Minolta 28mm f/2.0

[Dr. Harout]

I think the bokeh is somehow related to the location of the diaphragm within the lens.

[jcoffin]

The circular aperture I consider a moot point; bokeh mostly depends on other factors than the aperture's shape. Most of my Minolta AF lenses are older ones with non-circular apertures; I have only a few with circular apertures. I have yet to see any significant impact on image quality (particularly on bokeh) from circular or non-circular apertures.
-- Olaf

This guy don't know what he is talking about with regards to bokeh...the number of blades and the shape of the aperture can indeed affect the quality of the bokeh.

[ ... ]

-Sonolta

This guy doesn't know what he is talking about with regards to reading. What Olaf said was that bokeh "mostly depends on other factors", not that the aperture's shape can't have any effect at all. He's largely correct -- assuming the aperture is reasonably placed in the light path, it has to be quite a long ways from round before the lack of roundness becomes visible. Two lenses could have identically shaped apertures, and still have substantially different bokeh -- a smooth creamy bokeh vs. a harsh one is mostly a matter of correcting aberrations, not the aperture shape. With well corrected aberrations, most of the light will be transmitted toward the center, and less and less as you approach the edges, making the shape of the aperture less obvious. Many lenses, however, have sufficiently poor correction of aberrations that they produce a disk that's almost evenly lit, or has most of the light concentrated toward the edges. The more light that's "pushed" out toward the edges, the more visible the aperture shape will become.

And Dr. Harout is correct, but only to a rather limited degree. The placement of the aperture does make a difference, but you don't have a lot of flexibility about where you put it -- an aperture in the wrong part of the light path will cause vignetting.

[01af]

The circular aperture I consider a moot point; bokeh mostly depends on other factors than the aperture's shape.

This guy don't know what he is talking about with regards to bokeh ... the number of blades and the shape of the aperture can indeed affect the quality of the bokeh.

Two lenses could have identically shaped apertures, and still have substantially different bokeh -- a smooth creamy bokeh vs. a harsh one is mostly a matter of correcting aberrations, not the aperture shape.

Exactly.

And the other way it's also true: two lenses can have different shapes of apertures and still the same bokeh quality ... though many lens manufacturers will excrete marketing blurbs to tell you otherwise. Of course the apertures' shapes will show---but that's not what bokeh primarily is about. Those who believe bokeh mostly or solely depended on the aperture shape don't understand what bokeh actually is. Personally, I prefer clear, regular pentagons, hexagons, or septagons over shapes that try to be similar to circles but aren't really.

Of course, please do buy lenses with circular apertures if you prefer your blurred specular highlights round (kind of) rather than polygonal. But don't get fooled into believing this would also lead to good bokeh inevitably.

... an aperture in the wrong part of the light path will cause vignetting.

And distortions.

-- Olaf

[David Kilpatrick]

Check out what Konica Minolta has to say on the subject.

"The closer the aperture shape is to becoming a perfect circle, the more beautiful your defocused effect will be. That's why Minolta's specially-designed aperture blades produce a circular opening from their widest setting down 1.5 steps...

That's exactly the point. All lenses have a circular aperture when wide open, so the iris type is not an issue at full aperture. The Minolta circular iris only changes the shape to a circle between full aperture at 1.5 stops down - so, circular at f2, f2.4, f2.8, f3.5... and from f4 on down, the aperture is no longer a circle even with the 9-blade circular aperture iris.

But it is lot more of a circle than a five of six blade hexagon ever manages.

What makers are trying to recapture is the look of a lens built with as many iris leaves as a decent bouzouki has back staves:



This is a circa 1960 Minolta manual aperture Rokkor-TC 135mm f4 and it's really not a telephoto, more a long focus lens head on a tube, a bit like a Leica Elmar 135mm f4.5. I sold this one, maybe I should have tried to adapt it to fit the A900 if it was still around. They go for around $20-30 usually and the performance is superb, helped by the almost perfectly circular iris at all apertures, and the simple three-group design.

Oner of my shots on Alamy. It's rather funny to do my search - 'telephoto lens iris' as the only other shot which appears is, of course, a nice iris flower shot using a telephoto!

David

[David Kilpatrick]

Don, you have absolutely no idea what photography I've done. Nor did you appear to read my post. I have not said anywhere that the shape of circles of confusion is not directly determined by aperture shape. I've merely pointed out the a 9-blade or similar modern 'circular' iris may only have a significant effect - as Minolta themselves state - over a limited aperture range fairly close to wide open.

You are just weird. Ritalin needed.

David

[David Kilpatrick]

A 60-stave bouzouki... that's incredible. They must be like matchsticks in thickness, no wonder it's over $1000! All I have today is a baglamas - in bouzouki terms, a Minox

David