These two above feature (i) distorted size near-far subject matter, and (ii) leading lines. This is why lenses wider than around 35mm are best omitted from the 3D discussion. Any WA can make them, any one at all.
Some claim that low element count can be helpful in designing lenses with 3D pop. Wish I knew the validity of this claim, but one of the more reasonable comments on this is given by Erwin Puts in his Lens Compendium:
“The improvements in image quality of Leica lenses, introduced in the last fifteen years, are visible for anyone with an experienced eye. Some Leica users might defend the idea that the older lenses are as good as, or even better than the current ones. Image characteristics, like 3-dimensional representation, plasticity and smooth unsharpness gradients, cited as being instrumental for the unique fingerprint of older Leica lenses, are supposed to have disappeared. Some observers have tried to discern two different schools of design theory, the German and the Japanese philosophy, the latter putting all emphasis on image contrast and the former school going for a smoother image quality. In all fairness, these two schools do not and have never existed in this extreme juxtaposition. Of course designers differ in their balancing of aberrations and the rigor of aberration corrections. But generally any designer everywhere has the same range of tools and uses the same aberration theory. Why and how are the current Leica lenses improvements on the older designs and how do they differ from others? Modern optical design is a creative activity, based on experience, insight, a very thorough knowledge of optical theory and even a sprinkling of luck. Leica lens designs will start a model with the absolute minimum of lens elements. This approach is derived from the Berek adage that one must know aberrations before one can correct them. Aberrations are generated by lens surfaces and the fewer lens elements you have, the better is your insight. A corollary would be that it is smarter to avoid generating aberrations so that you do not have to correct them at all.”...Show more →
Ultraluminal wrote:
Some claim that low element count can be helpful in designing lenses with 3D pop. Wish I knew the validity of this claim, but one of the more reasonable comments on this is given by Erwin Puts in his Lens Compendium:
I saw that too but I haven't seen anyone substantiated that. Leica lenses tend to have less elements than the equivalent Zeiss but Zeiss is known to have more micro-contrast than Leica.
It would be fun to have a beer or two with top lens designers from Zeiss, CV, Leica . . . and get some answers.
For those who haven't read Erwin's book, he goes on to write:
“The rule of thumb is to keep the design as simple as possible. Higher-order aberrations are often induced by lower-order aberrations generated by the surfaces. A high level correction of the basic aberrations is more effective than introducing complexity. It is possible to construct an eight-element lens with spherical surfaces to be corrected such that all third- and fifth-order aberrations are zero. The same goal can be accomplished with a three-element lens with four aspherical surfaces, but this lens is expensive and difficult to manufacture. The solution space for a designer has these designs as extremes. The basic approach is not to try to correct the higher aberrations, but to prevent them occurring. The computer is indifferent to this rule and will happily produce high-order aberrations and will try to correct them. The main question for the designer is not to correct the higher-order aberrations but to ask why they exist in the first place. The answer will shape the lens design.”...Show more →
These articles are not very helpful and in fact, are erroneous. This outfit appears to provide tuition but the writer would be well-advised to stay well clear of the technical aspects of lens performance. It's important to realise that MTF (the technical term for micro-contrast measurement) can only describe lens performance at the focal plane.
We can hopefully agree that the two major definitions of three dimensionality are:
(i) subject-background separation - the definition pushed by modern industry propaganda to persuade you to buy and use very fast lenses at/near wide open to obtain 'the 3D look'; and
(ii) 'full DOF' imagery which needs apostrophes because once off the focal plane you can only work with longitudinal optical fade from best sharpness values (as noted in para 1). The latter is approximated the notion of circles of confusion and airy disks, and is the basis for the (really poor) hyperfocal focus method.
'Contrast', as in macro contrast, has nothing to do with micro-contrast. The latter is all about the measurement of structures or motifs in our images that can just be seen or which are just out of sight. Yes, not quite visible content is important to the perception of sharpness at the focal plane.
But the key point is that the two main notions of 3D have little to do with micro-contrast, because both types of 3D involve the depiction of content off the wafer-thin focal plane. This off-plane content will however be strongly influenced by other lens attributes, such as: field curvature, rates of focus fade, the depiction of tonal range inside bokeh, colour presentation, colour/contrast integration and scene attributes like Subject Brightness Range. It's obviously complex.
Some of the above helps explain why older lenses may portray 'full DOF' 3D much better than many modern 'sharper at the focal plane' lenses. A lot is also explained by optical priorities, and since bokeh qualities (balls, lack of onion rings, smooth edges, low contrast etc.) are currently paramount along with very sharp focal planes, designers are not attending to definition (ii) above, with often unintended consequences for typical full depth images.
