[davew]

On Nov 23, 2:00*pm, (Steve Pope) wrote:
Chris Bore wrote:

Having written at length, I reluctantly consulted Wikipedia and to my
amazement found it contained useful information.

http://en.wikipedia.org/wiki/DBFS

It does refer to the ambiguity, and in fact refers to an AES standard
(0 dBFS is rms of a full scale sine wave) and to a practial standard
(Euphonix sound level meters) that has 0 dBFS as the rms of a full
scale square wave (equivalent to the 'instantaneous' definition that I
suggested).

I find the former of these two definitions more widely used than the
latter. * Any signal whose RMS value is the same as that of a full-scale
non-clipping sine wave is 0 dBFS. *So a clipping square wave is
+3 dBFS.

Steve
And a digital implementation of a VU meter reads +4dBFS for a full
scale square wave or 0dBFS depending on whether the ratio of peak to
mean absolute value is taken into account. We add 4dB to make a sine
wave read the same value as the peak sample value.