This material, especially for Radozhiva, prepared Sergey Koveshnikov.
In my previous review on Nikon 135 / 2 I promised to share my thoughts on using Defocus Image Control. Let me remind you, this function is present on two Nikon lenses, this Nikon 105mm 1: 2D AF DC-Nikkor Defocus Image Control и 135mm 1: 2D AF DC-Nikkor Defocus Image Control.
In short, the whole point of DC is simply to control the spherical aberrations of the lens.
In the beginning, it is necessary for someone - to explain, to someone - to repeat the materiel about spherical aberrations. For those who are not interested, they can go straight to the "practical application" section, but nevertheless, I believe that it is impossible to fully understand and understand the DC function without delving into such a phenomenon as spherical aberrations. I will try to explain it here briefly and in simple terms. So, spherical aberration at a lens is an optical phenomenon, when light beams passing through the edge of an ordinary spherical lens are deflected more strongly than the same beams, but passing through the lens closer to its central part.
Ultimately, the beams of light passing through different parts of the lens will be focused in different places in space, in front of the main focusing plane, as a result, there is no single plane where all the beams come together. There is a point where the rays from the central region of the lens are focused and where, in an amicable way, all the others should be - we will call it "true focusing". Also, the diagram shows the area with the highest density of beams, it is located a little closer to the lens - if you place the focusing plane (in fact, the camera matrix) in this place, then the final picture in this position will have the highest contrast in comparison with any other possible positions (which is typical, even more than at the point of "true focus").
So clearly, as in the diagram shown at the beginning of the article, in real life only monocles behave - lenses consisting of only one convex spherical lens. In any normal lenses, additional corrective elements are introduced to eliminate this aberration, in the simplest case a concave lens. In this case, they say that this lens has spherical aberration corrected. Sometimes, however, due to the special design, these aberrations can become overfilled, i.e. the light beams passing through the edge will deviate less than required and, accordingly, will be focused beyond the main focusing plane.
In DC-Nikkor lenses, when you rotate the DC control ring towards F, the spherical aberrations become overcorrected (otherwise they are also called negative), when rotated towards R, the aberrations are undercorrected (positive). In these lenses, this is due to a movable group of two lenses located immediately behind the front.
By the way, as it is easy to see, the number of spherical aberrations is directly related to the aperture - after all, when aperture is used, the peripheral beams of rays are cut off, forming the “wrong” focus point, and the smaller the aperture, the fewer such rays remain. That is why the DC control ring is marked with aperture numbers: the instructions for the lens indicate that it is not recommended to set the DC ring value higher than the current aperture value. Those. for aperture, for example, f / 4, the recommended DC values are R4, R2.8, R2, 0, F2, F2.8, F4, and for f / 2 - only R2, 0 and F2.
Now, to make it clear what is happening with the picture, we’ll figure out what happens with the circle of confusion:
In the neutral position of DC, the circles of blur and behind- and in-front focus planes look like any decent lens with completely corrected spherical aberrations: the point source of light turns into a uniformly luminous circle, and the bokeh is “Hollywood”.
When the position of the ring in the R - circle, the blurriness behind the focus plane (background) increases, its edges are blurred:
The circles of blur of the foreground, on the contrary, are compressed and their faces become more clearly defined and bright, while the center of the circle is darkened, on the contrary.
When you move the DC ring to the F position, everything is basically the same, but exactly the opposite: the effects are reversed and what was happening in the foreground is moved to the background.
After each change of values, manual refocusing was carried out.
From the above, we can conclude that depending on the position of the ring, the degree of blur may increase, while, at the same time, the bokeh will become soft, "creamy", or the degree of blur will decrease, aggravated by the fact that the nature of the bokeh will begin to acquire sharpness and variegation (such bokeh also called "fish scales"). Once again, the nature of the bokeh and spherical aberrations are flesh from flesh, which is why I dwell on this in such detail.
At this stage, we have dealt with spherical aberrations for rays of the same color, but things get more complicated when we start considering the same situation for different colors:
As you know, rays of different colors are deflected in different ways, so the degree of "correction" of spherical aberrations will be different for different wavelengths (colors). Because of this, chromatic aberration and, as a result of a combination with spherical ones - the so-called spherochromatism. In other words, it is also called "fringing" and "colored bokeh" - this is a phenomenon when high-contrast areas out of focus are painted in green and magenta tones. And, if we increase the amount of spherical aberrations, due to the focus shift, we sort of "smear" the areas with colored bokeh, so that they begin to color the sharp area too.
