PDA

View Full Version : Let's Cut the Bokeh...



Scott Rosenberg
28-Feb-2005, 19:26
allright, i've been doing a lot of reading, a little thinking, and some head scratching. a recent thread (http://www.viewcamera.com/forum/viewtopic.php?t=68&sid=d9317220beaa7349dabfc061a724f2bd) on the view camera forum got me thinking about the whole concept of bokeh, which i've heard translated to bull$hit or senile. regardless, i'd like to hear what your theories are as to the factors that create this nebulous quality in the oof areas.

i'll mention the theories i've come across thus far...
1. the number of blades in the iris of the older shutters create a softer image in the oof areas.
i've come across this one several times, but then what of all the old lenses that have been moved into modern shutters? would this not make all lenses in, say copal shutters, identical in this regard?

2. before lenses were APO corrected, the red, green, and blue light all came into sharp focus at slightly different points, creating a tunnel of sharp focus, rather than a discrete point of sharp focus.
i can accept this in theory, but do not know enough about optics to form a real opinion on this one.

any other theories? could it just be a characteristic of the glass... just like the contrast, sharpness, or color cast is a characteristic of a particular lens?

any thoughts, theories, or ideas on this? i think it's high time we cut the bokeh and get to the bottom of this!

scott

David A. Goldfarb
28-Feb-2005, 19:43
The shape of the stop determines the shape of out of focus highlights, and round ones look a little less unnatural than pentagonal ones, I suppose, because the eye has a fairly round stop, but I don't really think this is a major factor that affects the overall look of out of focus areas (except in the case of mirror lenses that have a ring-shaped stop).

Usually overcorrected spherical aberration is cited as a cause for "bad bokeh," which manifests itself in various forms, like double-lines in out-of-focus areas or out of focus highlights that are brighter at the edges than they are at the center.

With some lenses, like the Verito, uncorrected chromatic aberration takes some of the edge off even when the lens is stopped down sufficiently to reduce the effect of uncorrected spherical aberration. Chromatic aberration may not have been so visible, though, in the days of orthochromatic films, and it can be corrected by using a strong monochromatic filter with B&W film.

It's only recently that the quality of the out of focus image has been called "bokeh," but if you read historical materials about soft focus lenses, all these issues are mentioned. They're just called "distracting double lines" and "cluttered backgrounds."

Jim Galli
28-Feb-2005, 21:25
"i think it's high time we cut the bokeh and get to the bottom of this!"

Scott, It's a bottomless pit. You've entered into the shifting sands of subjective opine here and there are no rules. No science. Just trial and error. And lots of opinions.

Oren Grad
28-Feb-2005, 21:55
The number of blades in the iris has some effect on the bokeh, but it's not anywhere near the whole story. The four major modern lens lines - Rodenstock, Schneider, Nikon and Fuji - all look different in the OOF areas, even though they use identical Copal shutters.

Bokeh is determined by the sum total of the optical design - which is to say, the combination of glass types, the number and surface curvature of the lens elements, and their grouping and spacing relative to each other. The handling of spherical aberration is certainly an important part, but tradeoffs in correction of other aberrations matter as well.

I don't think anyone has published a rigorous and exhaustive account of the optical design factors that determine bokeh, at least not in English. But if you haven't already done so, you should read the article by Harold Merklinger that appeared in the special bokeh feature that appeared in Photo Techniques magazine back in 1997:

http://www.trenholm.org/hmmerk/ATVB.pdf

David is mostly right about the relatively recent emergence of the term "bokeh". That's true for its widespread use in English, which I suspect can largely be traced to the PT feature. The Japanese usage, however, goes much further back. There's probably an interesting historical tale to be told about its true origins, waiting to be uncovered by someone who can read the Japanese literature going back many decades.

Alan Davenport
28-Feb-2005, 22:03
... the view camera forum got me thinking about the whole concept of bokeh, which i've heard translated to bull$hit or senile.

Good luck with this question. Most of the threads I've ever seen on "bokeh," ultimately translated to one or the other.

Peter Galea
28-Feb-2005, 22:07
Doesn't this belong in the Leica Forum?

Oren Grad
28-Feb-2005, 22:25
I have to disagree a bit with Jim.

There is science, even if it's not complete - just read Harold's article for some excellent science that goes a long way toward demystifying the effect, even if it doesn't account for all of the subtleties.

Selecting a lens with pleasing (for you) bokeh is also more than just trial and error, but as with other aspects of selecting equipment, you do need a bit of experience under your belt in order to have a language to describe it. It is, of course, exceedingly difficult to directly characterize the "look" of a lens is an unambiguous way - one ends up using vague adjectives that don't communicate well to someone who doesn't already know what you're talking about, and which understandably elicit a fair degree of skepticism. But once you know the "signature" of various lens types, it's not hard to characterize a newly-encountered lens by analogy to known types. If someone who pays a lot of attention to these things tells me that a lens looks, say, more like a Fujinon-W than an Apo-Sironar-N, or more like a coated Goerz American Dagor than an Apo-Symmar, I know *exactly* what he's talking about, even if it's hard to put into words.

