PDA

View Full Version : Influence of diaphragm shape on Bokeh



Thierry Schreiner
20-Jun-2008, 07:21
Hi to all,

A very quick question please.

What is the influence of the diaphragm shape on the bokeh of a lens?

By changing from an old shutter to a newer type, from a nice "round" opening with lots of blades to a "squarer" one, with less blades, am I changing the initial characteristics of the lens, especially its bokeh?




Thank's a lot for your imput.

Best regards

Thierry

Peter K
20-Jun-2008, 07:49
Also recent softfocus-lenses like the DC Nikkors are equipped with a "round" opening with many blades to get this special bokeh. So it's a good idea to repair the old shutter to avoid the change of the lense's character.

sparq
20-Jun-2008, 07:49
While bokeh and its appeal is highly subjective and is a product of many design factors, there is a simple relationship between the aperture shape and the way it renders out-of-focus highlights. The squarer the aperture is, the squarer OOF highlights it will produce.

There are plenty of resources covering that topic on internet, see i.e. http://www.luminous-landscape.com/essays/bokeh.shtml

Ken Lee
20-Jun-2008, 07:57
"By changing from an old shutter to a newer type, from a nice "round" opening with lots of blades to a "squarer" one, with less blades, am I changing the initial characteristics of the lens, especially its bokeh?"

Sans doute !

CCHarrison
20-Jun-2008, 08:37
From all of my reading on bokeh over the last 10 years, there is only a minor correlation between bokeh and aperture opening. The simplest explanation around this is this - when different lenses are ALL shot wide open, they do and can display different bokeh.

The most obvious impact of different aperture openings is when lenses are shot into a light source and that source takes on the shape of the aperture. Same thing with flare...

In the end, lens design impacts bokeh far more than aperture. I believe aperture shapes only play a very minor role in bokeh.

My on page on bokeh http://members.aol.com/dcolucci/bokeh.htm

Dan

Jim Galli
20-Jun-2008, 09:01
Yes it matters. If you have paid good money for an old lens with a pleasing bokeh character, you should not compromise it with a 5 6 or 7 blade aperture shutter.

Ole Tjugen
20-Jun-2008, 09:16
http://www.trenholm.org/hmmerk/ATVB.pdf is H. M. Merklingers explanation.

sparq
20-Jun-2008, 09:38
http://www.trenholm.org/hmmerk/ATVB.pdf is H. M. Merklingers explanation.

I've posted a link to that article already, only hosted on LL. ;)

CCHarrison
20-Jun-2008, 09:52
Jim

With all due respect, do you have sample images where bokeh was changed in any noticeable way when a low bladed aperture was used ?

thanks
Dan

Emmanuel BIGLER
20-Jun-2008, 10:16
I believe aperture shapes only play a very minor role in bokeh.

Depends on what we call Bokeh here.
For strongly defocused images, any bright point in the initial scene appears as the projection of the iris.
So if we include the shape of out-of focus bright points projected on film in the definition of bekeh, obviously the shape of the iris does matter.
See a provocative example here with a home-made star-shaped iris ! ;)

initial image (not super-sharp !)
http://farm1.static.flickr.com/127/354720922_024e496be7_b.jpg

strongly defocused and circular iris
http://farm1.static.flickr.com/133/354720559_a5e44c708d_b.jpg

strongly defocused and star-shaped iris
http://farm1.static.flickr.com/155/354720560_73e84bf2ef_b.jpg

Be for weakly defocused images and reasonably shaped iris blades, e.g. the good old 5-blade Synchro Compur pentagon, yes, I would support the idea that residual aberrations produce most of the specific character of each lens.

Emmanuel BIGLER
20-Jun-2008, 10:20
Another example by Michel Guigue (France), the strange bokeh of the Rodenstock Imagon lens is definitely correlated with the very special shape of the iris ; but residual aberrations here play a major role.
http://photo.guigue.free.fr/1_new_07_2007/imagon/imagon.htm
http://photo.guigue.free.fr/1_new_07_2007/imagon/2006.jpg

E. von Hoegh
20-Jun-2008, 10:33
Although I detest the term "bokeh" (along with all jargon); I have given a bit of thought to it.

I'm of the opinion that it's a result of residual abberations AND aperture shape. For an extreme example on either end, look at images from a Petzval and a catadioptric lens.

Ken Lee
20-Jun-2008, 10:46
If shape doesn't matter, then why not make lenses in rectangular shape to start out ?

Mark Sawyer
20-Jun-2008, 11:06
"Bokeh" has come to cover too many separate manifestations (soft-focus bokeh, swirly bokeh, curved-field bokeh, double-lined bokeh, aperture-shaped highlight bokeh...)

But specifically, "What is the influence of the diaphragm shape on the bokeh of a lens?" As already stated, internal reflections of very bright highlights can echo the aperture shape. In my observation, that's about it.

I've heard speculation that aperture shape has more effect on other types of bokeh, but I haven't seen it. Veritos are very well known for their "soft focus" bokeh, but no one seems to differentiate between the Studio Shutters (which have an almost sawtooth-edged aperture) and the barrel-mounts (which have a very rounded aperture) when buying a Verito.

