Re: Boke(h): a mythical construct

[Roberto Giaccio <giaccio@dis.uniroma1.it>]

Thomas Kachadurian wrote:
> ...
> But compare the 60mm lens on the Fuji GS645 to a Leica or Canon lens, 
> you'll see the difference.
>

This is interesting.  Recently, after some discussions about bokeh on the 
LUG, I was looking to the brochure of the Fuji GA645, which I think has the 
same lens as the Fuji GS645, and noticed how bad-looking were the 
out-of-focus areas in their photos.
I know that typographical reproduction may have changed things a little, 
but it was *evident* for me me that almost-in-focus areas were rendered 
similarly as a mirror lens.  For instance, almost-in-focus lines were rendered 
as double lines, where the border of the almost-in-focus area was brighter to 
the center.
This behaviour was visible only in almost-in-focus areas, so I can say that 
this lens seems to exibit a bad transition from in-focus to out-of-focus 
areas.

Furthermore, it seems to me that the work bokeh has been used to describe 
two different characteristics:

1. the rendition of out-of-focus areas;

2. an apparent increase of sharpness derived from a sort of optical
   trick used in the past by Leitz, expecially at the beginning
   of producing camera lenses; this has also been called "leica glow"
   in past messages.

I propose to assume for bokeh definition 1. in the future.

Finally, I don't understand why it can be described, measured and 
controlled; if different lenses, as if is evident for me, exibit different 
rendition of out-of focus areas, maybe the lens fingerprint can be measured 
looking at the light distribution, on the focal plane, of out-of-focus 
light points, for increasing distances: maybe a lens with good bokeh have 
just some smooth unimodal distribution with variance smootly increasing 
with the distance, while a lens with bad bokeh does not.
It may be that the fine rendition of textures that some leica lenses exibit,
is due to the fact that their light distibution in almost-in-focus areas
follows a bell curve: textures often have repeating patterns, and if
for instance, the light distribution has two peaks, the out-of focus
image becomes fuzzy when the distance between patterns is about the
distance between the peaks.

I hope that my message is not totally uncomprensible, or worse,
totally wrong.

Regards,

				Roberto Giaccio