If this sounds strange, remember that the same is true of many other everyday characteristics that we take for granted. For example, try describing a new flavor without drawing analogies to foods that you and your audience have both already tasted. Food scientists have a language for the elements that go into a given flavor that can tell part of the story, but for the rest of us, a well-drawn analogy is pretty much the only way we have to communicate what we mean.

Emmanuel BIGLER
28-Feb-2005, 23:43
There's probably an interesting historical tale to be told about its true origins, waiting to be uncovered by someone who can read the Japanese literature going back many decades.



At least some information about the origins of the word 'bokeh' has been presented here on the RUG in January 2004 by Todd Belcher



"Bokeh : The shape of the out of focus area of a photo." (Peter Kotsinadelis)



Then, Todd Belcher argues and explains the Japanese origin for this word:
(RUG, January 2004) http://digistar.com/rollei/2004-01/2521.html http://digistar.com/rollei/2004-01/2532.html



"I don't think that's the definition of Bokeh, Peter. Though that is
definitely an aspect of Bokeh. I think a definition of Bokeh would be
more along the lines of: the ability of a lens to render gradations. It
is considered good Bokeh when the gradations are smooth and poor Bokeh
when the gradations are harsh to the eye. The Harshness can be caused
by, as you say, out of round apertures, but can also be caused by the
lens itself. Symptoms of the latter are hard gradations and double
image out of focus areas.
Bokeh was coined by the Japanese and comes from the Japanese word for
gradation: 'bokashi'. The Japanese have, for many centuries, been
concerned with gradations in art, thus, it was not a big step to
consider it's implications in photography....
...
It is indeed related to Ukioye, Japanese woodblock printing, where
gradations were difficult to produce. The pressing of each sheet to
the woodblock had to have the ink wiped on to the woodblock 'just so'
to attain a nice gradation. In some areas, such as the face of a
character, very small areas may have had gradations and these required
a highly skilled printer to apply with consistency. The Japanese
prized the skill with which these gradations were made and effect they
had on the final image. these gradations made an appearance in
Japanese woodblock prints towards the end of the 1700s and early
1800s. --Todd Belcher"

Ole Tjugen
1-Mar-2005, 02:58
"Trotz scharfer Zeichnung bei hervorragendem Auflösungsvermögen hat das Heliar keine kalte und harte, sondern eine duftige Schärfe, die den Übergang von der schafen zur unscharfen Zone mildert, die keinen Bildteil unangenehm hart hervortreten lässt, sondern dem ganzen Bild eine harmonische natürliche Abrundung gibt."

> Despite sharp results with outstanding resolving power the Heliar does not have cold and hard, but a fluffy sharpness, which moderates the transition from the sharp to the indistinct zone, so that no picture part will stand out as unpleasantly hard, but gives a harmonious natural rounding to the whole picture. <

Advertisment for Voigtländer's Heliar lens in "der Satrap", March 1933.

Good Bokeh is nothing new, and lenses have been designed with that in mind for a long time. Only the word is recent.

Steve Hamley
1-Mar-2005, 05:11
Anyone that doesn't believe in bokeh contact me off line and I'll buy those lousy defective Heliars of yours at 1/4 the going price....

Seriously though, I do agree with Jim, not because bokeh can't be defined or illustrated, but because "bad" bokeh bothers different people to different degrees - some not at all and others it drives nuts. I tend to fall in the latter camp; looking at a picture with truly bad bokeh makes me think my eyes are crossed.

But there's another argument that "not so bad" bokeh with the beginnings of the double image acts like an unsharp mask in Photoshop creating an appearance of sharpness. Maybe, maybe not, but an interesting thought.

Steve

Jason Greenberg Motamedi
1-Mar-2005, 07:33
When I first read the word "bokeh" I thought the writer had misspelt "bouquet". To this day, I continue to think of the bokeh of a lens like the bouquet of a wine--I am fascinated by it, but don't really understand it. Beyond this, the language and vocabulary which is often used to discuss both bokeh and bouquet tend to obfuscate (!) the subject, making entry difficult for laypeople.

That said, there are clear differences between the ways that older lenses, particularly portrait lenses, render out-of-focus space. There seems to be a relationship between optical "correctness" and bokeh. To a limited extent, the poorer the corrections, the "better" the bokeh. The examples above, the Verito and Heliar, are certainly less perfectly "corrected" than a modern Plasmat, and their bokeh is both identifiably different and often judged more attractive. Petzval type lenses, which were used in the 19th and early 20th century for portraiture have a very easily identifiable "swirly" bokeh, which looks unlike any other and makes the f/64 crowd cringe.

For the record, "bokeru", not "bokeh", means senile.