And this is very bad, because if the "monocle effect" can be very desirable and often used for artistic purposes, then any color distortion is almost always an unwanted element of the image, it spoils the "picture".
Practical application of DC
As indicated in the rather laconic instructions for the lens - when you rotate the DC ring towards R, the background is blurred more, towards F - the front. But how much does this affect the final image? In a nutshell - almost imperceptible. If you do not set the degree of blur more than the recommended one, then in most scenes the difference can be seen only by comparing frames with and without this function. If I was asked to determine "by eye" whether the DC function was used or not in a single photo, most likely I would not be able to do this. If you set the ring further than the recommended values, a clear soft effect appears, which immediately gives out the use of DC. In addition, an increased amount of spherochromatism (aka fringing) is a "bonus" to this.
Frames shot one after another in a row, with the same settings (f / 2, 1/250, autofocus according to LiveView). The only difference was the installation of the DC ring.
Of course, the greatest influence from the position of the DC ring occurs when the diaphragm is fully open (remember from the theoretical part that the number of aberrations depends on the diaphragm) and, with its closure, the already not strong effect is completely nullified.
As already mentioned by Arkady in the review Nikon 105mm 1: 2D AF DC-Nikkor Defocus Image Control - after changing the setting of the DC ring, the focus will inevitably get lost and it is necessary to refocus again. But that's only half the story. In any position of the DC ring other than zero, focusing using LiveView and phase sensors will differ. This happens, again, because the presence of spherical aberrations distorts the image and it appears to be incorrectly focused. When focusing according to the phase principle, only the main (central) beam of rays will always be in focus, while the peripheral rays will be focused either in front of or behind the plane of the matrix. In LiveView mode, the contrast method works, i.e. the camera tries to give the highest contrast in the focusing area, and in this case it is somewhere between the point of true focusing (central beam of rays) and the focusing area of the peripheral ones. In general, when using DC, LiveView focusing (automatically or manually) usually gives more predictable and pleasing results.
In addition, when using DC, which is with contrast, that with the phase method - focusing becomes noticeably less confident, the larger the set DC value - the greater the uncertainty. Looking at the first diagram, one can easily guess why this is happening - how can you accurately focus something that does not have an unambiguous focus point by definition?
Another problem of the practical use of DC is that information about the set value is not transmitted to the camera and, accordingly, is completely absent in EXIF... Therefore, when parsing the footage, it can be difficult to understand if DC was used here and at what value; you have to strain your memory or try (often in vain) to determine it visually.
In what plots is the use of DC most justified? Of course, no one forbids you to shoot all the time only on R2, R2.8, etc. but still, I am convinced that each situation should have its own hardware settings. And the quality is declining, after all. So, the use of DC in the R area looks justified and even necessary in cases where the background has sufficiently sharp contrasting elements or the background itself is simply too colorful. It also helps to soften the rather high sharpness of these lenses, which is not always appropriate.
Often, the effect of monocle, which is achieved by immoderate (more than recommended for a given aperture) rotation of the ring towards R. This achieves the "dreaming" effect, which is often used in films and TV series to convey the feeling of unreality of what is happening. In addition, the stroke of this very ring is not limited by the aperture values marked on it, and effortlessly goes further, stopping approximately where the value of R (F) 8 could be. However, in non-recommended positions, the number of various kinds of aberrations (not only spherical) increases sharply, and the picture loses its attractiveness. One of the solutions to this problem is to convert the image to monochrome:
It is highly desirable, however, to avoid the background and the object with sharp and contrasting differences in brightness, in other words, the lighting to create the frame should be chosen as softly as possible. However, this recommendation is associated not only with optical imperfections and features of the lenses under consideration, but also in general - adhering to this rule is a good form in creating classic portraits.
Perhaps the most controversial thing in DC is the ability to set the ring to the F side.In fact, the overwhelming percentage of scenes when shooting (especially portrait) is that the subject is placed in front of the background, and the background itself (which is the first to suffer from motley blur in mode F) - occupies a significant part of the image; at the same time, the foreground is quite rare, and shots with a strongly blurred foreground, as a rule, are, unlike the background, unpleasant for the eye. For a couple of years of using such a lens, I never found a situation where there would be a real need to translate DC into F.
