[Philoslothical]

I've seen speculation that one difference, besides the number of blades, is the iris placement. Presets and other lenses without apertures tend to have apertures near mid-lens, whilst auto-aperture lenses see the iris closer to the lens base. Does this really affect bokeh? I don't know for sure.

Yep! This definitely appears to have an impact on both the quality of the overall bokeh, and the handling of specular highlights in it. I'm about 95% convinced that the iris position is the cause of that "vintage lens feel" produced by old presets, but there could also be contributing factors in the lens design. I don't know enough about that yet to say.

[Yamanobori]

Bokeh is a technical word and has a technical definition, just like sharpness. It is the rendering of the out of focus area of the image. The Bokeh should be smooth and should not cause hard-edge artifacts. Bokeh does not mean narrow depth of field--you have Bokeh wherever the image goes out of focus, even at f/11.

Now, if you like harsh Bokeh, that is simply preference. It is a valid preference, just like folks like soft-focus and plastic lenses , but the Bokeh is still "bad."

[Yamanobori]

The aperture position should be at a Fourier plane (pupil) and so should not affect Bokeh. The Fourier plane does not have any spatial information like the image plane--the only image at a pupil (Fourier plane) is the image of another pupil. When the aperture moves from the Fourier plane, it will cause vignetting--this is not mechanical vignetting which is cause by the lens barrel. This vignetting would be at all apertures which is why aperture placement is not random. (A simple lens (single element) obviously will have an aperture in the wrong place just because the aperture cannot be placed within the optics.)

Aperture shape will certainly impact Bokeh.

Vintage lenses tend to have some degree of spherical aberration which will certainly impact Bokeh.

There are no teachers inside a lens...

[Eruditass]

optics - What is bokeh, exactly?

Bokeh Test

On Bokeh

[Lowell Goudge]

Bokeh is a technical word and has a technical definition, just like sharpness. It is the rendering of the out of focus area of the image. The Bokeh should be smooth and should not cause hard-edge artifacts. Bokeh does not mean narrow depth of field--you have Bokeh wherever the image goes out of focus, even at f/11.

Now, if you like harsh Bokeh, that is simply preference. It is a valid preference, just like folks like soft-focus and plastic lenses , but the Bokeh is still "bad."

the problem with this definition is that the bokeh being "bad" or "good" is a function of whether it adds or detracts from the main subject. this in many cases is not a question at all of lens design, but of the relitive isolation through distance of the dubject to foreground and background. The problem is that the definition of good and bad is rather, ...excuse the pun...fuzzy

The aperture position should be at a Fourier plane (pupil) and so should not affect Bokeh. The Fourier plane does not have any spatial information like the image plane--the only image at a pupil (Fourier plane) is the image of another pupil. When the aperture moves from the Fourier plane, it will cause vignetting--this is not mechanical vignetting which is cause by the lens barrel. This vignetting would be at all apertures which is why aperture placement is not random. (A simple lens (single element) obviously will have an aperture in the wrong place just because the aperture cannot be placed within the optics.)

Aperture shape will certainly impact Bokeh.

Vintage lenses tend to have some degree of spherical aberration which will certainly impact Bokeh.

There are no teachers inside a lens...

the problem here is that in many instanes the aperture is not at the fourier plane. If it were, there would be no shape rendering of the aperture just change in illumination. Somewhat like the test for a mirror to see if it is ground correctly, block half of it and there shouldbe only a change in the illumination from an infinitely far source. (this is the simple test that the hubble telescope failed that would have prevented the opptical refit in space)

From my experience, certainly placement of the iris in the lens, from front oriented presets to rear oriented auto diaphrams has a big impact on the bokeh. I would agree on shape of the iris also, but only as it applies to near focal plane rendering, see my comment above about proper separation of background by distance for un-recognizable blurr.

[Yamanobori]

the problem with this definition is that the bokeh being "bad" or "good" is a function of whether it adds or detracts from the main subject. this in many cases is not a question at all of lens design, but of the relitive isolation through distance of the dubject to foreground and background. The problem is that the definition of good and bad is rather, ...excuse the pun...fuzzy

You are now speaking of personal definitions. I am referring to a technical definition. Separation of an object to a background has nothing to do with Bokeh. Just as Bokeh has nothing to do with depth of field. "Good" and "bad" refer to a defined ideal condition, not an individual value judgement.

the problem here is that in many instanes the aperture is not at the fourier plane. If it were, there would be no shape rendering of the aperture just change in illumination. Somewhat like the test for a mirror to see if it is ground correctly, block half of it and there shouldbe only a change in the illumination from an infinitely far source. (this is the simple test that the hubble telescope failed that would have prevented the opptical refit in space)

From my experience, certainly placement of the iris in the lens, from front oriented presets to rear oriented auto diaphrams has a big impact on the bokeh. I would agree on shape of the iris also, but only as it applies to near focal plane rendering, see my comment above about proper separation of background by distance for un-recognizable blurr.

