Archived posting to the Leica Users Group, 2003/07/16

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

Subject: Re: [Leica] digital glass
From: Peter Klein <pklein@2alpha.net>
Date: Wed, 16 Jul 2003 14:32:18 -0700 (PDT)

Phong:  It's interesting how well your D30 does despite being "only" 3
megapixels.  What print size do you consider the upper limit for a
full-frame D30 picture? 

Tom:  Thanks for posting the results of your tests.  That is really
valuable information.

One of the main advantages of Leica lenses is how they perform
wide-open or nearly so.  And I suspect (without being able to prove it)
that some of the differences show up better in black and white.

It does seem that a lot of people will tolerate a lack of resolution more
than they will tolerate grain.  Weird.  It's well known that a little
grain adds to the impression of sharpness.  Just as a little (audio) noise
can make a soft voice easier to hear.

Perhaps this is cultural and generational. People used to the Web tolerate
color juxtapositions that were considered anathama up until recently.  
And Web photos are often jaggy and blurry. The name of the game for a lot
of Web images is removing enough detail so that the the thing downloads at
a tolerable speed.  So maybe people have "baselined" that look. 

I would rather have grain and detail than neither.  I have spent
considerable time "de-noising" images with Neat Image, applying it very
sparingly (half the recommended amount of cleanup or less)--only to find
that I prefer a little grain to that "plastic" look.  Perhaps that's
because I spent a good portion of my teen years looking at black-and-white
photography from the 1930s onward. Smooth is good.  But plastic is bad.  
And I must say that I like a smooth, random grain better than digital
noise where a repeating pattern is discernable.

At any rate, it looks like we're dealing with some complex perceptual
things. If 6 megapixels is indeed enough to satisfy a viewer of a 12x18
print, then a 6 mp DSLR is all you need most of the time--for acceptable
commercial work.  As to what truly is "better," you would have to do a lot
more testing, taking into account that people's initial reaction may be
good enough for marketing, but not necessarily for art.  :-)

- --Peter Klein
Seattle, Wa


Phong wrote:

> Thank you for a most interesting report, though being
> a Leica and Canon shooter, the specifics of the new Nikon
> lens are of little more than an academic interest to me.
> Your conclusion regarding negative film, transparency,
> and digital parallel my own exactly.

Tom Lianza wrote:

>> I ran an interesting series of tests that utilized 5 camera/film
combinations (3 leica bodies (2M, 1 R) 2 Nikon bodies, 5 different types
of film) compared to the digital output from the D100.  I then took six
different scenes.  The film images were scanned in an LS4000 scanner and
then all the digital files where printed at 12X18 inches using a fuji
frontier printer. I then asked observers to rank-weight the images.

>> In 5 of the six tests, the digital camera (6 Mega pixel) was judged
"better".  In one instance, the observers generally chose an image shot
with an R6/Summilux 80 combination on a fine grain transparency film.  
One of the first observations was that whatever image quality advantage
the leicas had on the film, was lost in the scanning-printing process. It
appeared that grain noise was more objectional to the observers than
resolution issues and the sharpening utilized in the digital camera images
had very little impact on the apparent noise in the image.  Inspite of the
low resolution of the digital camera, the fact that it required only one
optical step allowed for less sharpening and lower noise recording than
the film combinations. I also suspect that the fuji frontier imaging
system also contributed to the image degredation.

>> So what I am seeing in my work and my tests is that the output path
(scanning-paper) is a greater limiting factor than digital resolution vs
analog film resolution.  Examination of the film images under a loop and
higher magnifications showed that the Leica images may have a bit of an
edge, but there was no obvious difference.  That accounted for the obvious
confusion in the evaluations of the reproductions after the digital image
was removed from the test: no single camera/film combination stuck out as
a winner.

>>At this point in time, I feel pretty comfortable with either digital or
analog workflows.  If I'm in a situation that requires a wide dynamic
range capture, I use negative film.  If I'm shooting for saturation and
sharpness, I'll use transparency film .  If I need a decent image with
almost no hassles or in a short time period, I'll shoot with digital and I
know that at modest reproduction sizes, it will do just fine. Now I can
stop screwing around with all this testing and just go out and shoot....


- --
To unsubscribe, see http://mejac.palo-alto.ca.us/leica-users/unsub.html