[Leica] Rangefinder accuracy

[Erwin Puts <imxputs@knoware.nl>]

A number of years ago, when comparing the accuracy of the M6 against the G1
I noted that the accuracy of the M6 was within a 3% range. The figures of
the G1I will mercifully keep a secret.
For my analysis of the M3, M6 and M6 HM I used this figure as a basis and
conducted a series of tests using the Noctilux at apertures 1 to 2 and the
Summilux 1,4 at apertures 1,4 to 2,8. The distance I set at 2,5 meters and
I placed objects at 10cm distance from each other, just within the DoF of
the Noct at 1 and the Summilux at 1,4. I also conducted tests with a 90mm
at about 100meters, just short of infinity.All situations had strong
backlighting in order to make visibility of the object a little strained.
As the mechanical linkage for all three cameras is identical, the only
difference in accuracy is caused by the viewfinder enlargement, givng a
slightly larger displacement of the rangefinder spot,making it visually
easier to see the separation between the in and outoffocus subjectlines.
Full results shortly at my site (where you can find the expanded results of
the 21mm test), but here the condensed conclusions.
With the Noctilux at f/1 and f/1.4 and the Summilux at f/1.4 and /2,0 the
M3 and the M6HM outperformed the M6 significantly. At f2 and f/2,8 the
advantages  shrinked to  smaller margins, as the extended DoF now masked
the fine differences in focus error.
The M3 still has a slight edge (what did these designers in the 50's had in
mind: a 0,7 lens?), but is more than offset by the easier to view
rnagefinder and the greater clarity. The prominent 50mm lines in the M3
viewfinder are more distracting than the smaller lines of the HM. The HM
has two sets of framelines of course, but not so forcefully in the image.
The upshot: when demanding highest accuracy and working in difficult
situations with wide apertures, there is only one new choice: the HM. (it
also proves the validity of my earlier postings on weird mathematics). I
also took care to note if the blanking out of the rangefinder spot (so
lamented with the M6) still occurs. The HM is certainly an improvement, but
perfect, not yet.
Take care to ask yourself the required amount of enlargement from the
negative. All DoF tables are based on  the not very realistic 5x
enlargement factor.
If you are a little more lenient in your demands and/or use mostly
apertures from 2,8 and/or wide angle lenses the choice between M6 and HM
can be made purely on ergonomic or convenience arguments.
Erwin