The only instance I can think of where aperture shape affects more than highlight shapes is in soft focus lenses that use a very odd shaped aperture so that the lens can be closed down but still use the outside area of the lens to preserve soft-focus aberrations. The h-stops in Imagon lenses, similar stops on Fuji sf-lenses, and the old Verito enlarging stops (below, from an old Wollensak catalog) are examples of that.

http://i55.photobucket.com/albums/g139/Owen21k/veritoa.jpg


BTW, I've wondered whether an internal center-filter of the type normally used on ultra-wide lenses could be used internally like a waterhouse stop to make a slightly soft lens still softer by emphasizing the exposure from the outer, more aberration-prone, area of the lens. I know, "go try it yourself..."

Frank Petronio
20-Jun-2008, 11:08
I love the circular apertures of the nice Leica lenses, the pro Nikkors, the old portrait lenses. I don't really care about the science or calculations -- those lenses just plain work well.

Oren Grad
20-Jun-2008, 12:08
But specifically, "What is the influence of the diaphragm shape on the bokeh of a lens?" As already stated, internal reflections of very bright highlights can echo the aperture shape. In my observation, that's about it.

I've heard speculation that aperture shape has more effect on other types of bokeh, but I haven't seen it. Veritos are very well known for their "soft focus" bokeh, but no one seems to differentiate between the Studio Shutters (which have an almost sawtooth-edged aperture) and the barrel-mounts (which have a very rounded aperture) when buying a Verito.

The only instance I can think of where aperture shape affects more than highlight shapes is in soft focus lenses that use a very odd shaped aperture so that the lens can be closed down but still use the outside area of the lens to preserve soft-focus aberrations. The h-stops in Imagon lenses, similar stops on Fuji sf-lenses, and the old Verito enlarging stops (below, from an old Wollensak catalog) are examples of that.

Mark's observations are consistent with my experience.

You can use a computer to run the convolutions of different diaphragm shapes and see what effects they'll have under different focus situations. But for most users and practical purposes the only way to sort out conclusively whether the glass or the diaphragm shape has a larger effect on OOF character in general (as opposed to the shape of OOF highlight points) is to test the same glass in different shutters / diaphragms on the subject configuration of interest.


BTW, I've wondered whether an internal center-filter of the type normally used on ultra-wide lenses could be used internally like a waterhouse stop to make a slightly soft lens still softer by emphasizing the exposure from the outer, more aberration-prone, area of the lens. I know, "go try it yourself..."

And a different thought - if Nidec Copal wanted to make a few bucks, they could do a special run of shutters with a slot for waterhouse stops.

Jim Galli
20-Jun-2008, 13:09
And a different thought - if Nidec Copal wanted to make a few bucks, they could do a special run of shutters with a slot for waterhouse stops.

Interesting that many of the 355 G-Claron Barrels have a slot for those that want a pure round stop? Or perhaps there were other stops like the square ones provided with some of the Voigtlaender APO Skopar's.

No I don't have good science to back up my former statements, just what I've read in articles. My point is if you're spending a lot of money on a Heliar for example because of the lovely bokeh or an 8 3/4" Verito, would you take a chance on compromising the $$$ spent by putting it in a shutter with 6 blade aperture? I wouldn't.

Oren Grad
20-Jun-2008, 13:35
Interesting that many of the 355 G-Claron Barrels have a slot for those that want a pure round stop? Or perhaps there were other stops like the square ones provided with some of the Voigtlaender APO Skopar's.

Jim, the slot in process lenses is indeed probably there more for users who need square stops, not round :) :

http://www.largeformatphotography.info/forum/showthread.php?t=6535


No I don't have good science to back up my former statements, just what I've read in articles. My point is if you're spending a lot of money on a Heliar for example because of the lovely bokeh or an 8 3/4" Verito, would you take a chance on compromising the $$$ spent by putting it in a shutter with 6 blade aperture? I wouldn't.

If you have to make a decision in the absence of definitive comparative information, and the cost of being conservative in that way is not prohibitive, that's certainly a reasonable position to take.

Turner Reich
20-Jun-2008, 15:45
Frank, which Pro Nikkor's have the round apertures?

Dan Fromm
20-Jun-2008, 16:48
This discussion on photo.net http://photo.net/bboard/q-and-a-fetch-msg?msg_id=00Pq30 may be of interest.

Mark Sawyer
20-Jun-2008, 17:11
In the photo.net thread, someone made a very long, narrow aperture shape, and inserted it in the lens. It's the most extreme example I can think of in changing the aperture shape, and it's still hard to tell what it did. The vertical elements in the out-of-focus background *might* have been enhanced by the vertical aperture, but they were so strongly vertical already, it's hard to tell. I wish she'd rotated the camera 90 degrees and made a similar image for comparison. I also thought there should be less depth of field along horizontal lines than vertical lines, but again, it was hard to tell.

As I said, it was a most extreme example. I still doubt whether the difference between a hexagon, an octagon, or a circle could be detected except in internal lens reflections of very bright highlights showing the exact aperture shape.

sparq
20-Jun-2008, 18:48
As I said, it was a most extreme example. I still doubt whether the difference between a hexagon, an octagon, or a circle could be detected except in internal lens reflections of very bright highlights showing the exact aperture shape.

