The guide rails are machined and each shall be very "linear". As the lens
mount is lay flat on the table, measure four corners are good enough.
My point is about the different measurement data you got, it is a disaster
if you get 0.19mm different from one side to the other, it is almost
impossible. You will get a very bad focal plan and image will never sharp on
the whole frame unless heavy stop down. This is not related to focusing
For the film flatness issue in 35mm system, it is over emphasized, no camera
manufacturer care about it except Contax who made RTS III with vacumm
BTW, my table is just not so flat for precision meassurement and it is still
ok. To claim good measurement data we usually use a professional table made
----- Original Message -----
From: "Moose" <olymoose@xxxxxxxxx>
> C.H.Ling wrote:
>> Sorry, Moose. 0.19mm (46.08-45.89) difference seems a lot!
> Yeah, I agree, and I may have exaggerated it by using the largest
> possible measuring error
>> I remember Canyon New F1 claimed they controlled the tolerance to 0.02mm.
>> I just made some measurements on an OM1n and OM2n, the four corners of
>> the guide rails are no more than 0.035mm difference, this already
>> included the errors from my not so flat table (but I did measured more
>> than two times at different locations of my table).
> My table's pretty flat. And maybe I grabbed a worst case body at random.
> I didn't just take corner reading, but also middle and in between. In
> the case of a casting, assuming that dimensional distortion will be
> linear, rather than radial, isn't necessarily a good idea. I don't
> recall which were which, as my point was simply that obsessing about
> whether the nominal register distance is perfect is wasted effort.
> As long as you bring it up again, I'll add viewfinder optics, mirror
> position accuracy and precision of focusing screen positioning to film
> characteristics and film flatness as factors that contribute to focusing
> accuracy beyond register distance precision.
> If one has a problem with focus accuracy, it's worth tracking down which
> of multiple factors contribute to it. Other wise, take pictures, be happy.
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx