Hi Speedskater,
It's more complicated terminology than I thought.
I just had J. Gordon Holt's 1990 book "The Audio Glossary" out and here are a few definitions:
phase
(5) An expression of the relative polarities of a stereo pair of devices or audio signals. Out-of-phase means connected in opposite polarity (180° out of phase).
polarity inversion
Reversal of absolute phase, so as to reproduce original compression waves as rarefactions and vice versa.
polarity switch
A user control which reverses the absolute phase of a device's output signals.
and then there is:
absolute phase
(1) The correct acoustical polarity of sound waves, in which original air compressions are reproduced as compressions rather than as rarefactions.
(2) The electrical polarity of both stereo channels, which determines (1).
Much as I admire Gordon's work, I think it can be stated more succinctly:
Phase is not polarity and polarity is not phase. But there is a relationship.
Phase involves time and polarity involves direction.
Phase deals with relative timing between two signals. Though the phrase "out of phase" is often used to indicate two signals of opposite *polarity*, it actually means they are not in time (or "in phase") with each other. In fact, the polarity of both can be the same but one might be a fraction of a wavelength behind or ahead of the other. The audible results can manifest themselves in a number of ways, from an odd frequency response to one musical "attack" being heard more than once. (Anyone old enough to remember television via antenna might also remember "ghosts", pale "echoes" of the main signal, seen slightly offset to one side from the main image. This is another manifestation, this time with video of course, of phase issues.)
Polarity, as I mentioned above, deals with the direction of the signal, its positive motion and its negative motion. Sound waves in nature, including from musical instruments and voices, are *not* symmetrical. As with everything in audio, different folks will have different sensitivities to this. Some will not hear a polarity reversal if both channels in a stereo signal are of the same polarity. Others will hear differences they'll describe as "subtle" and others will insist on being able to correct this on recordings that have it wrong. (This is complicated by typical multimic, multitrack recordings were different signals are in different polarities.)
A somewhat recent example of a recording that is in phase (with itself) but in reversed polarity from the original release is the Beatles' 2009 CD remaster of "Revolver". Compared with the original 1987 CD, one of the differences is the polarity (which I only tested -and confirmed- because I heard it when doing a listening comparison of the two). I'm told the 2009 matches the master tape. If that is the case, my ears make me suspect the error is then in that master tape. To my ears, the '87 CD, its limitations aside, sounds more "in focus".
In sum, the correct term *is* "absolute polarity" (as opposed to "relative polarity", a condition which, as I mentioned above, is often, mistakenly as I see it, referred to with the term "out of phase").
Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com