question about speakers and phase

[Special K]

I don't have any standard speakers so I cannot test this out, but what would happen if you were to reverse the polarity of the cables at one end of one speaker (of a 2 speaker setup)? Reversing the polarity of an AC signal is equivalent to a phase shift of 180 degrees, right? So if I were to reverse the polarity of only the right speaker in a two speaker setup, the right signal would be 180 degrees out of phase with respect to the left signal, correct?

So a 5 kHz component of an audio signal would be shifted by (1/16,000) th of a second, right? (5 kHz = 10,000*pi rad/s, so a phase shift of pi radians = 1/16,000 second.)

That seems like it wouldn't even be perceptible to human hearing. Does this mean that the phase response of an audio amplifier really has little effect on how the music sounds? You would have to have a pretty dramatic phase shift to notice the difference, right?

 

[brentkiosk]

If you reverse the polarity of one speaker in a standard two speaker stereo system that is set up to produce stereo, you will definitely notice the difference. At the mid point between the two speakers, it will sound sort of dead. As you move to either side, you will hear the closer speaker to a much greater extent than you would with the corre ct speaker connections. In fact, if there were no reflections of sound from walls, floor, etc., with the speakers out of phase you would hear no sound along a line half way between the speakers.

The question about phase response of an audio amp is one debated by audiophiles. Many people say the sound is "better" for amplifiers with constant phase response (as a function of frequency). Others say the effect is pretty minor or totally unobservable. I think amplifier design tends to concentrate on low harmonic distortion first with phase respons being far down the list.

 

[Special K]

Originally posted by: brentkiosk
If you reverse the polarity of one speaker in a standard two speaker stereo system that is set up to produce stereo, you will definitely notice the difference. At the mid point between the two speakers, it will sound sort of dead. As you move to either side, you will hear the closer speaker to a much greater extent than you would with the corre ct speaker connections. In fact, if there were no reflections of sound from walls, floor, etc., with the speakers out of phase you would hear no sound along a line half way between the speakers.

The question about phase response of an audio amp is one debated by audiophiles. Many people say the sound is "better" for amplifiers with constant phase response (as a function of frequency). Others say the effect is pretty minor or totally unobservable. I think amplifier design tends to concentrate on low harmonic distortion first with phase respons being far down the list.

Why would the difference be that dramatic? What is wrong with my calculation of 1/16,000 second above?

 

[imported_electron]

Hmm, I highly doubt that one would ever hear dead silence in the middle of 2 speakers out of phase. It's true that 2 identical AC signals going into each other 180 deg out of phase would cancel each other out, and if you had a meter of sorts there instead of 2 speakers then they would cancel each other out completely. but in this case you have speakers and speakers turn AC signals into sound. What happens to that sound is what sets it apart from the cancelation you expect in theory that almost never happens in reality.

You see what will happen is while one speaker will be pushing it's cone out, the other will be pushing it's cone in. First of all, stereo recordings almost never put the exact same signal to both speakers anyway. So now that I think about it even that meter experiment will not produce 100% cancellations.

Not to mention all the reflections of sound you'd get not only from the walls of the room but also the speaker cabinets themselves.

So about the only way i can see perfect cancellation between the 2 speakers is if

1) Everything is in a perfectly noise absorbing room
2) Both speakers are fed the same exact signal only 180deg out of phase (note this is not what you'd typically have in a stereo musical recording or it would sound like mono.) and
3) both speakers are ideal perfect speakers with no reflections from their own enclosures. There are no speakers like that I've ever seen.

So in conclusion, while it is possible, it would almost never be seen in practice.

 

[Howard]

You only have to worry about it at lower frequencies and when you invert the polarity after the crossover.

 

[Born2bwire]

It wouldn't be dead silence, but it is very noticeable, particularly on the low end. It sounds like the sound is coming from elsewhere and not from the speakers (probably the reflections). There is a noticeable dead zone where the center drops away, though not completely.

As to why it happens, it doesn't matter that the time delay is very insignicifant. In ideal reproduction, the wave will experience perfect deconstructive interference and be negated. So despite being like a 1/16000 second time difference, you can't hear sound whose amplitude has been cancelled out.

As to phase response from amplifiers, it isn't significant if you do not have a phase difference between channels. More importantly, the next best thing to having no phase change is linear phase change. A nonlinear phase shift across the frequency band creates time delays and smears the sound. The rough math is associated with the derivative of the phase response of the signal. If the response is constant or linear (f(omega)=a*omega), then the phase shift is either zero or a constant factor at all frequencies. So even though there is a phase shift, it's applied evenly to the whole signal and it just comes across as the insignificant delay that you calculated. But if the response has an order greater than one, then the time shift is frequency dependent and then you may have problems. One example that has been posed is that in a particular set of ruins, if one claps their hands, the echo sounds like a bird chirp. The reason being is that for some reason, the frequency phase response of the reflected sound is nonlinear. The timeshift of the frequencies increases the width of the sound impulse (clap) such that instead of being a sharp sound it is now a slightly drawn out chirp. Needless to say though, your original problem statement is about constructive and destructive interference due to the phase difference between channels, not so much time delay.

 

[alcoholbob]

Phase is usually the last thing considered for most speaker manufacturers (as different order crossovers can affect phase) because its generally considered the least audible out of all the other quantifiable measurements, such as impulse, IR, imaging, driver ringing, etc.

 

[CrispyFried]

Originally posted by: Howard
You only have to worry about it at lower frequencies and when you invert the polarity after the crossover.

yup, its pretty easy to demonstrate.. switch the leads of one speaker and play a mono music track thats heavy on the bass. youll find the bass response is way down (unless your speaker placement is awful).

other than that, at higher freqencies in stereo youll find that imaging suffers, ie its "blurry."

 

[0roo0roo]

while back i found a saite with wav files of sounds in phase and outa phase. u can probably google it.

 

[hemiram]

I had a passive surround sound adaptor I built for my system back in 74 or so, you hooked up the two speaker cables to the input, and there were terminal blocks for the four outputs. There were several relays that switched the phasing between the two sets of right and left, and between the rears and fronts. For maximum bass, you wanted all of then in phase with each other, for a kind of "wide" sound, the front and back speakers were flipped to opposite phasing, and for truly bad/weird sound, you threw all of them out of phase and it sounded terrible.

I had a ton of bass with 4 12" woofers in a small room, so I normally ran it in the wide mode. It sounded better than a lot of the surround modes do on modern surround sound receivers, since there wasn't much being done to the signal.

On certain recordings, that little box did some amazing things. There was one record I had that had a horn sound as it it was floating around the room above your head.