The shapes of OOF highlights have nothing to do with internal lens reflections. An OOF spot light source will always have the exact shape as the aperture.

Oren Grad
20-Jun-2008, 19:25
This discussion on photo.net http://photo.net/bboard/q-and-a-fetch-msg?msg_id=00Pq30 may be of interest.

Dan, interesting thread - thanks for posting. I agree with one of the posters there, finally something useful to do with a Meyer Domiplan! :) Hard to interpret her results, though, without having examples without the cat's eye or with it in different orientations.

Very oddly shaped irises certainly can do interesting things when the subject and focus conditions are just right. I've got snapshots filled with OOF lozenges from the funky two-bladed diaphragm of my Minox 35GL. The Olympus 35RC that I just picked up looks like it, too, is going to provide considerable entertainment value in that way.

Frank Petronio
20-Jun-2008, 21:37
Turner -- I wouldn't call the Nikkors totally round, but I think all of the current $1000 plus f/2.8 zooms have at least nine-bladed apertures. I've had their older 85/1.4 and 50/1.2 AIS lenses and they also had nine (or eleven blades maybe?) and I am sure their other fast prime lenses were similar.

I know the short, cheap AF primes that I like to use on the dslr have five blades. I just shoot wide open and they're round ;-)

Mark Sawyer
20-Jun-2008, 21:46
The shapes of OOF highlights have nothing to do with internal lens reflections. An OOF spot light source will always have the exact shape as the aperture.

You get one image of the aperture per lens surface, and they even pick up the different colors of the AR coatings. Them's reflections...

Don't take my word for it: pg. 97, "The Ansel Adams Guide, Book One, Basic Techniques of Photography" (Dr. John Schaefer, 1992/99):

"Figure 3.27: Light reflections off lens elements. When a bright light is in or just outside the picture frame, each surface of a compound lens will reflect a percentage of the light striking it. Reflections off the lens elements may project the outlines of the diaphragm blades on the film, spoiling the image..."

(BTW, this isn't the image in the book, just one I pulled off the web to explain it to my class, but you get the idea...)

Shen45
20-Jun-2008, 22:16
When I got my first Studio shutter it carried a 9" Verito and I have to say I was very unsure if it would be ok as the aperture at f8 -32 is very ragged and not really all the circular. The results are wonderful. Transitions are smooth and I cannot detect raggedy highlights. Another studio shutter I have has a 12" Wollensak and again the aperture shape is not really very good but the images show little real change if i put the same lens into a #5 Betax [Also made for the same lens] which has a very smooth aperture. I'm sure aperture plays a definite role but I agree with Jim that the original manufacturer put together a lens combination that gives a certain look that is probably best not messed with -- too much :)

Oren Grad
20-Jun-2008, 22:17
Turner -- I wouldn't call the Nikkors totally round, but I think all of the current $1000 plus f/2.8 zooms have at least nine-bladed apertures. I've had their older 85/1.4 and 50/1.2 AIS lenses and they also had nine (or eleven blades maybe?) and I am sure their other fast prime lenses were similar.

I know the short, cheap AF primes that I like to use on the dslr have five blades. I just shoot wide open and they're round ;-)

Frank, my 17-35/2.8 AF-S has a lovely, near-circular diaphragm. But I'd characterize the bokeh as tolerable. It's certainly nothing special; I've used plenty of lenses that are nicer in that respect, including many with simpler, more overtly polygonal diaphragms.

aduncanson
20-Jun-2008, 22:54
I bought a 240mm G-Claron in a barrel to use as an enlarging lens. I got a glimpse of the aperture and was horrified, there were no more than five blades and they appeared to be slightly convex rather than concave, meaning that rather than approaching a circle the clear aperture looked a bit like a five pointed star. I thought about it and realized that in enlarging, or any typical process application where a planar subject is focused onto planar sensitized material, it did not matter. Nothing is out-of-focus. I quit worrying.

Cesar Barreto
21-Jun-2008, 06:39
I've been working on a series, shooting macrophotos from old coins and have used different solutions, among them enlarger lenses reversed and mounted on an old Compur shutter. It did work out fine, but I was getting increasingly unsatisfied with rendering on out of focus areas, with their clear pentagons shapes and some odd effects inside them.
So I tried to use the shutter diaphragm, wich was some 2cm above the lens and the result was really funny.

Mark Sawyer
21-Jun-2008, 13:40
...Another studio shutter I have has a 12" Wollensak and again the aperture shape is not really very good but the images show little real change if i put the same lens into a #5 Betax [Also made for the same lens] which has a very smooth aperture...

Steve, I'm curious how you mounted a lens from a Studio Shutter (where the shutter has male threads) to a Betax (where the shutter has female threads). I'd like to move couple of my lenses from Studios to Betaxes, but never knew there was a way. How was it done???

Thierry Schreiner
24-Jun-2008, 15:54
Dear all,

Thank's a lot for your respective and respected inputs.

Lots to learn from you guys out there, the same as always.

Best regards from France

Thierry:)