[Baumli]

Dear GAS gentry,

As many of you know, I enjoy throwing audiophile words around, and I certainly enjoy music, but often there are concepts I just don't understand in audio. Usually I understand the general meaning, but I get hung up on the details. One of these involves phase.

I know what it means to have speakers in phase with each other. I know that the "back-wave" of a dipole is out of phase with the "front-wave," and that with a bi-pole the forward information is in phase with the rear information, although by the time all that information coming from different directions gets bounced around in the room not much is in phase anymore. And of course, depending on some rooms, you can have speakers that are thoroughly in phase "with each other" and yet there is musical information all over the room that isn't in phase with other musical information in the room. Also, what is salient here, some people like a presentation that is relatively out of phase. A dipole or a 360-degree resonator has a lot of information that isn't in phase with the rest of the information, but this makes for a more omnidirectional sound, and all this resonance and reverb is exactly what some people like.

I prefer everything to be as in phase as possible. Hence, my choice of speakers, and the placement of my two subs, their crossover slope, etc. But I want to ask about other speakers.

I put my question to a renowned circuit-design engineer, and he said he just didn't know, that I would have to ask a speaker manufacturer. So I asked a speaker manufacturer, but he is from a different country, and either he wasn't understanding me or I wasn't understanding him or both. Regardless, I didn't get my answer. I asked an audiophile who has considerable knowledge of these things, and he said that in, for example, the Acoustic Zen "Crescendo," the information coming out of the transmission-line opening at the bottom is indeed out of phase with the speakers up above-- he suspected about 45 degrees--and that since this affects only the bottom frequencies few people would notice the absence of phase, and in fact since there is out-of-phase musical information in the lower frequencies reaching your ears this give a bass presentation that gives the illusion of having more bass. ("Illusion"--that makes me suspicious.) So I have pondered this matter considerably, and I'll try to here present my question as clearly and simply as possible:

If you have a speaker with normal drivers, but it has either a port or transmission-line opening at the bottom (let's say either are forward-firing to keep things simple), wouldn't this mean that the information coming out of the port is out of phase with what is coming directly from the drivers up above? After all, the information reaching the bottom is coming off the back of the drivers (primarily), and also there is a time delay caused by the cabinet dimensions which means the sound (compressed air) reaching that opening at the bottom isn't getting down there and out of there as quickly as what is coming from the driver. To my ears a speaker with a port, or with a transmission-line opening, just plain sounds out of phase. Maybe only slightly, but I am sure I hear it. Or am I hearing something else? So please tell me (no; please explain to me, thoroughly and patiently) what a bottom, forward-firing port or transmission-line opening does by way of keeping the musical information in phase, or perhaps making it out of phase.

Any cogent, educated, or omniscient elucidation on this matter would be very much appreciated.

All the best,

Francis Baumli

[michaelhigh]

I elect our daves to answer this in his most succinct way.

Or he'll Google it first. 

[JohnH12]

From a designer's perspective...acoustical phase is the time information associated the frequency coming from the source.  The phase information describes how the frequencies from the sources will sum. You obviously can’t hear phase but you can hear when the frequencies sum to create a bump and (maybe) a dip in the frequency response.

The speaker is (usually) designed with the response of the port included in the frequency information.  Most designs are for a flat first arrival frequency response.

So then the question becomes are you hearing changes in the frequency response because the port is out of phase, or because there’s a change in the frequency response due to the crossover design, age, room or placement issues?

I’m not discounting that the port is what you are hearing, but more often there’s a room interaction that is changing the speaker’s frequency response.

[G Georgopoulos]

most speakers don't have phase issues,
i have been at this game for over 35 years
and the only phase issue i know is
absolute phase,--source to speaker phase,
and this isn't an issue because you can
invert 180degress with an inverting amp
and the audio is the same...

[Baumli]

Dear folks,

I do not well understand the theory of all this, but I do not agree with what is being stated. First of all, I have been in this audiophile game almost 50 years, and I certainly can hear a difference when the switch on an amplifier is thrown, allowing inversion. I acknowledge that some people can't. And I'm not bragging in stating that I can. It's always puzzled me that some listeners can hear certain things, and others can't. For example, I can hear when a time-alignment between speakers and a sub are off by more than one millisecond. One afternoon spent with Harry Pearson many years ago showed that he couldn't hear a difference until it was off by over three milliseconds, although Frank Doris who was his right-hand man at the time could, like me, hear a difference if it was more than one millisecond. But there are things I can't hear that other people can. For example, I can hear a difference in CD players, digital links, and so on, but I have never, ever been able to hear a difference based on a CD treatment--whether it be green ink, damper rings, "shavers" that "round" the disc, etc.