Naturally, the out of focus area of an image is the result of the intersecting sensor plane being neither at an image place or a Fourier plane. This is why you can see an image of an aperture (Fourier plane) in specular highlights because the unfocused light is taking a different path through the lens as the image forming light and the image of the aperture is visible and the Fourier plane is no longer in the correct location for that particular condition. However, the lens design requires the aperture to be placed at a Fourier place for the condition of the in-focus image--there is not a lot of choice here. (Where else can you really put it?)

I have not seen anything to suggest that placement of an aperture within a lens impacts Bokeh. Nor have I read anything that states that. That would require some very controlled tests to make that determination not only in the fact you need to judge Bokeh at small apertures (hard to do).

The iris shape impacts the shape of the Airy disk. This is regardless of lens type.

[Lowell Goudge]

You are now speaking of personal definitions. I am referring to a technical definition. Separation of an object to a background has nothing to do with Bokeh. Just as Bokeh has nothing to do with depth of field. "Good" and "bad" refer to a defined ideal condition, not an individual value judgement.

the problem is that bokeh is not mathematically defined but personally defined.

Naturally, the out of focus area of an image is the result of the intersecting sensor plane being neither at an image place or a Fourier plane. This is why you can see an image of an aperture (Fourier plane) in specular highlights because the unfocused light is taking a different path through the lens as the image forming light and the image of the aperture is visible and the Fourier plane is no longer in the correct location for that particular condition. However, the lens design requires the aperture to be placed at a Fourier place for the condition of the in-focus image--there is not a lot of choice here. (Where else can you really put it?)

I have not seen anything to suggest that placement of an aperture within a lens impacts Bokeh. Nor have I read anything that states that. That would require some very controlled tests to make that determination not only in the fact you need to judge Bokeh at small apertures (hard to do).

The iris shape impacts the shape of the Airy disk. This is regardless of lens type.

there was a posting a while back of a PHD student attempting to model bokeh, and wanting test shots of the bokeh, specifically related to the uniformity of illumination in the OOF aperture shape, and the "hardness" of the edge. TO me, that suggests that there is something related to placement of the aperture. within the lens.

Also note, that specifically looking at two very similar lenses optically, a tele lentar 135F2.8 and a rikenon 135 F2.8 where the major difference was the placement of the aperture within the lens, the tele lentar being a preset with the aperture very far frorward and the rikenon an auto aperture with the aperture towards the rear of the lens, but otherwise the same elements and groups, the bokeh was quite different when shooting the same subject from the same distance at the same apertures. The preset had for lack of a better word, more "depth". By this, I mean the out of focus detail behind the focus plane seemed to be much softer and less defined, implying more depth to the image, than the auto aperture lens.

I can't explain it other than aperture placement. a big iris at the front vs a small iris at the back

[ducdao]

We need some test shots of lenses with similar optics but different iris positions and blade counts. Any volunteers?

I'll play. I have plenty of 135mm so I can try those first. Any suggestions as to the test setup (e.g. distance between subject and camera, distance between subject and background as well as what kind of background (busy, smooth, colors, etc...))?

[Lowell Goudge]

I'll play. I have plenty of 135mm so I can try those first. Any suggestions as to the test setup (e.g. distance between subject and camera, distance between subject and background as well as what kind of background (busy, smooth, colors, etc...))?

Check the 135mm lens club, and also the thread called "definitely not the flat earth society"

I did some test shots using day lillies, from the minimum focus distance of the lenses, about 4 feet, at difrferent apertures, to compare the tele-lentar 135/2.8 preset, the rikenon 135/2.8 and an SMC Tak 135/3.5

this can give you an idea of setup etc, to make similar shots, they were nothing special.

I could, if I find some more lillies (mine have died off for the year) run some shots with my SMC(K) 135/2.5 for comparison.

maybe the best thng would be a greenhouse, since we might be able to get a "standard shot" at multiple venues

[Lowell Goudge]

Vintage lenses tend to have some degree of spherical aberration which will certainly impact Bokeh.

let's hold that thought a minute.

I will modify slightly one of my observations, and I believe all will become clear (pun intended)

In looking back at some lens designs, i may have been incorrect in stating that the same optical formula but with the aperture moved to the rear of the lens on an auto aperture lens as opposed to the front of the lens on a preset.

in looking at the lens designs, they are not the same even though they state the same number of elements / groups. WHat has changed is that the optical design on many presets has the fourier plane closer to the front of the lens, than later auto aperture designs,

THe optical formula is different and as a result, the intersection of the aperture with the out of focus detail is potentially quite different also.

Other aberrations may also add to what is a more pleasing bokeh, such as the longitudinal CA seen in many lenses in what I would call the Near focus regions.

You mention spherical abberation aslo,

so the question is, have the newer designs, which in come cases have moved the fourier plane within the lens, as well as correcting abberations, led to harsher less pleasing bokeh?