Mathematically, the RMS value of a series of points (or voltages, in this case) is calculated by squaring each point, taking the average of the squares, and then dividing this by the number of voltage points measured. You then take the square root of this number to get the RMS value. (It's literally the __root__ of the __mean__ of the __squares__.) For a simple sinusoid, this works out to be 0.707 times its peak amplitude. It's a type of averaging, though not the same as taking the arithmetric mean of a series of values.

Meters with RMS ballistics are meant to convey some information about the perceived loudness of a signal. They do this better than a peak meter would, as our hearing apparatus perceives loudness not by peak level, but by something closer to RMS level.

The RMS of a signal doesn't tell the whole story of its perceived loudness, since one can have two signals with the same RMS and make one sound louder than the other by filtering out its very low frequencies that we do not efficiently hear. It's even possible for a somewhat lower RMS signal to sound louder, depending on its frequency content/balance. This is why simply looking at an RMS meter to decide if your master is competitively loud is insufficient on its own.

-Ben B

*
Last edited by Ben B; 15th September 2011 at 08:52 PM..
Reason: incomplete
*