People in the last century used to shoot a much larger proportion of 'full DOF' images, so designers gave higher priority to those needs. Users had prints made to the limit of film res, or viewed slides, and didn't have 100% viewing at their fingertips, obviously. And 100% crops tell us nothing about full DOF 3D, which is a full image attribute. Times change, and lenses change in line with what the industry wants to feed you. And even there, makers all have different predilections, methods and abilities.
I've personally come to the conclusion that for me the main factors are:
* Not too shallow depth of field with clearly discernible background, and a smooth and visible transition (so not 135/2 headshots with far away blurred bg) -- this is maybe because of how we see the world day to day and thus looks familiar. Examples, are the x100f shot of the woman in the cafe/restaurant with building in background, peacock on the bench, or the veteran with chocolate lab (though the dof is almost too shallow and makes the foreground look a bit cut out), or the pictures posted earlier (or different thread?) With GF 45/2.8.
* Good micro contrast (maybe because of next point).
* Foreground/background contrast difference (in-focus vs out-of-focus).
Strangely, at least for me, busy bokeh often seems to enhance the 3d subject pop (or is it the other way around and do the characteristics that create 3d illusion at the focus plane lead to busy bokeh?).
Wide aperture tele lenses create lots of bg blur and separation and wide angles often leading lines and exaggerated foreground relation but I find images most 3d at focal lengths that are normal-ish, 28-65 or thereabouts, at moderate apertures.
Oh, and the right kind of field curvature seems to help, see Zeiss 28/2 and 35/2 imo.
Thank you Philip PJ for your enlightening and well written thoughts on "the 3D look." As a newcomer to digital photography I have chosen Fred Miranda as a go to site because there are so many knowledgeable people here who are dedicated to the art of photography and are ready to take a lot of their precious time to carefully read and respond to the posts of others. And what is extra cool is that this passion does not turn (at least not often!) into bomb throwing, which seems to be all too common on some other photo blogs.
As I try to digest what is going on with the 3D look, I am quite certain, as others have pointed out, that the nature of the mind is one key ingredient. After all, the image is ultimately a part of your mind. For example, I am currently reading a neat book by Jeff Schewe called "The Digital Print." In it he describes how many professional photographers develop the ability to see very fine gradations in color that most other people cannot see. And so one might imagine that such a person could see a 3D look in a photo while others might not see it. There was also some interesting posts in this thread about the 3D look in paintings. So for fun, if you have the time, you might enjoy the 10 minute video below. It talks about the "resolution" of the human eye and bit about how the mind sees.
This thread reminds me of similar discussions in large format blogs about contact prints vs digital prints, discussions that were launched in 1985 when the Iris printer came out and then of course later when Epson's high-resolution inkjet printers became available. The thing about contact prints is that they have a look, a depth, that you have to see to understand. You can't just take a picture of a contact print and post it online. And as awesome as digital prints are, I've never seen one that can capture that look. And this look is not so dependent on the lens: pinhole photographs (where every part of the image is more or less in focus) will have the same feeling, as if the photo itself has depth, as if you are looking through a window at a strange 3D monochrome universe. But I think there are some commonalities between contact prints and this thread on the 3D look in digital imaging. I will quote DrPablo from the archived Fred Miranda thread "Where does the 3D look come from?" (it's a very long thread with many interesting posts):
But honestly, format size in my experience has a lot more to do with it. The reason is that gradual transitions of color and tone are recorded in such tremendous detail and with such subtlety, that the picture really comes to life. Medium and especially large format exhibit this effect almost as a routine.
A lot of people are showing this 3D look using photos that have shallow DOF, but I think it's actually even more evident in well-chosen shots with infinite DOF. I can tell you that the 8x10 contact print I made yesterday, with infinite focus, absolutely has that 3D effect.
It's a bit of a fad for people to go for completely obliterative bokeh, but I think it's compositionally sloppy a lot of the time. It takes the subject completely out of context, fails to create interplay between foreground and background, wipes away suggestions of depth, and rather than looking lifelike it just looks imprecise and isolated.
I do concur with others that transitions in and out of focus help the 3D effect, but that doesn't answer Grant's and my point that you can see that 3D look even if the entire image is in perfect focus. Obviously DOF has nothing to do with it in that case.
The more I look at images described as 3D, the more I think that having real luminosity (with texture) is key. ...Show more →
Yes, I think part of it is a matter of the sheer abundance of detail in larger formats. Irrespective of subject size, a larger format will give you an image that is captured much closer to life size, so a print will have more of the textural details we expect to see. Every transition, including gradations of tone, color, focus, and texture, is projected over a much larger recording surface, so the subtlest real life details seem to make it into print.
