Archived posting to the Leica Users Group, 1999/04/27

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

Subject: [Leica] Leica revisited & production tolerances
From: "Folmer, Derek" <derek.folmer@rtit.com>
Date: Tue, 27 Apr 1999 14:50:57 -0400

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

- ------_=_NextPart_001_01BE90DE.E8AEE140
Content-Type: text/plain

If I may comment and expand on something on what was mentioned in an earlier
post. I really don't want to drag this on & would rather learn more about
relevant Leica subject matter than discuss statistics or quality control.
Now I'm not a mathematician, so I might miss some of the subtleties and am
trying to keep this general.

Erwin wrote:
Any component and any manual adjustment in the Leica production environment
is 
governed by tolerances that are set to very small values. But every value is

also ruled by the statistical fact that 95% of instances will be within the 
tolerance bandwidth  and 5% will be beyond this value span. Sometimes these 
tolerance deviations cancel out, sometimes they do not.

The engineers at Leica would select their "confidence interval" that they
want their range of specifications to fall within. If this is 95% or not, I
am not sure. 95% is a common acceptable range, but it could be tightened to
99% or even more in areas with critical consequences.

Generally one standard deviation has a confidence interval of 68%. That is,
if you take a set of samples, determine the average and standard deviation -
68% of your products will fall within a range of plus/minus one standard
deviation. 

Likewise 95% confidence corresponds to roughly 2 standard deviations. 

What does that mean for pressure plates? Assume that Leica designs a gap of
1 mm for the pressure plates. The do their quality control tests, determine
that the average gap is 1mm, with a standard deviation of, lets say 0.02mm
(Give me a break, I'm just making these numbers up!). So roughly 95% of
products will fall within the range 1 mm +/- 0.04 (which is 0.96mm to
1.04mm). So (exam question), will your film get scratched because of the
pressure plate?

So to get back to Erwin's post - yes, I am quite sure that Leica
manufactures to tight tolerances and what that means is that the standard
deviation will be quite low & that statistically there is little chance that
the product will be out of spec. However, there is no "statistical  fact"
that 95% will be within this bandwidth and, though there is a tendency for a
'regression to the mean' I would not be willing to live with 'two wrongs
make a right' or that the tolerances will somehow work out sometimes.

I have not completely followed this thread, but if there are enough cases of
pressure plate or other problems in new cameras then Leica does have a
problem. Either way, if you have a new camera and there is something
justifiably wrong with it, even if you are the only one, Leica must feel the
obligation to solve the problem for you.

That said, the M3 and lenses I use date from the mid '50's. In the future I
know I will have to put money into them to keeep them 'as good as new' but
in the mean time I am willing to live with their faults and am quite happy
when I'm out taking photographs and seeing what is around me with a
photographers eye as well as trying to interpret what I see as creatively
and interestingly as possible. If I had a new camera it would be a different
story, and I would expect everything to work.

Derek




- ------_=_NextPart_001_01BE90DE.E8AEE140
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Dus-ascii">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.0.1461.43">
<TITLE>Leica revisited &amp; production tolerances</TITLE>
</HEAD>
<BODY>

<P><FONT FACE=3D"Courier">If I may comment and expand on something on =
what was mentioned in an earlier post. I really don't want to drag this =
on &amp; would rather learn more about relevant Leica subject matter =
than discuss statistics or quality control. Now I'm not a =
mathematician, so I might miss some of the subtleties and am trying to =
keep this general.</FONT></P>

<P><FONT FACE=3D"Courier">Erwin wrote:</FONT>
<BR><FONT FACE=3D"Courier">Any component and any manual adjustment in =
the Leica production environment is </FONT>
<BR><FONT FACE=3D"Courier">governed by tolerances that are set to very =
small values. But every value is </FONT>
<BR><FONT FACE=3D"Courier">also ruled by the statistical fact that 95% =
of instances will be within the </FONT>
<BR><FONT FACE=3D"Courier">tolerance bandwidth&nbsp; and 5% will be =
beyond this value span. Sometimes these </FONT>
<BR><FONT FACE=3D"Courier">tolerance deviations cancel out, sometimes =
they do not.</FONT>
</P>

<P><FONT FACE=3D"Courier">The engineers at Leica would select their =
&quot;confidence interval&quot; that they want their range of =
specifications to fall within. If this is 95% or not, I am not sure. =
95% is a common acceptable range, but it could be tightened to 99% or =
even more in areas with critical consequences.</FONT></P>

<P><FONT FACE=3D"Courier">Generally one standard deviation has a =
confidence interval of 68%. That is, if you take a set of samples, =
determine the average and standard deviation - 68% of your products =
will fall within a range of plus/minus one standard deviation. =
</FONT></P>

<P><FONT FACE=3D"Courier">Likewise 95% confidence corresponds to =
roughly 2 standard deviations. </FONT>
</P>

<P><FONT FACE=3D"Courier">What does that mean for pressure plates? =
Assume that Leica designs a gap of 1 mm for the pressure plates. The do =
their quality control tests, determine that the average gap is 1mm, =
with a standard deviation of, lets say 0.02mm (Give me a break, I'm =
just making these numbers up!). So roughly 95% of products will fall =
within the range 1 mm +/- 0.04 (which is 0.96mm to 1.04mm). So (exam =
question), will your film get scratched because of the pressure =
plate?</FONT></P>

<P><FONT FACE=3D"Courier">So to get back to Erwin's post - yes, I am =
quite sure that Leica manufactures to tight tolerances and what that =
means is that the standard deviation will be quite low &amp; that =
statistically there is little chance that the product will be out of =
spec. However, there is no &quot;statistical&nbsp; fact&quot; that 95% =
will be within this bandwidth and, though there is a tendency for a =
'regression to the mean' I would not be willing to live with 'two =
wrongs make a right' or that the tolerances will somehow work out =
sometimes.</FONT></P>

<P><FONT FACE=3D"Courier">I have not completely followed this thread, =
but if there are enough cases of pressure plate or other problems in =
new cameras then Leica does have a problem. Either way, if you have a =
new camera and there is something justifiably wrong with it, even if =
you are the only one, Leica must feel the obligation to solve the =
problem for you.</FONT></P>

<P><FONT FACE=3D"Courier">That said, the M3 and lenses I use date from =
the mid '50's. In the future I know I will have to put money into them =
to keeep them 'as good as new' but in the mean time I am willing to =
live with their faults and am quite happy when I'm out taking =
photographs and seeing what is around me with a photographers eye as =
well as trying to interpret what I see as creatively and interestingly =
as possible. If I had a new camera it would be a different story, and I =
would expect everything to work.</FONT></P>

<P><FONT FACE=3D"Courier">Derek</FONT>
</P>
<BR>
<BR>

</BODY>
</HTML>
- ------_=_NextPart_001_01BE90DE.E8AEE140--