
Although not a fan of triplets (not because of rendition, but because you have to stop down a bit to achieve acceptable sharpness and thus they become impractical), my favourite lenses are the 4 element Tessar and Elmar types.
The Nikon history site is interesting in many ways because, besides historical facts, it adds the human element to camera and lens design. Somewhere they talk about the modest and overlooked F-301/N2000 which despite the the plasticky appearance got quite a bit of attention and love from the designers, and several people within Nikon has it as their favourite manual focus camera - go figure.
Sometimes as much human ingenuity went into creating the modest marvels as in the prestige projects.
NickTrop
Mentor
Great points, I agree in full.
Ronald M
Mentor
Usually at Leica newer version has less elements.
50 2.8 Elmar M is a wonderful lens. I have two. 90 4.0 macro is same. Only one.
50 2.8 Elmar M is a wonderful lens. I have two. 90 4.0 macro is same. Only one.
Phil_F_NM
Camera hacker
These days I love seeing the results from low element designs. Having gotten into 16mm work over the last 2 years, many lens designs are based upon Cooke Triplets, of which I have a few.
What surprised me the most with low element designs was the fact that if I stopped down my 1913 Kodak meniscus just a bit, it is almost impossible to see the difference against the same scene taken with a new Fujinon-W EBC on large format images.
I think what it comes down to is what each of us find pleasing for our own work and if two elements do it, awesome; but if a thirteen element lens does it for us, that's awesome too.
Phil Forrest
What surprised me the most with low element designs was the fact that if I stopped down my 1913 Kodak meniscus just a bit, it is almost impossible to see the difference against the same scene taken with a new Fujinon-W EBC on large format images.
I think what it comes down to is what each of us find pleasing for our own work and if two elements do it, awesome; but if a thirteen element lens does it for us, that's awesome too.
Phil Forrest
CMur12
Well-known
Thanks for posting this, Nick. Very interesting that Mr Sato did not consider MTF to be an adequate test of image quality.
I must admit to being in possession of a bit of a purist streak. Perhaps from that, I tend to believe that the fewer lens elements needed for the job, the better.
I have a soft focus filter (Bay 1) for my Tessar and Tessar-type TLRs. (I like Tessar and Tessar-type lenses.) However, it made more sense to me to start with a softer lens (for portraiture), rather than degrade the image from a sharper lens, so I bought a Yashica D with the 3-element Yashikor. The lower contrast, softer focus, and vignetting at wide apertures seemed to offer natural advantages.
Then again, I haven't actually done any further experiments in window-light portraiture since the early 1990s, so I never actually tested my hypothesis with the Yashica D. I would be curious to know if anyone else here has such experience.
- Murray
I must admit to being in possession of a bit of a purist streak. Perhaps from that, I tend to believe that the fewer lens elements needed for the job, the better.
I have a soft focus filter (Bay 1) for my Tessar and Tessar-type TLRs. (I like Tessar and Tessar-type lenses.) However, it made more sense to me to start with a softer lens (for portraiture), rather than degrade the image from a sharper lens, so I bought a Yashica D with the 3-element Yashikor. The lower contrast, softer focus, and vignetting at wide apertures seemed to offer natural advantages.
Then again, I haven't actually done any further experiments in window-light portraiture since the early 1990s, so I never actually tested my hypothesis with the Yashica D. I would be curious to know if anyone else here has such experience.
- Murray
If you like lenses with low element count... It would seem that having aspheric-ground elements helps that... It's said that each aspheric surface can replace one element in the lens design. I suppose this could be taken to extremes, such as a one-element lens with two ASPH surfaces equaling the optical quality of a triplet?
Also of note is that the recent rash of APO primes are generously packed with elements!
Also of note is that the recent rash of APO primes are generously packed with elements!
Nitroplait
Well-known
Didn't use my Yashica D for portraits when I had one, but you are likely correct in your assumption. For my use, I had to use it at f:8 or 11 to get the sharpness I needed, but below that it would likely have worked well for portraits.
Pleasantly soft without looking artificially smeared like soft focus filters sometimes do.
Freakscene
Obscure member
MTF shows three dimensional characteristics of lenses perfectly if it is presented three-dimensionally, but it basically never is presented that way. A manufacturer typically measures or calculates MTF at 50x the focal length, or, increasingly and frustratingly, wherever MTF is calculated to look best (“our Bragicron is better than your Corvus!!!”). To show MTF this way you need either a mutidimensional model output or, to visualise it, a chart per f stop and frequency response. It makes the whole thing vastly harder to read, quite a challenge given how basic 2-D MTF charts are typically misinterpreted.
