Archived posting to the Leica Users Group, 2004/04/15

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

Subject: [Leica] Slides the cat has gone to kitty heaven
From: jean.louchet at inria.fr (Jean Louchet)
Date: Thu Apr 15 16:57:15 2004

>I did not say they were not giving. I implied the $5000 pet operation is
>immoral because it's being spent on a beast rather than a human being. A

It could be spent on buying or repairing a Leica as well.

>beast who exists merely to amuse its owners.

Yes, photography is another expensive amusement.

>> African child, and then rushes in the Volvo to the Vets to have the dog
>> operated on for cancer at a cost of $5,000.

Anything against having a volvo? (designed and built like Leicars, at 
least the old ones, and relatively cheap to run)

>> Vanity pets have no place in a world of suffering human beings.

Suppressing them wouldn't alleviate these sufferings more than suppressing
art photography or flowers in gardens. I have no shame taking a shower in
the morning (sometimes :-) ) even if some people are thirsty in another
continent. I am not sure if taking too many showers is really harmless
though :-)

Jean

p.s. (more on topic) about different recent threads

 -chromed vs. black lenses: leica chrome lenses are sign. heavier than
their black equivalents. My hypothesis is that the black ones are anodised
aluminium alloy, the chromed ones are chromed copper/zinc alloy.
Copper/zinc alloys have a very much lower coefficient of dilatation than
aluminium alloys (better in thermal gradients). I like chromed lenses in
sunny environments as they do not heat up as much as the black ones.

 -cheap 135mm: I have got a Jupiter 4/135 (LTM). old bright alum version,
and I am really very glad with it. Lightweight and surprisingly good
results and sharpness. In mountain landscapes it gives pictures very
comparable to the Elmar-C 4/90 except that is it a bit more
flare-sensitive, but same contrast. Not yet tested on shorter distances.
I can't say as many good things on my other Russian lenses. My Russar 20mm 
is a dog, nice on the shelf (black vodka-break version). The ultra-compact 
6/28mm (chromed) is good, nothing spectacular but a good lightweight thing 
with the CLE. The collapsible industar 3.5/50 is good around f/5.6 or 8 
but a bit prone to flare. The Zorki 1.5/50 (variant of Jupiter) is 
acceptable and I use it in 
"risky" events with one of the old zorkis, but there is a world between it 
and the bright and saturated CV 1.5/50 Nokton in terms of image quality 
and bulk. The Jupiter 2/85 is just a disaster: good/very good glass but 
wrong rangefinder cam thread, so it is unadjustable...

 -Just bought a Jupiter-6 (2.8/180), adapted to the F3 with a little work
(internal adjustment for inf. focusing + machining outside where it would
rub the big F3 prism), no results for the moment but looks very good, very
little flare due to its huge body and virtualy no vignetting thanks to its
huge front elements. Should be adaptable to any 24x36 slr as the Ni
bayonet is one of the furthest from the film, and there is plenty of space
for adjustment inside the lens. Very stable with its 1.5 kg... and cheap
(100 euros on ebay)

 -about the Gaussian filter: I am not a Photoshop user but did some
digital image processing (theory...). A true Gaussian filter will
progressively kill all the high frequencies in the image (the higher the
freq., the more it will be reduced). On a computer it will be approximated
by a "truncated" Gaussian filter, with reasonably similar effects. I think
that a "portrait lens" is in theory exactly equivalent to a "perfectly
sharp" lens plus a truncated Gaussian filter.

Reducing the resolution to 50% then restoring it to 100% will at its best
give the same result, but this depends on the interpolation algorithm
inthe software, and I would bet that for computation cost reasons this
will be a rather crude approximation.

A uniform blurring filter (also called "local averaging filter) is quite
different from a Gaussian filter. All photographers know that an "out of
focus" picture is not the same as a smooth picture. It can be simulated
with a perfect picture processed with a "uniform blurring" fliter. [For
math addicts, the ever-oscillating shape of the corresponding Fourier
transform shows that] an out-of-focus picture still contains most of the
original high frequency information (which a "smooth" picture does not
contain). This is why out-of-focus blur can be compensated (not completely
but significantly) by computer calculations (I suppose there must be such
a thing in photoshop?).

Motion blur is far more difficult to correct. One funny thing is
out-of-focus blur with catadioptric lenses, this can be corrected too but
the formula is more complicated. Gaussian blur cannot be corrected. The
human eye tends to consider a Gaussian blurred image as "sharp".

Scratches on the lens won't simulate Gaussian blur, but intriduce lots of 
flare and fog on the image. May work to hide unwanted moustaches but 
careless work on the enlarger will do about the same... without killing a 
lens.

Jean 

  -- 
 ------------------------------------------------------------
 Dr Jean Louchet       Equipe COMPLEX      INRIA Rocquencourt
                       BP105   78153 Le Chesnay cedex, France
 Jean.Louchet@inria.fr     http://fractales.inria.fr/~louchet
 +33 (0)1 3963 5582/5104              fax: +33 (0)1 3963 5995
 ------------------------------------------------------------ 


Replies: Reply from crgrbrts at verizon.net (Craig Roberts) ([Leica] Slides the cat has gone to kitty heaven)