M3 rangefinder v.offset/focus

[gosfield@DOLPHIN.UPENN.EDU]

hello all

I have an M3 which has served me well for 10 years of so.  Recently I 
have noticed a slight vertical offset in the rangefinder. I.E.: when 
converging the images on point light sources, there is a very slight 
discrepancy in the vertical axis between the two rangefinder patch images. 
Photos taken at large depth of field/infinity have continued to be fine.

However...i recently took some photos of a pet cat sitting on a fence at
about 4-5 feet, in poor light, at very low shutter speed and medium
aperature, on tripod.  The entire cat was out of focus, (so was the fence)
and i can not find any 'micro-streaks' under loup mag. to prove camera
movement, nor any consistent smears to show cat movement. I am most
inclined to believe this is due to my own error in focus or stability, 
possibly due to the low light (dusk).

I have read (here or on rec.photo.advanced) that there are actually two 
separate focus adjustments on the M3, one for far focus, and one for near 
focus, and that either one may be out of adjustment separately.  I have 
also read of service problems in attempting to correct these 
adjustments, even from Leica service (Remember the "RF innacuracy in M 
Leica" thread on rec.photo.advanced ?)

What are the easiest ways for the 'civilian' to check these.  Obviously i
can measure some distances, focus and shoot, rack focus forward and
backward by small increments and 'bracket focus' recording the actual
distance of the focus points, and check in resulting proofs for any
consistent focus shift. 

Is there any visual indication in viewfinder of focus problems in the
close up range?  My slight vertical offeset has certainly existed during
the period of acceptable service (sharp 8x10 with Tech Pan at
infinity/near infinity focus).  When i took the camera to a local
repairperson they told me there was no RF problem, just from visual
inspection (probably considered me another obsessive-compulsive
Leicaphile.) Nevertheless, the point sources are still detectably double
when in focus.  Is there an 'acceptable range' of RF image convergence or
should it be exact. 

Advice from the technically experienced/sophisticated would be
appreciated.  I do a lot of macro, and have done SEM/TEM, (although i don't
mess around inside cameras) so be as technical as you like (i'll ask
questions or look up what i don't understand). 

best wishes to all,

ted
gosfield@dolphin.upenn.edu