You're not slow at all - it's one of those subjects where the phsyics is simple but the implications are not.
When you move, the difference you hear is actually what's called an interference pattern.
Think back to science class - when you dropped two things into a pool of water - and the waves spreading from the two things met and created a pattern of dark and bright spots.
That's precisely what happens when the sound coming from two speakers meets in the air in your room.
(And, if you consider a two-dimensional version, and move your point of view along a line, it gets brighter and darker - which is what they call "a comb filter effect".)
And, when you move your head, you're moving your ear between "a bright spot" and "a dark spot" - so the sound gets louder and quieter.
(And remember that, while we're talking about a single sharp "pattern", the pattern is actually different at each and every frequency.)
"Absolute phase" is always sort of relative... because it's related to time.
Lets start with one speaker, playing one tone, continuously.
Now, let's put a microphone at one spot, and take an "instant snapshot" of the sound there... and let's just "note" what it is.
Now, at the exact same time, let's put another microphone 12" further away from that speaker.
The sound that we measure at our second microphone will have been delayed - compared to the sound at the first microphone - because it took it longer to travel those extra 12" (it will be delayed by about 1 millisecond).
Therefore, assuming our original source was a sine wave, at each of the two microphone positions (at the same exact time), we'll be looking at different parts of that sine wave.
We can look at that difference as a difference in milliseconds; but sometimes it's more convenient to look at it as a fraction of the wavelength of the frequency we're thinking about.
The reason TO look at it in terms of phase rather than distance is that whether the sound adds or cancels out at a particular spot depends on whether it's in phase or out of phase at that point at our particular frequency.
(So, if the speakers are 1 foot separated in distance, we might get a cancellation or an addition, depending on the frequency;
but, if we talk about phase, sounds that are in phase will always add, and sound that are 180 degrees out of phase will always cancel.
However, the time difference it takes to "make the sound 180 degrees out of phase" will be different for each frequency we may consider.
And, if we shift one speaker one foot, whether the result is in or out of phase, and how much, will be different at different frequencies.)
However, your confusion at this point is actually quite reasonable.
You are quite right that, whenever we have the same sound coming from two speakers, we end up with a complex pattern of additions and cancellations.
(Remember those waves from science class.)
At any given frequency, the room will contain a complex 3d pattern of points where they add together and where they cancel each other.
The distance between those points, and the amplitude variation between them, will depend on things like the distance between the speakers and the room acoustics.
It also depends on the frequency, so the pattern will in fact be different for each and every frequency.
Luckily for us, in most rooms, the patterns aren't especially sharp or distinct, and our brains are
VERY good at sorting this type of stuff out.
In fact, our brains use this sort of information to figure out where things are (or appear to be).
The reason we worry about things like making sure we're sitting right between the speakers is that our brains are especially sensitive to variations in symmetry.
If both speakers are wrong in exactly the same way, we still perceive the sound as coming from the center.
(You can't hear if both speakers are six feet away from you or seven feet away, but you will hear if they're
different distances away from you.)
The reason is that, if they're different, either because they're different distances from us, or because something else is different between them, then we get "conflicting cues".
(For example, if the sound is the same loudness in both ears, then that's a cue that it's in the center; but, if it arrives at our right ear 1 millisecond sooner than the left,
that's a cue that it's a foot closer to the right; if those two cues "conflict", then our brain has trouble figuring out where the source of the sound "really" is.
While some cues are more important than others, and we can live without some of them, having all the cues we do have agree, and avoiding ones that conflict, always makes things sound "more solid" or "more real".
(Cues include relative amplitude, angle of arrival {look up HRTF}, time difference on first arrival, phase difference between channels, arrival times of reflections from walls and floor, and several others.)
(If our brain decides that the main frequency of the piano key is in the center, but the first harmonic is a foot to the left, it doesn't sound "right".)
The result there is what we perceive as "bad imaging" or "a blurred sound stage" or "it doesn't sound like a real piano".)
Note that the whole mess is even more complicated because what we hear is really a mix of direct sounds, and sounds reflecting off the walls, and reflections
OF reflections.
If you use omnidirectional speakers, you aren't changing the initial arrival times (which are based strictly on distance from the speaker).
However, when you use omnis, you are increasing the proportion of reflected sound vs direct sound (they bounce more sound off the walls in all directions, so you hear proportionally more reflected sound).
Since reflected sound in general contributes more to "a sense of air" and less to specific object placement, onmis give you a less distinct image placement to begin with, but one that is also less sensitive to minor imperfections.
Unfortunately, with real speakers, and real rooms, the details get VERY complicated....
However, the bottom line is that, as long as the situation is symmetrical between left and right, then our brains usually manage to keep things straight in terms of sound stage and imaging.
But, as soon as that symmetry gets "broken", like by having a phase shift that ends up being different between left and right at some frequency, or more reflections from the left wall than the right at some particular range of frequencies, then the imaging gets "funny".
And the reason why Dirac does better at room correction than systems that only correct frequency response is that it
DOES correct these phase differences and errors when it can. That's why it can have such a major effect on imaging in certain circumstances.
(Note that frequency response differences between the channels also have a major effect on imaging - which is why systems that only correct frequency response can also sometimes offer big improvements. Dirac is just correcting
more things that may need correcting.)
The short answer is that, with a single speaker, phase shift would be largely unimportant.
And, when you have two speakers playing the same signal, there are always going to be cancellations and interactions between them at various points in the room.
(To hear this, just play a steady mono tone from both speakers, and move your head slowly a few inches side to side.)
However, as long as this is more or less symmetrical, it isn't usually too annoying.
But it gets bad when it's not symmetrical.
The easiest way to "simulate" some asymmetrical phase shift is simply to sit in your normal stereo listening position, then move one speaker a few feet further away from you (keep it at the same relative angle).
If you have a balance control, then adjust it so they're still at the exact same loudness.
Making the distance between yourself and the two speakers unequal will introduce "a frequency-dependent phase difference" between them.
(The phase shift will be 90 degrees at 100 Hz, and about 900 degrees at 1 kHz, if you move one speaker about 2-1/2 feet further away.)
This example will yield a phase shift that, while it shifts with frequency, will vary
smoothly with frequency.
The effect is somewhat more annoying if something like an equalizer introduces a phase shift that varies non-smoothly.
I'[m trying to digest your points one by one. So forgive me if I'm going slow as molasses. So when I shift in position slightly and perceive a change in sound that could be phase shift? Huh. That explains a lot of my experience. Small shifts away from the sweet spot and I go ...wait what happened here? (small room)
But it's also a bit dismaying that unless one is in the absolute center one cannot experience a stereo image without phase shift. (I guess without omnidirectional MBL speakers I assume). So basically stereo allows for a pretty good reproduction minus height for ONE LISTENER? That explains a lot of what has me confused @keithl unless I am misintepreting this.
