My earlier comments applied to the great outdoors.
On the other hand, if using incident light metering indoors (e.g., light source is a window, lamp or flashgun) then distance from light source does matter and the inverse square law (subject to light source, but not subject to camera) does apply and readings should be taken as close as possible to the subject, meter pointing towards the camera.
Folks, incident light metering (like the tides) does work in a predictable fashion. If you want to explain just how it is that it works, there may be a PhD. thesis in it for you.
Incidentally, (pun intended) for those who agonize about metering a 3-dimensional subject half in the light and half in its own shade (say, a subject's face), that is exactly why the meter has a dome (some domes being more sophisticated than others) which, if the meter is pointed correctly, results in an in-meter averaging of the required exposure with a single reading (the dome being a model for the face).
In fact, this is what makes incident light metering so convenient: this relatively simple device -with a single reading- results in the same exposure outcomes as is generated by the averaging of a number of spot metered readings, shade and light averagings taken by reflected light metering, and any number of "you beaut" "matrix" and "honeycomb" metering contraptions.
Last edited by Galah; 02-09-2011 at 05:11 PM. Click to view previous post history.