[edit, huh?]As you know,
polarity is absolute, phase is relative.I meant..polarity is either/or, where phase is a matter of degree.There are two very different types of phase differences:
1) The common type is phase differences due to a time offset between signals. This is the common experience of a sound interfering with a delayed copy of itself and comb filtering. The effect is frequency relative, that is to say the difference in phase between the two signals causes cancellation or reinforcement that varies with frequency. Aligning tracks by sliding them on the timeline adjusts for arrival times for a single sound, which is usually what you want for recording. For non-coincident microphone pairs, sources that are not in located in exactly same place will show various phase differences of this type between channels, which is normal. This is type of phase difference that is adjusted by moving mics, changing delays, or sliding tracks on the DAW timeline.
2) Constant phase displacements across all frequencies. This is very different in that the amount of phase offset between signals stays constant regardless of frequency. This doesn't occur acoustically, only electrically or numerically in the digital realm. It's usually accomplished with all-pass filters. The term 'phase rotation' and the mention of a single angle of phase difference usually means this type since the degree of phase difference is constant with respect to frequency and that difference can be expressed as a single number.
Two common uses of this is stereo width controls and matrix surround encoding for the decoders in home theater receivers. For Matrix surround, material with no phase difference between channels is routed to the center speaker and material 180 degrees out of phase is routed to the surround speakers. Matrix encoding works by manipulating this type of phase difference to move sounds evenly between the front and back. In contrast, non-constant shift phase differences of the first type moves the out-of-phase portions of a sound towards the surround speakers and the in-phase portion of the same sounds towards to center speaker, possibly spreading ambience around to nice effect, but not moving broad spectrum sounds to a precise front-back location.
I use Samplitude and it has a feature where you can switch the panning mode on stereo tracks to a stereo 'width' control which uses this type of phase manipulation. When panned center to 'mono' both channels are in-phase, when panned 'wider' than standard 100%, the out of phase information between the two channels is increased. At the extreme, the signals are 180 degrees out of phase at all frequencies. When I route the stereo output of the DAW into my surround receiver and turn on any of the basic matrix surround processing (Dolby PLII, DTX Neo6, Lexicon Logic-7, Circle Surround, etc. or even wire up multiple speakers to a single stereo amp using the old Halfler sum/difference wiring technique) that 'width' control becomes a front back stereo panorama control. Turn it all the way wide and the sound goes all the way to the back, turn it all the way mono and it goes all the way to the center speaker.
Even if you care less about surround, checking your mix by sending it through a matrix decoder can be helpful as an overall phase relationship check. If too much goes to the surrounds then there is too much out of phase information.
A vectorscope is a common visualization available on most DAWs that displays the phase relationships between two signals graphically as a wiggly Lissajous curve.
Not sure if that’s what you’re looking for, but hope that helps.
[windbag ]