David du Busc
1-Mar-2005, 07:41
I believe we've entered the circle of confusion.... so let's talk about that too.
Tally Ho!

paulr
1-Mar-2005, 12:18
one thing to remember: since bokeh has to do with areas that are out of focus, it will make little difference to you if your style of photography involves using movements and small apertures to bring everything as close to in focus as possible.

i don't usually have a lot of visibly unfocussed things in my pictures, so naturally i tend to care more about how the lens makes the focussed parts look. but this is really just about stylistic choices.

i read an excellent article about bokeh, which attributed much of what the bokeh fetishists talk about to the way spherical aberation is corrected. the article (which i'd link if i could remember where it was) showed different examples of out of focus points of light against a dark background. in all cases, as you'd imagine, the point becomes a circle as it's defocussed. but the character of that circle is different from lens to lens. in a lens that has "perfectly" corrected spherical aberation, the circle is uniformly shaded and has a smooth border against the background. this is, of course, perfect as defined by modern optical engineers. the bokeh people consider this to be only ok, in general. in the lenses with slightly undercorrected spherical aberation, the point was brightest in the middle and faded gently into the background. most of the bokeh people like this the best. in a lens with overcorrected spherical aberation, the effect is the oposite, and actually looks like a circle of light with unsharp masking applied: it has a bright halo and an abrupt transition to the dark background. few people like this effect.

here's the catch: if a lens has undercorrected spherical aberation (and smooth transitions from light to dark) in areas that are farther than the focal plane, it will tend to have overcorrected spherical aberation (and abrupt, ugly bokeh) in areas that nearer than the focal plane. and vice versa. so it's unlikely that you'll have a lens that looks great both in areas that farther and closer than the focal plane. this may be a reason people have different lens preferences: anyone who tends to defocus the foreground will like a different set of lenses than someone who tends to defocus the background.

David A. Goldfarb
1-Mar-2005, 12:24
In the case of Petzval lenses (I just acquired one myself) the swirly look is also coming from the very visible curvature of field associated with Petzvals. Since they were designed as fast portrait lenses (f:3.6 or thereabouts), sharp corners could be sacrificed for speed.

Oren Grad
1-Mar-2005, 12:35
Emmanuel - Thanks for sharing that fascinating post from the Rollei board. I did a bit of homework with my Japanese dictionaries, and discovered that the same character (kanji) is used for both "bokasu" and "bokeru" - that is, it's used to express both the concept of tonal gradation, and the concept of blur or being out of focus.
That's pretty remarkable, I think. Although they are different concepts, together they go much of the way toward capturing what's distinctive about the character of a lens. In current usage in the Japanese photo magazines, sometimes bokeh is used to refer strictly to the character of the OOF blur, but sometimes to refer to the optical character overall, including the smoothness of tonal gradation.

Jason - Although they are homonyms, the kanji used for bokeru "senile" is different from the one used for bokeru "blur" / bokasu "shade off", though there may some overlap in actual usage.

Paul - From your account, the article you're referring to is probably Harold's - see link above. Your reminder about the tradeoff between background bokeh and foreground bokeh is a good point. Regulars here may remember that I'm a big fan of the Apo-Sironar-N and-S lenses, which to my eye have exquisite background bokeh. But the foreground bokeh, while generally tolerable, is in many situations not nearly so pretty. There are a few lenses that manage to get around this, though - Mandler's 35 Summicron-M (the last pre-ASPH version), for example, trades off other forms of correction to deliver bokeh that's pretty well controlled in both foregrounds and backgrounds. I woudn't be surprised if there are some older LF lenses that achieve the same. David - what's the Heliar like in this respect?

David A. Goldfarb
1-Mar-2005, 13:35
The Heliar is pretty well controlled in front and in back, but I tend to put more background out of focus than foreground, and my instinct is that it's smoother in back than in front. I couldn't find a handy LF Heliar shot that I've scanned with OOF areas in front and in back, but this is close--a shot on 35mm with a converted 100/3.5 Medalist Ektar, which is a coated Heliar type:

http://medfmt.8k.com/mf/medalist/moth.jpg



and it's pretty sharp in the sharp area for a handheld shot--



http://medfmt.8k.com/mf/medalist/mothdtl.jpg

Bill Smith
1-Mar-2005, 20:29
I think the literal translation of boke is "blur".
This whole subject is boring.

paulr
3-Mar-2005, 09:21
Thanks for the clarity, Bill. for a minute there I thought it was interesting.

tim atherton
3-Mar-2005, 09:25
"I think the literal translation of boke is "blur". This whole subject is boring."

quite possibly almost as boring as "what movements do you use on a technika" or "what dos Angulon mean"....

Scott Rosenberg
3-Mar-2005, 09:41
but not so boring as to keep you from reading the thread and posting a reply. i really wonder about the folks on the forum that spend their time replying to threads without actually contributing to the exchange of information therein. just practicing your typing i suppose? why not simply send yourself emails - at least then you're only wasting your own time.

for those of you that actually posted with useful information, i appreciate it very much - the links were very informative. those, along with some others i came across, have really cleared up the whole bokeh thing for me. it’s actually quite a fascinating study, at least I think so, and looking into it has really taught me some new things about optics and the developments made over the last century.

again, thanks for the education,
scott