In general, the DC function allows you to “play” with spherical aberrations without unnecessary problems, try to use them in a creative way. It should not be expected to be of any special practical use, and to many it may seem just superfluous, but, as it seems to me, it can be a good tool for photo artists. In the end, the additional ability to somehow influence the drawing is only a plus.
Well, as a conclusion - I can’t come up with anything better than quoting Arkady from his 105mm 1: 2D AF DC-Nikkor Defocus Image Control review, joining his words:
I’m not used to Defocus Image Control, it’s a pretty thin tool that you need to be able to use.
Thank you for attention.
You will find more materials from readers of Radozhiva here.
Well written, I certainly will not buy DC)
Been waiting for this article
Thank you A very interesting material on the use of lens movements to change the optical properties.
This idea is already more than 150 years old - even in J. Petzval's portrait lenses, a movable third lens was used, which changes the character of the image. In addition to SfA, the sign of the coma also changes there - the character of the picture along the edges changes.
Most likely, the defocus control indicated here can be provided on any manual Planar lens by displacing its front lens, for example (the shift of the half affects the coma rather than the spherical aberrations) - there are ample opportunities to “play around”.
By the way, Canon has a similar lens - 135 / 2.8 Soft Focus, but in it, as it seemed to me (when I was turning it in my hands), this function is released one-sided - SfA can be changed only in one direction.
Yes, that's how it is. It is only worth noting that the modern reprint of Petzval from the Krasnogorsk plant is completely devoid of the function of adjusting the SA, attention only to the notorious "twisting" of the background.
Relatively speaking, is some unusual depth of field moving closer or further relative to the focus point?
Yes, you could say that. Plus the depth of field is becoming a little wider and more implicit.
I think "F" is needed not to blur the front, but to give some sharpness to the back. Such situations also happen, and quite often. When there are two people, one is a little further away. Or some interior detail, a bush with flowers in the back, a table lampshade, whatever. Maybe when a portrait of a head, so that a larger face is sharp, not only one eye.
Thank you for your work, Sergey!
So for this it’s enough to cover the diaphragm and that's it, no DC.
DC does not reduce aperture.
Yes, besides covering the diaphragm, bokeh and blur begin to disappear, a certain element of artistry is lost.
What kind of bokeh are we talking about, if you need to sharpen the back side? You decide whether you are either checkers or go.
Dear Andrey, here we are talking about the depth of field zone. Bokeh is located behind this zone.
I recommend that you read the bokeh article on this site before writing such passages.
Dear Peter I am well acquainted with the articles on this site. And you will argue that covering the diaphragm does not increase the depth of field? And you can shift it by changing the focus point. And I recommend that you read this PASSAGE, at least in the articles on this site.
No, sir, I do not intend to fight with you. To do this, I need an opponent smarter.
To do this, you must comply yourself, and not think that you meet the requirements. Self-conceit killed many)))
Good job +++
Fine. Thanks to Sergey and Arkady.
A wonderful article, the first time I read such a systematic, well-illustrated and thoughtful material on this topic. Thanks to the author!
Thank you.
By the way, for some reason, the article turned out to be a photo of rather poor quality. Full sizes can be found here: https://cloud.mail.ru/public/Neg9/9etdo4v2P
Thank you.
Very accessible about not the easiest things.
I read it with pleasure.
Thank you!
An interesting article, excellent presentation of the material!
The mechanism of operation of this function is well illuminated, but the practical significance of the function itself is small.
Good article, informative
Thanks to the author for the correct presentation!
Thank you for the article! And now I got the opportunity to bring the theoretical basis to my reasoning that the inaccuracy of one of my manuals is not a bug but a feature! :)
A useless function ... It is not at all clear how one could have thought of this before ... There is only one explanation - to add to the cost ... Purely selfish interest ... And there is no airiness in the picture, there is simply no sharpness ... That's all ...
Sharpness is there. Spherical aberrations do not distort sharpness (namely, the resolution) - the contrast drops, so the previews look soapy.
GIF animation clearly shows how the 10th glass of cognac looks on any arbitrary background.
I think that the control of the SA can only make sense in b / w shooting. In color, this "mystery" effect loses its appeal.
Yes, actually, that’s how it is written in the article.
Great lens. I have 105 DC
Good article. Everything is clearly described.
Prior to this, I have not even met such lenses.
I wonder if it is possible to “spoil” an ordinary manual prime lens in such a way that it gives out-of-focus circles with a bright edge?
Probably - yes, but I suppose that not everyone and that it will not be easy, like, say, conversion into a monocle.