Archived posting to the Leica Users Group, 2008/04/20

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

Subject: [Leica] Noctilux DOF Film vs M8
From: pklein at 2alpha.net (Peter Klein)
Date: Sun Apr 20 21:05:16 2008

Further musings on M8 focusing:

The tolerances needed for everything to be exactly right on an M8 are 
tighter than with a film M.  Film has depth.  As long as the rays come to a 
focus somewhere within the emulsion, they will excite the most silver 
halide molecules at this point, and you have an in-focus image.  Not so 
with the M8. Its sensor is a 2-dimensional plane as far as the image is 
concerned. You're either in focus or you're not.

Now, suppose a lens focus shifts. Then the question becomes, "what exactly 
is correct focus, anyway?"  At f/1.4, f/2 and f/8+? Or at the stops 
between? Or some compromise that makes it within acceptable limits most of 
the time?

Because on the M8, you might have to choose.  And to face a question most 
Leicaphiles don't want to ask:  "How much error is acceptable?"  We want 
perfection.  With some lenses, we ain't gonna get it.

My guess is that with film, Leica set up a standard that put most lenses in 
focus with most films, regardless of aperture. As a practical matter, focus 
shift was mostly taken care of in the tolerance given by the thickness of 
the emulsion. Maybe with film, we had a fudge factor of 3 gnat's eyebrows. 
Now it's got to be within half a gnat's eyebrow.

Leica has provided some information about all this. See the LFI article 
that showed photos of wine bottles taken with the 35/1.4 ASPH at various 
apertures.

As I've said before, on the M8 I can detect a bit of backwards focus shift 
on *all* my lenses of f/2 and faster.  I've tested with tape measures at 1 
and 2 meters, and cereal boxes on the roof of my car at 2-6 meters, and I 
always get the same basic result.  I too have noticed that with the M8, at 
middle apertures especially, it's best to assume that most to all of the 
DOF is behind the point of focus, not distributed in the classic 1/3 front, 
2/3 back configuration we knew with film.  This may be why some people say 
that "all lenses back focus on the M8."

Focus on the closest thing you absolutely must have in focus, and you'll be 
OK most of the time.

The good news is that with many lenses, the error may not be 
photographically significant at sizes that most of us print at.  I'd advise 
people to stop pixel peeping at 100% and assuming something is wrong. If 
the screen looks good at 50%, prints up to 11x14 probably will, too.

The Noct is a special case.  Its focus shift was already significant with 
film.  It is even more significant with the M8. Since Noct owners bought 
the lens for f/1, I'd suggest using it at f/1, and if it works there, chalk 
the middle stops up to the universe and physics (or learn to the 
appropriate amount in front of what you really want).  But if adjusting the 
camera to your Noct at f/1 puts your other lenses' focusing off, I'd 
suggest putting the RF back the way it was, and getting the Noct adjusted.

The big unknown in all this:  Has Leica has upgraded their equipment enough 
to adjust fast lenses to best advantage on the M8?  From some leaks and 
rumors I've read, maybe they have in Solms.  Have they done so in NJ?  I 
don't know.  There have been an awful lot of stories about people sending 
their lenses in, but they come back focusing no better, or worse. There 
have been too many of these stories from too many good photographers for me 
to attribute it all to photographer error.  This is why I trust DAG more 
than Leica in terms of lens adjustment these days.

--Peter


Replies: Reply from kididdoc at cox.net (Steve Barbour) ([Leica] Noctilux DOF Film vs M8)