The last thing of great importance to dimensionality, really fundamental, is that of the portrayal of objects in image space, inside the bokeh field. You don't hear much about it, do you?
(We can put to one side the issues involved in DOF and Airy Disks and think more of deep bokeh fields.)
Because they were not designed for it, many modern fast lenses do very poorly at showing differences in definition in objects at different distances in image space. But imagine if your own vision showed objects in your visual field at 3m, 5m and 10m with the same level of definition - or more accessibly described as the level of abstraction?
Excepting 'blanket bokeh' used for particular aesthetic preferences, this is a highly undesirable trait in lenses used for optimally showing the vital extra dimension of image depth. Some don't see depth and perhaps never will. The best real world analogy for the loss of image depth might be driving in fogs and mists and trying to estimate the distances of various objects - not easy. Depth perception may or may not be a learned experience.
Discussions about bokeh focus almost exclusively on matters such as: contrast levels, harshness of edges of objects, double lining, object light spreading, front and back bokeh qualities and balance, and the involvement of color (saturation, casts, etc.) and not to forget the presentation of 'bokeh balls' (asph grinding residue, edging, edge spreading, size and (the horror) the degree of non-circularity of outer frame specular highlights).
Fortunately, the recognition of the importance of dimensionality is growing rapidly, albeit from a low baseline. At the end of the day, all we that see it and want it desire are a decent range of options in the lenses available to us. No one builds houses like they used to, and it would be tragic if that fate befell photographic lenses. Because the older lenses designed by the geniuses of the last century are getting older and harder to maintain.
I love the way many Zeiss, Voigtlander lenses render and provide depth. I also believe this is a design choice they make, sacrificing chromatic aberration for depth.
This guy has gone (and is still going) down the lens design for depth rabbit hole. While he tries my patience, if you stick with him, he seems like he's thought about this a lot:
Hard for me as a layman to know if what he says is right, but it sure seems like he is directionally correct.
As an aside, he says telephoto lenses were initially made white so they wouldn't heat up as much in sunlight which risks distortion due to calcium flourite (high expansion coefficient) being used in some of the lens elements.
He has a decision tree for choosing lenses that may have more depth that he admits has limitations. Here he goes through some guesses about what characteristics of lens design produce more depth:
The decision tree is based on the number of low dispersion elements, % of low dispersion elements, and a ratio of how many elements vs groups in the design.
Zeiss Loxia 35mm. Shot at 2.8 or so. Cloudy, but the geometry lends itself to 3D. Agree pr Disagree.
Low number of lens elements.
This is a perfect example of 3D pop and I'm not surprised it was taken with a Zeiss Loxia 35mm.
I wonder if field curvature plays a big part here. You can see that the rocks on the left side of the dog are blurry, while the dog (and the rock he's standing on are sharp). Or is it just the "midzone dip" (in sharpness) that this lens is known for?
I found an article on Nikon's website where a Nikon lens designer talks about how adding more glass elements to a lens will increase vieling glare and reduce light transmission. Could this be why lenses with many elements render "flatter" images (less tonality and color)?
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. Why do you think this is? Depending upon how they are used, lenses are able to compensate for specific types of aberration. However, at the same time, they also generate other types of aberration. Therefore, 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. While some may think that increasing the number of lens elements represents the evolution of lens design, reducing the number of elements can help to rid the design of that which is not necessary. Lenses have reflective surfaces. Increasing the number of lens elements by just one increases the number of reflective surfaces by two. The greater the number of lens elements in the structure, the more veiling glare there is. This results in imaging that is not as sharp or clear. An increase in the number of reflective surfaces is also linked to increases in ghost and flare. So what if we apply a good coating? The answer is simple. No matter how effective a coating is, it can never completely eliminate reflectance. By reducing the number of lens elements by just one, reflectance for two surfaces can be completely eliminated. 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?...Show more →
jaygould wrote:
I found an article on Nikon's website where a Nikon lens designer talks about how adding more glass elements to a lens will increase vieling glare and reduce light transmission.
Interesting read. Not exactly news, though.
Could this be why lenses with many elements render "flatter" images (less tonality and color)?
Not really, as for decent modern lenses this is generally not true. Lens design is always a tradeoff. While less elements means less accumulated issues, it also means less options for the lens designer to maximize IQ. They do not add elements for the fun of it, but because the improvements due to the additional and of course well coated elements (sharpness, CA, etc) outweighs the degradation (contrast, transmission, accumulated surface roughness, etc). As for "less tonality", that is mostly a myth that may have made sense for old rough and/or uncoated glass. All (decent, modern, etc) lenses produce significantly more tonality than your eye can distinguish.
Daran wrote:
All (decent, modern, etc) lenses produce significantly more tonality than your eye can distinguish.
Nikon's lens designer (and all his mentors) don't agree with you. And what he said about loss of light and veiling glare rings true from my experience as well. So I'm going to trust Nikon over you here. Unless you can prove your statement that modern technology can overcome physics.
Zeiss Loxia 35mm. Shot at 2.8 or so. Cloudy, but the geometry lends itself to 3D. Agree pr Disagree.
Low number of lens elements.This is a perfect example of 3D pop and I'm not surprised it was taken with a Zeiss Loxia 35mm.
I wonder if field curvature plays a big part here. You can see that the rocks on the left side of the dog are blurry, while the dog (and the rock he's standing on are sharp). Or is it just the "midzone dip" (in sharpness) that this lens is known for?
You are serious, right?
If that's the "3D pop" holy grail that some folks think will make their photographs great... we're doomed. ;-)
It's my photo and it isn't a great photo. I like it because it is my dog, but I believe it shows an image that has 3D pop so I put it up. I think the lens had something to do with it.
jaygould wrote: Daran wroteAll (decent, modern, etc) lenses produce significantly more tonality than your eye can distinguish.Nikon's lens designer (and all his mentors) don't agree with you.
You need to slow down and try to understand what he actually wrote (which is correct) and how it differs from your interpretation (which isn't).
And what he said about loss of light and veiling glare rings true from my experience as well. So I'm going to trust Nikon over you here.
I never asked you to trust me. I only point out errors in your interpretation. Feel free to ignore me and keep your beliefs, if that makes you happy, I certainly won't force you to learn anything new.
Unless you can prove your statement that modern technology can overcome physics. . You don't have the first clue what physical limits of recording tonality are, do you?
For looking at 3D we mostly focus on more open scenery or urban locations, at longer focal distances. Here is a look at portrait lenses, showing a close examination of comparative images.
Two high end 35mm lenses here (a screen capture, courtesy of Dustin Abbott). Which do you prefer and why? Just to think about it a little, while I tell you one lens has five ED elements and the other one has but one ED element. I'll do a deeper dive into this very interesting topic in the 3D thread, looking at exactly what are the differences in these images and speculate as to why they differ so much.
Their designs are almost identical at the macro level - 15/10 vs 14/10. I wish there was a ready made formula to explain dimensionality from the lenses themselves rather than their images. But it's elusive, and the reason is that it's a complex matter, involving more factors that need to be picked apart - from looking at the images themselves.
two high end lenses for Sony E-mount, both shot at f1.4