Phase and phase analysis

I'm having trouble getting my head around this correct phase and phase analysis thing. I'm happy to admit the features in AA2.0 have only highlighted that I don't really understand what I thought I understood!
I know the basics of phase and wave polarity (I think!) and how you get situations like phase cancellation and such, but how can certain frequencies in a stereo wave form be more out of phase than others? Like, could a bass track in a multi-track project be "out of phase" and this would show up more readily in the Phase Analysis window with a mixdown? I thought that an entire wave form- even one half of a stereo wave -was either in phase with the other or not.
I'm asking here if anyone can either take the time to explain, or maybe point me towards a website that offers a real nuts-and-bolts, basic (and other metaphors) explanation of phase and phase issues.
I always reckon dumb (?) questions are better than dumb mistakes.

Cheers, Graeme.

"In" and "out" of phase only refer to a relationship wherein the two waveforms are 0° and 180° in respect to each other. Otherwise, there is the rest of the circle to consider... Is this clear? They are just descriptive names assigned to these two conditions; there are no other ones for the rest.

If you have something which causes a time delay to a signal (as many circuit elements do), the phase shift will be different for each frequency component because the time delay will be a different portion of the period and hence a different phase shift expressed in degrees.

This is only the beginning of the explanation, but you may (or may not) be able to take it from here.

Can I try asking another way? Is it theoretically possible to create the "perfect" mixdown which analysed as having every frequency in phase (according to AA2.0)?
Regardless, I'm doing some music today. I'll dabble with the feature in the extreme and try to learn stuff.
Cheers, Graeme.

Can I try asking another way? Is it theoretically possible to create the "perfect" mixdown which analysed as having every frequency in phase (according to AA2.0)?

I think the answer to your question is "no".

I think you need to back up here - rewind a bit and simplify your questions.  Rather than discussing a "mix" - lets talk about two or three tracks of material and how phase issues might or might not arise.

You seem to have a fundemental misunderstanding about possible phase cancelation and it's making you think way to hard about problems you likely aren't having.

Give a simple, specific example of your recording process - 3 tracks max.  Tell us your concerns about phase cancellation - I think it will be a lot easier to explain once we know where you're coming from and have a simple example to discuss.

FWIW, I've never managed to get much useful information from the phase analysis window.  I usually have it open because it looks cool, but  my ears work better for me than a visual tool.

Thanks Karl, but...

"FWIW, I've never managed to get much useful information from the phase analysis window. I usually have it open because it looks cool, but my ears work better for me than a visual tool."

This kind of says it all! I'm not actually experiencing any problems. I was more after an interesting discussion/thread about an AA2.0 feature that's intriguing me. I'm just curious about the information the phase analysis window offers and why it seems, to me, to be either something that provides serious and complex data I don't understand or maybe it's more a bunch of pretty colours that aren't really a useful tool.
Basically, I kind of think I know what I'm doing (mostly). But the PA Tool had me also thinking, "Wow, there's a whole bunch of important stuff I don't get".
Maybe not.
I can see how it might provide useful phase information for two or even more selected tracks, but a mixdown? Now, I'd say no.
So yep, it's back to the two Phase Analysis tools on either side of my head.

Cheers, Graeme.

It can tell you something about your stereo mix. e.g. if both channels are identical (i.e. mono) you will get a straight line angling upward at 45 degrees in Left/Right mode. If one channel is out of phase, the line will angle downward. For stereo files it should roughly follow the upwardly angling line. If the result is roughly circular or roughly follows the downwardly angling line you may have a problem. e.g. the latter case indicates a probable phase inversion of one channel.

Phase analysis is most helpful when mixing down elements to mono or when you know your recordings may get used that way. A multichannel recording with out-of-phase channels with similar content can get nasty when mixed down.  

Thanks Alan- and good point, Bruce. Especialy since I do the odd radio commercial. It'll be handy to see if they still reproduce on your standard, bashed-up brickie's radio on the bonnet of the ute!

Can also indicate a DC offset by the diagonal being unsymmetrical in the opposite quadrants. The phase display is only intended as a rough indication; it can only produce meaningful exact measurement data when the waves being compared are single-frequency pure tones. Also in respect to phase, the x-axis is usually time, but nothing says it has to be. The actual relation that exists has to do with the system through which the signal passes; this could be as diverse as air to light, to electrons, to virtual (i.e., digital manipulation)... each will have their complex mathematical function, which may or may not be linear or constant.

Its a big chunk to try to understand all at once. Fortunately there is a lot of material on the web to consult; it may not make sense all at once, but with diligence, it will.