Archived posting to the Leica Users Group, 1997/11/20

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: Re: Bokeh:a mythical construct
From: Erwin Puts <imxputs@knoware.nl>
Date: Thu, 20 Nov 1997 22:39:06 +0100

Gee, I now wished I had kept my big mouth shut.
Sorry for not answering within a day. Sometimes I have to work for a couple
of days in circumstances where my mind is not tuned to Leica stuff.
To start: I am very much aware of the fact that any lens can only give a
sharp rendering of an object in a flat plane. So any  3D object  has one
sharp plane and every other extension of that object in front or behind
that plane is unsharp in various degrees. The 'thickness' of the sharpness
plane is governed by the familiar circle-of-confusion, which is also the
basis of the various DOF tables. All these tables are based on the fact
that the permissible circle of confusion is 0.03mm. This is nowadys
hopelessly inadequate as anyone who enlarges his negatives or projects his
transparencoes can readily see for him/herself.
So the phenomenon of unsharpness as such is a fact of optics.
The way the unsharpness is rendered by any lens is a result of two kinds of
computations. First of all the computations that ensure the sharpness of
the in-focus-plane are relevant for the unsharpness regions. The better the
optical correction of the sharpness plane, so will be the rendition of the
unsharpness areas. BUT: if the corrections of the sharpness plane are
extremely well done, then the visual or perceivable diffence (that is the
gradient from sharp to unsharp) will be quite steep. This happens with all
modern Leica lenses of the recent (3) generation. Older lenses have a
smoother gradient, not because of any special characteristics  of the
unsharpness rendition, but simply because the in-focus-sharpness was not as
well corrected as it is today.
Second: any lens designer can concentrate on maximizing the characteristics
of the sharpness (more general: the rendition) of the in-focus-plane and
,speaking relatively, neglect the correction of optical defects in the
out-of-focus areas.
That is what most japanese lensdesigners do as they optimize for high MTF
values in the plane of focus. Therefore the noted diffences between for
instance Leica and Nikon/Canon.
Nothing new or mythical here. The effect of the placement of the diaphragm
and the number of blades is greatly exaggerated. The basic optical
characteristics are laid down by the optics not the mechanics. Only in
highly unsymmetrical lenses the diaphragm placement has a certain role to
play. The number of blades has an insignificant effect.
So 'bokeh' as a construct has nothing new to reveal. But as soon as we
start stating that the 35 Summicon (nonASPH) has better bokeh than the
actual Summicron 35 ASPH we make the wrong turn  from a factual view.
Certainly if we imply that Leica lenses are specifically designed to show
good bokeh (whatever that is) we are sliding into mysticism.
Just as we can rationally discuss the correction of optical aberrations in
the sharpness area, we can do the same for the unsharpness area. To make a
bold statement: if Canon (which I give a higher value optically than Nikon)
were to decide to design a lens according to Leica standards, the rendition
of the out-of-focus-areas would be identical.

>    Erwin, are you saying that you believe that all lenses of
>    the same focal length render out-of-focus areas identically
>    when opened to the same aperture?
>

No I do not believe that all lenses of same focal length focused at an
object at the same distance and stopped down to the same diaphragm render
the o-o-f areas identically. But that is as stated a function of the
corrections of the sharpness areaand the careful attention to the optical
aberrations in the unsharpness areas. Optical aberrations are not different
for the sharpness and the unsharpness areas. The same laws and mathematical
equations work here and there.

As for testing. As always you can only compare in a meaningful way any
differences when you take comparative pictures with the same film and the
same object at the same distance with different lenses set at the same
aperture. Pictures taken in these settings with the Summicron 35 (non-ASPH)
and Summicron ASPH revealed hardly any  visual difference in the rendition
of out of focus areas. The superior bokeh of the nonASPH version I failed
to note.
Erwin