Quality Control

[George Huczek <ghuczek@sk.sympatico.ca>]

At 02:58 AM 30/08/97 -0400, you wrote:
>In einer eMail vom 30.08.1997  05:41:46, schreiben Sie:
>
>>On the other hand, if what many people fear is 
>>true, that the frame counters of 75% of M6s are breaking in the first 5 
>>years of ownership, then Leica should bite the bullet and recall the 
>>cameras for a fix.  
>
>I counted no more than 9 (in words: nine) different users complaining on
>their broken M6s. Leitz probably made and sold some more ...
>
One can not conclude that the quality of the M6 has deteriorated simply by
a few anecdotal incidents posted on the LUG.  The folks at Solms would be
in a better position to determine the failure rate of their cameras based
on production rates and the number of cameras being sent in for repairs.  I
am quite confident that if they felt there was a major problem, then they
would do something about it. 

Having said that though, it should be noted that methods used for quality
control have changed since the days when each item coming off the assembly
line was given a thorough check by a human operator before being allowed to
pass inspection.  Now, many companies rely on random sampling.  I think
Leica uses this for subjecting cameras to maximum and minimum temperature
tolerance tests.  Using statistical methods, they check a certain number of
items coming off the assembly line in a random manner.  If those items pass
a thorough inspection, they assume that the entire batch is acceptable.  If
you happen to purchase one of these items that happened to undergo thorough
testing, then chances are you will not experience any problems.  If, on the
other hand, your merchandise proves to be defective in some way, then you
have become the unfortunate victim of the vagaries of statistical analysis.
 Random sampling is faster and less expensive than individual testing.

- -GH