In article <01bc9cf5$f3860b20$034561ce@tima.com>, ronk@kiva.net says...

> One bit of mis-information I keep running across is that because
>of segmented meterings use of computer generated algorithms that this
>type metering is "ROLL-O-the DICE."

Hmmm, well, as the author of the "R-O-T-D" description of matrix metering,
let me defend my description...;-) If you are trying to accurately control
your exposure, based on solid, dependable metering of subject tonal values,
HOW can you do that when the metering system is tossing in variations in
the readings - and variations that are hard to assess, so one cannot
therefore apply corrections to undo the (therefore equivalent to random...)
variations in the meter readings added by the camera algorithms...???
If you want to trust these algorithms while taking essentially careless
meter readings, they may give good results much (but not all!) of the time.
But if you wish to make careful meter readings (to increase your
correct-exposure success rate), allowing the camera to operate in "roll-o'-th'-dice" mode would remove your ability to take accurate meter readings by introducing a greater element of chance, hence the description...
Hope This Helps
(David Ruether - http://www.fcinet.com/ruether )