R8 metering (technique)

[Erwin Puts <imxputs@knoware.nl>]

Well Elliot,
here some background info on metering calibartion. All handheld meters, 
incident, reflected and spot (also reflected but narrow angle between 1 
and 3 degrees) are calibrated to simulate the so called 18% reflectance 
grey card.In-camera meters follow the 'whims' of the manufacturer (see 
below), except in the case of the in-built spotmetering which is again 
according to 18% reflectance. 
Now the assumtion behind the 18% is the idea that the average 
distribution of light and dark tones in a scene is equal to 18% 
reflectance of the incident light. One stop more (overexposure) gives 36% 
reflectance, two stops more: 72% and 100% is reched at about 2.5 stops. 
On the underexposure we have minus one stop: 9%; minus two stops: 4,5%; 
minus three stops: 2,25%. So we have an overall contrast from deep shadow 
to highlight of 5,5 stops. Therefore the standard 18% calibration assumes 
two parameters: a total contrast of 5 stops and a particular distribution 
of light and dark tones. This assumption is not always met in practise, 
but the normal integrated metering (as in the R8) uses it as a basis for 
calculations. So if the assumption is realised in practise then the 
integral metering does a fine job. 
In reality the distribution of light and dark scenes follows the standard 
statistical Bell curve. If you adopt this in the exposure calculations, 
then the average scene has a reflectance value of 13%, which is more 
realistic. So many 'average' scenes are slightly underexposed, giving the 
rich saturated colours (more saturation equals less mixture of white 
light). 
The matrix metering is calibrated according to the kind of statistical 
analysis most camera manufacturers do: analyse typical scenes, they 
assume the typical clientele will take and adjust the algorithms 
accordingly. Take pictures out of the bandwidth of this adjustments and 
you go wrong. 
The spot metering is accurate as long as you know what you are doing. 
Avery area you are aiming at is interpreted as an 18% reflectance 
whatever the actual luminance value. This you can use as follows.Measure 
the darkest part of your object you want with some detail in it, then 
adjust your reading minus two and a half stop: then all your shadow areas 
will show beautifull details. Measure a highlight area and over expose 
two stops. 
If in doubt use these two time honoured rules; 
The reflectance of fresh green grass is 18%, so metering on this area is 
identical to metering on a grey card.
The sunny f/16 rule. Midday sunny light should give an exposure 
equivalent to f/16 and the reciproke of the ISO value of your film. That 
is a very easy way to check the accuracy of your exposure system.
Erwin