Modern glass, coating and manufacturing techniques make arguments about the number of elements moot - lens designers can use as few or as many as they choose. One overlooked factor is that the fewer elements you have, the greater the mechanical precision required for manufacture, because the same proportional misalignment causes a greater optical change in the imaging characteristics.
Marty
Modern glass, coating and manufacturing techniques make arguments about the number of elements moot - lens designers can use as few or as many as they choose. One overlooked factor is that the fewer elements you have, the greater the mechanical precision required for manufacture, because the same proportional misalignment causes a greater optical change in the imaging characteristics.
Marty
taffy
Well-known
Very interesting, thanks to OP for sharing.
Comment of @Marty is very insightful too.
Comment of @Marty is very insightful too.
Archlich
Well-known
Some point & shoot cameras have lenses with amazingly few elements.
The 28-75/3.5-8.9 lens of the Minolta 75w Riva has 4 elements in 4 groups.
The 24-50/2.8-5.6 (!) lens of the Fujifilm Silvi F2.8 has 6 elements in 5 groups.
The 28-56/2.8-5.4 lens of the Fujifilm Natura Classica has 6 elements in 5 groups.
The Minolta is quite soft with hefty falloff. No idea about the "3D quality". But I do appreciate simple designs.
The 28-75/3.5-8.9 lens of the Minolta 75w Riva has 4 elements in 4 groups.
The 24-50/2.8-5.6 (!) lens of the Fujifilm Silvi F2.8 has 6 elements in 5 groups.
The 28-56/2.8-5.4 lens of the Fujifilm Natura Classica has 6 elements in 5 groups.
The Minolta is quite soft with hefty falloff. No idea about the "3D quality". But I do appreciate simple designs.
David Hughes
David Hughes
Another low element lens was the one in front of the Minolta Riva Zoom 90. It went from 37.5 to 90mm and was just 4 elements also. Alas, I sold it one day or gave it to a charity shop. Here's a sample photo taken many years ago:-
Regards, David
Regards, David
Sonnar2
Well-known
I'm not a specialist in zoom lenses and barely use them ((at the moment I use a Tamron 70-180/2.8 AF on my Sony A7ii quite often photographing horses)) --
but I didn't knew that a zoom lens with just 6 elements in good quality is even possible.
I stay with the KISS approach. Wise words of a lens designer, especially about quality needs of a "kit lens"... indeed, the cheapest high volume lens can build or destroy reputation of a company.
but I didn't knew that a zoom lens with just 6 elements in good quality is even possible.
I stay with the KISS approach. Wise words of a lens designer, especially about quality needs of a "kit lens"... indeed, the cheapest high volume lens can build or destroy reputation of a company.
f.hayek
Well-known
1+ ^^
Modern Zeiss M-mount and SLR lenses, renowned for their image "clarity" & "3D pop" are anything but simple in design nor rely on the fewest possible elements.
benlees
Well-known
As a lover of non-bazooka cameras, it's a shame curved sensors haven't been developed to the point of marketability. My knowledge of optics is poor but I understand curved sensors would reduce the 'pressure' of a lens to perform, so a drastic reduction is size and complexity could result.
My only criteria for a lens... does it work for photography without ruining a photograph? 
Well, and size...the smaller, the better.
Well, and size...the smaller, the better.f.hayek
Well-known
Curved sensors is a great idea if the field curvature of every lens in a series can be made to curve in the identical fashion.
Wouldn't it be wonderful if sensors could be variably flexed to accommodate any particular lens.
Phil_F_NM
Camera hacker
This is a case where a system analogous to the Kodak Retina IIIC and early Retina Reflex could be employed.
Phil Forrest
Mackinaw
Think Different
Personally, I'm a big fan of Tessar-based lenses. Be it my Canon 50/2.8 LTM, the 80/3.5 Yashinon on my 124G, or the 135/4.7 Optar on my Crown Graphic. Very nice look.
Plus I wonder what role economics plays into lens designs. It would seem the fewer lens elements used, the cheaper to produce. Everybody looks at the bottom line, at some point.
Jim B.
Plus I wonder what role economics plays into lens designs. It would seem the fewer lens elements used, the cheaper to produce. Everybody looks at the bottom line, at some point.
Jim B.
zuiko85
Mentor
Speaking of field curvature does anyone here remember the 24mm f2.8 Minolta MD VFC?
A wide with a ring to control the field curvature, through convex to the film to flat to concave.
Have no idea how the designers did that but interesting nonetheless.
A wide with a ring to control the field curvature, through convex to the film to flat to concave.
Have no idea how the designers did that but interesting nonetheless.