And still, I don't understand what is being said about a port being "in phase" with the speakers. I'll try to accept on faith that it is--or at least that it can be. But how?! It seems to me that sheer physics--so basic as to be called mechanics--would dictate that the information from the back of the driver, by the time it makes its way down to that port, is going to be issuing from that port at a somewhat later time than what is coming from the front of the speaker. How could it be otherwise?

Even speaker placement in my listening room suggests something is going on. If I put ported speakers in there, the crossover slope to my subs has to be set at 18 degrees to be in phase with the speakers. If I use my non-ported Dunlavy's, then I set the slope at 24 degrees.

So allow me to humbly state: I would prefer explanations, not assertions. But I know I ask this at my own peril, because the moment things get technical, then I am lost in a thicket of concepts that elude my cerebrum.

Welcoming wisdom,

Francis Baumli

[G Georgopoulos]

Dear folks,

I do not well understand the theory of all this, but I do not agree with what is being stated. First of all, I have been in this audiophile game almost 50 years, and I certainly can hear a difference when the switch on an amplifier is thrown, allowing inversion.

Francis Baumli

Mr Francis you can hear the difference of inverted audio,are you kidding me??it's the same audio!!how can you hear a difference??... ...you must have phase sensitive ears!!

anyway i wish you find the answear of your question... 

[JohnH12]

Many people can hear when a speaker is out of phase with another speaker, but to be clear you're hearing a change in the FR. 

You can play with the crossover, move the speaker, tilt the speaker, or other time delays to move the source phase to get the FR you prefer.

The term "in phase" and "out of phase" is a bit loose as phase varies with frequency and observation point.  It's easier to talk about the resulting summation.

[Baumli]

Dear folks,

In answer to Georgopoulos--well, yes, I do have phase-sensitive ears. It is sometimes difficult to hear; other times it is obvious. Joe Grado's ability to hear phase shift was much better than mine. I've known other people who can't hear it at all.

But on further reflection to what was exchanged recently: The fact remains that even if one inverts the phase of the amplifier 180 degrees, then it would remain the case that if the port of a speaker is out of phase with the information coming from the driver, the same degree of mismatching in phase information would be involved. I.e., if A (the driver) and B (the port) are out of phase with each other by 25%, then if you invert both A & B a full 180 degrees, then they both will remain 25% out of phase with each other.

So this question remains moot (in the correct meaning of the word "moot"--which almost no one uses). But no; come to think of it, the question doesn't remain moot. It remains entirely unaddressed ... so far.

I am hoping to get an answer to this question on our forum. If I don't, then I suppose I'll start phoning speaker manufacturers whom I don't even know.

Optimistically,

Francis Baumli

[JohnH12]

Francis it's too bad you have dismissed my responses, as perhaps we could have met and demonstrated measured phase.

I'm still hopeful of a local GAS GTG.

John

[Baumli]

Dear John,

Please be assured that I have not dismissed any response you gave. I didn't even see your response until tonight. I am only getting a few of the notices about responses (the usual glitch in this site!), plus there has been a problem with me being right in the middle of trying to use this site and I get logged off. So I am now hoping for a brief stay on this site. And please don't be miffed about my supposedly dismissing your response. Your way of dealing with this matter seems cogent, clear (if complicated), and promising. So I am nothing less than grateful. I'll do my best to respond to what you have to offer. (And at the moment another intervening variable is menacing--a lightning storm is coming in.)

Your words are trenchant in stating that what counts is the "summation" of the frequency response. And indeed this can be altered by many factors, not least being the room response, positioning of the speakers, etc. And you state what I myself was getting at: what one wants to start with is a flat frequency response coming from the drivers to the listener. Then (and the "then" is upon one almost immediately) the other variables intervene, especially the room interactions.

I would love to have a conversation about this, and perhaps we could arrange to do so? My phone # is: (314) 966-2167. (I try to be available by phone afternoons and evenings.) Perhaps subsequently a listening session?

But aside from a conversation, many equations, and the "ideal" summed response, my "theoretical" question remains: When a cabinet is designed, how do you make that cabinet so that what is coming off the back of the speaker is going to be leaving the port at the same moment the information from the forward-facing driver is moving toward the listener?

Thank you for your patience. I'm glad to know we're again having a dialogue.

All the best,

Francis Baumli

[JohnH12]

Hi Francis, maybe consider removing your phone number before the spammers find it. 