I do think that luminosity has a lot to do with the 3D effect, which means having textured highlights. Placing important details on Ansel's zones 6 and 7 is a very good strategy, especially in the context of a long transition from deep shadows through midtones....Show more →
The phrase "I know it when I see it" is a colloquial expression by which a speaker attempts to categorize an observable fact or event, although the category is subjective or lacks clearly defined parameters. The phrase was used in 1964 by United States Supreme Court Justice Potter Stewart to describe his threshold test for obscenity in Jacobellis v. Ohio.[1][2] In explaining why the material at issue in the case was not obscene under the Roth test, and therefore was protected speech that could not be censored, Stewart wrote:
I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description ["hard-core pornography"], and perhaps I could never succeed in intelligibly doing so. But I know it when I see it, and the motion picture involved in this case is not that.[4]
The expression became one of the best-known phrases in the history of the Supreme Court.[5] Though "I know it when I see it" is widely cited as Stewart's test for "obscenity", he never used the word "obscenity" himself in his short concurrence. He only stated that he knows what fits the "shorthand description" of "hard-core pornography" when he sees it.[6]
Stewart's "I know it when I see it" standard was praised as "realistic and gallant"[7] and an example of candor.[8] It has also been critiqued as being potentially fallacious, due to individualistic arbitrariness.[9][10]
This simple phrase, embedded in a plurality opinion, carries with it many of the conflicts and inconsistencies that continue to plague American obscenity law. In effect, “I know it when I see it” can still be paraphrased and unpacked as: “I know it when I see it, and someone else will know it when they see it, but what they see and what they know may or may not be what I see and what I know, and that’s okay.”
It looks 3D to me, falling into the category that Philip PJ describes as
(i) subject-background separation
And the warm colors up front and the cool colors in the background also help to support the 3D look: darrellc wrote in this thread
Interesting quote from that article, as one could possibly apply the same strategy to photography by looking for these patterns in the natural or man made environment or perhaps enhancing in post: “Some painters, notably Cézanne, employ "warm" pigments (red, yellow and orange) to bring features towards the viewer, and "cool" ones (blue, violet, and blue-green) to indicate the part of a form that curves away from the picture plane.”
Also, the bright foreground leads to the base of the building, which then recedes (perspective) into the inky dark high sky.
1. This f8 75mm image shows some characteristics in what you might look for in 'full' DOF 3D images. Focus is on the front building: object definition remains strong and 'photographic' to the front, and covers the rear located buildings even as the distance from the focal plane increases. Colour is very lifelike, shows excellent and consistent integration with tone gradation (local contrast).
This shows the importance of the quality of lens imaging well off the focal plane, which is where 3D appears or does not. No one looks at it, from what I read. I sure do.
Wide DR (a7rII at ISO 100, 42mp) keeps most content in the middle two thirds of the range, where colour is expressed most. Colour takes a lot of boosting with this lens before over-saturation, great for low colour output (CMYK printing).
Contour shaping - the faithful shaping of objects on axis (towards/away from camera) across the frame in 3D space - is satisfyingly good, buildings are as the eye saw them as they appear to the back/sides of the ridgetop. Fine detail is well drawn but not with overly high acutance (old Carl Zeiss look). Light processing.
Thank you for those wonderful examples. And since this is a thread on lenses that can give that 3D look, I'd like to give a shout out as well to the 75mm Color Heliar f/2.5. For one thing, perhaps partly because it doesn't get a lot of love from lens reviewers, it's cheap! For those who are interested in reading more about this lens, here are a couple of links from the many that I've collected over the years, reviews that may touch a bit on what Philip was talking about above:
In the link below the reviewer writes
This lens has a very smooth transition between in and out of focus and the colors just blend together. I find it very pleasing in general . . . . The in focus areas are very well defined followed by just smooth out of focus backgrounds . . . . Yes, the Leica Summarit is similar, the Leica summicrons are amazing as well. They are not $300 however. What lens can you go buy today for $300 that will be so amazing? . . . . The Color Heliar is amazing from wide-open . . . . This is one of the few lenses I have ever used that I cannot really find fault with, especially when compared to any alternatives. Voigtlander might not have the reputation of Leica when it comes to lenses but this lens is honestly fantastic. A true gem amongst lenses. I couldn’t pinpoint what was so different at first, but I know now. I absolutely am in love with the rendering. I especially love the rendering in low contrast lighting. So much so that I began a project that completely relies on the look that only this lens has given me. I will have an e-book available soon that showcases this lenses artistic rendering. Every time I look at my files from this lens I am constantly surprised at how details come out, or how smooth areas are, etc… This lens gives great compliment to any lens kit. It is so small and light, that I never notice I have a 75mm lens with me!...Show more →
I’ve had ample opportunities, here at B&H, to try many of Sony’s own offerings, as well as other native E-mount lenses, and am typically impressed by how good and consistent pretty much every lens is nowadays. And maybe that’s the problem. I enjoy using lenses for their character—a somewhat indescribable quality that cannot really be evaluated or shown in tests. I’m interested in rendering, not necessarily perfection.
