Why absolute phasing matters to you...

Imaging is a matter of psychoacoustics, fooling your brain into thinking there is a live performance in front of it, with a height, depth, and width..
It is a matter of sending a signal to your ears which is realistic, that duplicates what the actual event would have sent to your ears, as exactly as possible!

This may seem simple, but it can get complex... And most of the issues that can impede your efforts to accurately reproduce the true sound don't have to do with frequency response or amplitude. The reason that it is more difficult to establish proper imaging in a car has everything to do with "absolute phasing"... which is the phase relationship of the signals coming from all speakers in your car playing common frequencies as they combine at your ears.. Or, in layman's terms, it's how the sound from all the speakers blends at your ears.

This balance can easily be disrupted by many things; having an offset listening position, having multiple sets of speakers reproducing the same frequencies, and even having multiple speakers in the installation that are properly separated with crossovers to isolate the frequencies that they play.

This isn't to say that you can't make a car image properly, and have correct frequency response if you have an offset listening position.. ...or rear speakers.. or a center channel.. or three-way component sets, but the issues that need to be dealt with to achieve success given these inherent handicaps are signifigant, and should not be disregarded.

First, why "absolute phase" matters more than "relative phase":

Relative phase - Speakers are fed AC electricity of varying frequencies. When "correctly" wired, when the positive portion of the current reaches the speaker, the cone moves out.. and when the negative portion of the current reaches the speaker the cone moves in (this happens hundreds and thousands of times per second, depending on the frequency!). If you wire one speaker "backwards" (+ and - inverted) and one forwards, relative to the each other they are wired relatively "out of phase". If both are wired in a like manner, they are relatively "in phase"

Absolute Phase - This has everything to do with how the sound generated from the speakers reaches your head, which may have very little to do with how they are actually wired. If you had exactly the same pathlength between your two speakers, the sound would always reach your head at the same time from your two speakers. This would mean the "absolute phase" of the two speakers would be perfect.

However, if one speaker is further away than another, the sound from the farther speaker will take more time to reach your ears. Even if that is just hundredths of a second, it can mean that a pulse of positive pressure (from the speaker cone moving out) of the one speaker to be reaching you while a pulse of negative pressure from the other speaker is reaching you at the same time, and the end result is that they cancel each other out. This would mean that the "absolute phase" of the two speakers is "out of phase" at that frequency.

Interestingly, if this scenario exists, it is frequency dependent... meaning that some frequencies will be "in phase" while others will be "out of phase", due to the way that each frequency's wavelength lines up in combination across the two pathlengths in effect.



Illustration showing how two speakers, relatively wired in-phase, can yield results that are absolutely out-of-phase:



Example
If at a certain frequency, exactly 5 waves fit on the left side pathlength, and exactly 6.5 waves fit on the right side pathlength, the waves as they arrive at your head are exactly 180 degrees out of phase with each other, and they will cancel.

Now, with the same pathlengths, at another frequency, a higher frequency, maybe at that higher frequency exactly 9 waves fit on the left side pathlength, and exactly 12 waves fit on the far side pathlength. At that frequency, the waves combine constructively.

Any time you have two speakers playing the same frequency, of different pathlengths, phasing anomalies will occur that make some frequencies cancel, while others combine constructively.


Why absolute phasing is an issue with imaging:

Waves from multiple speakers (common frequencies) combine when they are in phase, cancel when they are out of phase, and combine or cancel to varying degrees depending on exactly how "in phase" or "out of phase" they are.

When you have two speakers of inherently different pathlengths playing the same frequencies (like front and rear speakers), at certain frequencies the wavelengths will line up and combine properly, at other frequencies they will combine negatively and cancel to some degree.

The higher the frequency, the more often in the frequency spectrum this happens... more phasing anomalies per octave. Also, the higher the frequency, the more sensitive your hearing is to minor distortions and disturbances in the sound. This is why people often suggest at a minimum running a low-pass filter on the rear speakers... to keep the damage constrained to least impacting range as possible.

Your subconscious is the key behind imaging...

Interestingly many of the differences in the reality of the sounds being heard are not factors that are easily listened for by the conscious ear, even to a trained acoustical engineer.. However, it is fairly easy to sit between two speakers for some critical listening to consciously decide whether you think it sounds "real" or not!

It is your subconscious which determines if you believe the sound or not, your conscious mind is simply aware of the status of what it is hearing... Just like your sense of balance is subconsciously controlled, and your conscious mind is simply aware when you are losing your balance.

The biggest damage that occurs when multiple sound pathlengths are encountered from your listening position is that your subconscious is not fooled, the "image" is not there like it could be. The cancellations are so narrow in bandwidth that you would most likely be fairly hard pressed to detect any missing frequencies or harmonics in the sound, yet those are subtle clues that your subconscious will pick up on and simply "will not be fooled"!

Try this experiment at home, to see how phasing fools your subconscious, without necessarily seeming to mess up frequency response...

Sit between two speakers at home, equal distance. (And verify your speakers are wired properly first!!)

Play some good reference music, or even better, talk radio.

The voice (or music) should sound well centered, the voice emanating from exactly between the speakers.

Now, shut the stereo off, and flip the phase on just one speaker (swap the + and - wires).

Now listen again.
See?
The music and voice still sound the same, you can hear all the content, haven't completely seemed to lose any tones..
But now there is no "image".. it sounds diffused, unrealistic, smeared, right?

THAT'S the sort of damage phasing anomalies can cause...

Common Questions:

These are actual questions that I collected from www.caraudioforum.com:

You nailed it dead on...
You can make the near speaker quieter...
But that doesn't do a thing about the phasing problems! maybe minimize them somewhat since one speaker isn't as loud, but that's not 100%.

The best thing you can do is equalize the pathlenghts as much as possible...

Kick panels are better (more equal) than low door locations...
Low door locations are better than dash locations...
Dash locations are better than A-pillar locations...
A-pillar locations are better than high door locations...
You see what I mean?

Another factor to consider is that speakers have directionality properties, meaning they are louder when listened to on-axis..

Since the far side speaker is more on-axis to you than the near side speaker, it should sound louder as well...

Effectively pulling off your "balance knob" trick, without any balance adjustment required..
And without screwing up the image for the passenger side of the car...

Oh yeah, if I do ANYTHING to artificially adjust the amplitude or phasing on that near side speaker...
...to balance it down a little..
...or maybe I am getting really trick and delay it a little so it effectively sounds further away...
In doing this, I'm screwing up the passenger side image all that much more!

Hmm...


Ahh, challenges!



Written by Geolemon

http://www.teamcaf.org/geolemon