I apologize, I'd like to give a more detailed answer but I'm short on time until next week. The short answer is for ported speakers we measure the combined response of the driver plus cabinet and port.  Any port leakage is swept up the measurements. 

[Rusty Jefferson]

......If you have a speaker with normal drivers, but it has either a port or transmission-line opening at the bottom (let's say either are forward-firing to keep things simple), wouldn't this mean that the information coming out of the port is out of phase with what is coming directly from the drivers up above? After all, the information reaching the bottom is coming off the back of the drivers (primarily), and also there is a time delay caused by the cabinet dimensions which means the sound (compressed air) reaching that opening at the bottom isn't getting down there and out of there as quickly as what is coming from the driver. To my ears a speaker with a port, or with a transmission-line opening, just plain sounds out of phase......

No, not necessarily.  Here's a link to to the Wikipedia page about Transmission Line Speakers.  https://en.wikipedia.org/wiki/Transmission_line_loudspeaker   Note this statement: ......"The length is specified to reverse the phase of the rear output of the drive unit as it exits the vent. This energy combines with the output of the bass unit, extending its response and effectively creating a second driver......"

Also, some speakers will have a (low) woofer out of phase with a (mid) woofer and tweeter to improve response at the crossover point between the 2 woofers.

..... To my ears a speaker with a port, or with a transmission-line opening, just plain sounds out of phase. Maybe only slightly, but I am sure I hear it. Or am I hearing something else?........

I don't know what you're hearing.  In some cases you may be hearing time domain issues.

.... For example, I can hear when a time-alignment between speakers and a sub are off by more than one millisecond.......

If you can hear this in a controlled double blind test, then there probably isn't a speaker type whose natural anomalies you can't hear, or be bothered by. 

[Baumli]

Gentlemen,

Thank you for your several responses. There are times when, yes, I am hearing time-domain problems. As I stated before, there are a lot of imperfect artifacts going on when reproducing sound. These confuse matters, I know, and sometimes one can identify what the confusion is and what isn't.

I looked at the entry on Wikipedia for transmission line speakers, and I have two problems with it: First, the prose is awful, and I can't be sure of what is being asserted. Second, the content consists primarily of assertions instead of explanations utilizing even basic math. So I am not satisfied with this; I want a better explanation before I am going to believe that what is emitted from the open end of a transmission line "port" is actually in phase with what is coming from the driver.

I received a phone call from an audio wizard this evening which was vastly informative (and enjoyable); I will pursue what he explained to me further. Meanwhile, I shall await the promised phone call, with all due patience. And again, I thank all of you for your input.

Now, for some music. No. I don't have time. Maybe instead some chocolate? Nothing like this "food of the gods" to stimulate the mind and calm the nerves.

In the spirit of incipient somadi,

Francis Baumli 

[daves]

Francis, you are trying to wrap your head around one of the nasties in audio.

How can a music impulse that is created in space, either in a 2Pi or a 4Pi geometry, be successfully modeled and reproduced by an eight to fifteen inch paper/composite Frisbee, with the least amount of time and phase artifacts?

As long as you are dealing with a port, or 1/4 wave transmission line, you will notice at least a hint of breathing or chuffing at some point. In poor examples, it is glaring. Speakers where the pistonic woofer travel is too large, and the inertial moment of travel is too great to effectively control, is going to have bloated bass. The AR3 is a perfect example of this.

A second example would be a backloaded horn speaker, like the JBL C40 Harkness I have in the family room. The backloading effect is akin to a screen door - on a bass note, you have the forward piston pulse as the primary wave, and then as the woofer rebounds back, you have the secondary wave. The secondary wave travels about four feet further than the primary before it exits the plane of the speaker, so you will ALWAYS have a hinge effect(delay) in the bass region of 1.23 meters/343.2 meters/sec, or 0.003586 seconds...3.6 milliseconds.

This cuts two ways with you. Since you have a threshold acuity of 1 millisecond, you will definitely hear a phase discontinuity at 0.001 second X 343.2 m/sec = 0.3432 meters, or 13 inches. Any speaker that has a port, or is horn loaded, backloaded, or has a transmission line where the center point of the woofer to the longest exit point, is greater than 13 inches, will make you cringe.

The calculation I did is simply travel distance/speed of sound.

