Re: [Leica] Consistent underexposure problem

[Godfrey DiGiorgi <ramarren@bayarea.net>]

>> First, I'd do a little testing with a gray card and a
> ... Are you saying that after spending $3000 on a camera
>body, I have to check it to see if the factory spent the few minutes 
>necessary to calibrate the meter?

Fine instrumentation always requires testing and calibration, Anthony. 
When I worked at JPL/NASA, it was unthinkable that we would buy a 
half-million dollar scope and not test, calibrate it before use. Why 
should a camera, even a $3000 camera, be any different? 

As I said, I don't expect it to be off. I would do and have done exactly 
the same thing with every camera and meter I have ever bought, regardless 
of price or supposed quality. In all the years and all the cameras, I've 
found three that were malfunctioning as I bought them and either returned 
them or had them repaired before use. 

A defective-from-manufacture unit is a fact of life in manufacturing and 
does not imply that either the manufacturer or the sales people selling 
the item are crooks or want to burden you with bad merchandise. That's 
what warranties are for. The build quality of Leica cameras is very very 
good and it's very rare that there are actually defective examples which 
make it out to the customer, but it can, and does, happen occasionally. 
The same thing happens with Nikons, Ferraris, Mercedes Benzes, Kenmore 
dishwashers, Sony televisions, everything.

>> In the meanwhile, you can adjust the ASA setting to
>> compensate for the error and go take some more pictures
>> to see if that solves your problem.
>
>I considered that, but I felt that changing the ISO rating would be 
>cheating.  I should be able to get things right with the meter set correctly.

Certainly. That's one reason for running a reference test against a 
known-accurate meter, to be certain the meter is properly calibrated. Any 
meter can be off, but it's infrequent that they are. M6s generally have 
very accurate, reliable meters; you're just eliminating the obvious here.

I'm perhaps overly cautious: I like to test that equipment is functioning 
to my expectations before I start working on what I'm done wrong. Part of 
my engineering background. I've spent too much of this email discussing 
the outside chance that the meter in the camera might be off. Don't sweat 
it, I fully expect that you'll find the meter to be right on the money if 
you test it.

>Looks like I'm going to have to buy a meter.  Which brand is a
>good brand? Why an incident meter and not a spot meter?  I
>usually take pictures of things that are relatively far away
>(landscapes, buildings) and an incident meter might not be very
>practical, unless it can be both incident and spot, somehow.

Gossen, Sekonic, Minolta and Calculite, amongst others, manufacture an 
excellent range of meters. Of the digital models, I prefer the Sekonic 
L308B and L328F the most due to their size and ease of use. I like the 
Gossen Luna Pro SBC and Luna Pro F for their excellent analog read out 
ergonomics. Since I use hand held meters primarily for incident metering, 
I chose the L328F due to the swiveling sensor head: it makes it easier to 
position the sensor when metering such that there are no shadows from 
your holding the meter to get in the way of the light. The L328F also 
allows use of a spot meter attachment for reflected light, limited area 
metering.

Why incident vs spot or reflected meter? You're trying to understand why 
you're getting underexposure, which is most likely to be related to your 
choice of metering targets and their relationship to the 18% neutral gray 
reference that the M6 meter is calibrated to. Any reflective meter's 
exposure recommendations can be fooled if it's not pointed at the 
equivalent of that 18% gray target or it's equivalent. An incident light 
meter is harder to fool because it measures the light falling on it and 
gives an exposure reading which is calibrated to reproduce that 18% gray 
target accurately in that amount of incident light. You could not make 
the assumption if your metering test had in its operation the same 
fallibility that the meter in your camera has, so you use an incident 
meter as a reference to learn what 18% reflectance gray should look like 
by taking a reading with it, then looking around with the reflectance 
meter to see what comes up with the same reading.

An incident meter is fine for metering subjects you can reach as well as 
distant subjects. What is important to an incident meter is that it 
meters the light falling upon the subject. If you're subject is a 
building 400 yards away, illuminated by the present sunlight or overcast 
sky, you can take a reading by stepping into a nearby patch illuminated 
by the same light and taking a reading. Spot meters require more thought 
and evaluation: something like the Zone System notions of placing the 18% 
gray and then reading the spread for the darker and highlight values is 
necessary to get the most out of their use. I've usually gotten better 
results, faster, with an incident meter due to the ease of use and 
because I don't like to spend quite that amount of time doing the 
metering and evaluation: it's easier and faster to bracket a couple of 
frames on a tricky exposure than it is to spend a lot of time evaluating 
with the meter.

>The sunny-sixteen rule works when it is sunny out.  In shadows, under an
>overcast, indoors, etc., it is not very useful.  Furthermore, every specific
>scene is different, and I like to _know_ what the correct exposure is, and 
>not guess.

No meter will necessarily give you the "correct" exposure. They all give 
you an exposure recommendation, based upon how they are used. Your Nikon 
F5 meter has a very sophisticated computer which reads multiple parts of 
the scene it's presented with and then compared them, in conjunction with 
the focus distance and the distribution of colors, to a reference library 
of scenes to obtain a similar reference which it then presents to you and 
the camera automation systems. A handheld meter, and the meter in your 
M6, simply reads the light you present it with and recommends an exposure 
based upon it's assumed reference calibration. It doesn't know whether 
that is correct or not, and it's less sophisticated than the Nikon F5 
meter in making that recommendation. You have to evaluate whether that 
recommendation is correct or not, and adjust exposure accordingly. 

The meter is only a tool, not an arbiter of truth. Only you can know 
"correct" exposure.

Godfrey