[Leica] Filter transmission measurements

["Erwin Puts" <imxputs@knoware.nl>]

Tina wrote in part:
"In room light, without the UV filter in front of the meter I get an f
number of 2.89.  With the UV
filter, I get an f number of 2.86.  Not much difference, it's true, but
enough to matter when you are taking pictures in the dark!"

Let us first present some general principles. Most consumer measuring
instruments have an accuracy within 5% as does the mechanical tolerance of
the aperture stop. Restricting ourselves to the Minolta exposure meter, we
should allow for a tolerance of 5%.
The calibration of the Minolta meter is for the 2800 Kelvin range.(tungsten
light)
The B&W data sheets show that for the red part of the spectrum (around 600
nanometers) the UV filter has a transmission efficiency of 97 to 98%.
Theortically then the transmission loss of about 2% would be within the 5%
margin of tolerance and thus be not detectible with reliability. Further to
be sure one should do a series of readings to see if there is statistical
validity.
I did some tests myself with a UV filter and noted
In tungsten light and without the UV filter, my Minolta Autometer V gave a
consistent reading of 2.8 + 9/10 with 10 readings.With the filter on the
measuring dome and without my hands around the filter (to avoid shadows from
my hand) I noticed a drop of only 1/10. When I held the filter in my hand
and positioned it in front of the measuring cell, I noticed a wider range of
readings, within 1/3 stop, that changed a bit depending on the distance from
filter to measuring dome and the angle at which the reading was done. Here
we should find at first a very clear and repeatable lab situation.
With a Minolta soptmeter, I did not notice any difference!
I am not inclined, based on this evidence and the theory,  to assume that a
UV filter will drop the transmission by 1/3 of a stop when photographing in
ambient light with a higher than normal proportion of the red spectrum.
Even if this were the case, consider the effect of it on the density of the
negative.
On the assumption that a true 1/3 stop difference in exposure is real, we
note that a full stop difference would change the density of the negative by
a value of Log 0.3. A third stop would be a change in density of 0.1 (log
scale). BUT this assumes a characteristic curve of inclination 1. Generally
we develop to an inclination of 0.6. The density change would become 0.6 x
0.1 = 0.06. Such a density difference cannot be accounted for in normal
consitions for developing and printing.


Erwin