The second cut is you will most notice this with the bass, because the travel path of a single wave is so large. To wit,
Hz            Travel length of a single full wave
20            56 feet
40            28
80            14
160          7
320          3.5
640          1.75
1280        10.5"
2560        5.3"
5120        2.64"
10240      1.32"
20K          0.68"

Dealing in hypothetical situations, let's calculate a three way crossover time error with a 10% phase shift, crossover points of 400 Hz and 5000 Hz.
At 400 Hz, the wavelength is 0.8575 meters. A 10% phase error in time domain, worst case,  would be 0.94325/343.2 meters, or 2.7 milliseconds. Best case would be 0.77175/343.2 meters, or 2.2 milliseconds.

At 5000 Hz, the wavelength is 0.0686 meters. A 10% phase error, worst case, would be 0.22 milliseconds, while the best case would be 0.18 milliseconds.

The other confounding aspects are the amplifier damping factor interaction with woofer impedance, and the woofer composition re weight and fabric stiffness/strength.

Sorry for the dissertation, and sorry for any conceptual errors, but this will get you thinking about the physics, which is where a lot of the problems roost.

[JRace]

With a port you are not hearing the back-wave leaving the port (which would as you suggest, be late compared to the original front wave.

Instead think of the air sitting in the port as the "other driver". This air responds almost as quick as the front wave. Yes, is is out of phase, but slightly. Given the lower freq nature of the port our brain is less concerned with phase information (or less bothered by it).

[daves]

JRace, the air sitting in the "port" of my JBL C40 is directly behind the woofer, not sitting at the exit point. This is a mechanical system. ALL air, when dealing with speakers, must obey the speed limit, which is measured at 343.2 meters per second on Earth. Because air is compressible, and because the amount of air moved is strictly defined by the swept travel volume of the speaker piston, there is a time delay. This delay is what Francis, and some other people, have been blessed or cursed with hearing.

Before we had computers and TS parameters to aid in speaker design, volumes and ports were often poorly designed, Ports were often too small, too large, too shallow, or too close/too far from the woofer(s.) We had slot ports, and speakers with twenty ports, and speakers with no ports. I have heard them all, and any number of them have chuffed, whistled, barked, moaned, and vibrated. These suckers have to be tuned, they have to deliver just the right air capacity of laminar flow air for the woofer to hit its optimal back pressure for its size and travel. Then you need to have the right amplifier with the proper damping factor for everything to live in the listener's sweet spot. I think this is why Francis is running his solid state McCormack DNAs.

[JRace]

I was referencing this from the OP:

If you have a speaker with normal drivers, but it has either a port or transmission-line opening at the bottom (let's say either are forward-firing to keep things simple), wouldn't this mean that the information coming out of the port is out of phase with what is coming directly from the drivers up above? After all, the information reaching the bottom is coming off the back of the drivers (primarily), and also there is a time delay caused by the cabinet dimensions which means the sound (compressed air) reaching that opening at the bottom isn't getting down there and out of there as quickly as what is coming from the driver.

I agree there is a delay, just not as severe as possibly expected.

Keep in mind that we do not process phase information below 100 Hz very well in the first place.

The air in the port is the driver, and it is being 'pushed' by the back wave from the speaker.

[daves]

And the back wave from the speaker is a motive force from the damping factor of the amp reacting with the magnetic structure of the speaker, influenced by the construction of the speaker enclosure. This is one of the largest rabbit holes in speakerdom.

We aren't going to argue over the significance of the delay. Francis started the thread, he hears the delay and recognizes it as a problem for him, and he asked for an explanation. I merely am giving him some tools to get part of an effective answer for him.

[KenSeger]

What a can of worms.

First, GG, the ability of a person to detect absolute polarity is the Wood Effect, I can't remember his first name.

Francis, I'm sure you have run across the sub-continent Indian saying, "Chess is pool which a gnat can drink dry or an elephant can swim in." Well, 'phase' in hi-fi is not quite that, and by more than half a bubble off.  More like an ocean that swept the poor elephant away to its doom via a tsunami.

Let me don my swimming trunks and try not to emulate the elephant.

I'll type up something that covers resistance, capacitance, inductance, the phasor angles of those three, their mechanical analogues, and expand upon what DaveS was saying about wavelengths.

Oh, and I actually did read the first four entire paragraphs of the Wikipedia article on transmission lines.  I consider that penance for all sins, both of commission and omission, for my previous 64 years and for any sins in my future. Good grief Charlie Brown. I knew there was trouble ahead when the author used (generally folded) and (usually folded) in back to back sentences.  That is a classic example in which a sentence of errors takes a chapter of corrections.  Not unlike the catch phrases on bumper stickers during any election cycle.
More later,
Ken

[Russell Dawkins]

And the back wave from the speaker is a motive force from the damping factor of the amp reacting with the magnetic structure of the speaker....

Really?