Does Phase Distortion/Shift Matter in Audio? (no*)

[KSTR]

To evaluate the equation, it is very helpful to have a programming language available that can perform the required complex number arithmetic. Languages such as Octave, MATLAB, Python, and Fortran come to mind.

Or LTspice, it can handle transfer functions in direct Laplace notation.

 

[Tom Danley]

Please keep the answers online. The length isn't a problem. Thanks.

Hi
A point i was trying to get across is measurements are a "View" of some phenomena and are not necessarily the phenomena itself. Describing the measurements does not normally connect strongly to exactly what one hears first hand and is usually only a pale descriptor of what is experienced.
A couple years ago, back when there were still trade shows, Doug and I presented a couple classes at Infocomm. The challenge to me was demonstrate two speaker systems in stereo that had the same response but had entirely different stereo images. One system was to be a "single source" but NOT using speakers from work.

The difference between the two systems was greater than i had expected but the "good pair" are simple and easy, makes me wonder if there are those curious enough to try it themselves and see what a single point source without crossover sounds like.

If so, do this;

Locate and obtain 2 of these drivers

3FE22 (8Ω) by FaitalPRO

FaitalPRO combined the latest R&D methods and technologies with the most advanced industrial automation techniques and all FaitalPRO drivers are made in Italy with the highest level of quality, consistency and durability.

faitalpro.com

This is an unusually good full range driver, small enough to be an actual point source up to a couple Khz and obviously no crossover phase shift. It has an on axis rise up high but that is easy to eq out if desired. It also goes surprisingly loud but is not a subwoofer haha.
Get a 2 & 3/4 inch hole saw and two pc of 1/2 inch (nominal) plywood about 2 feet square, for the class i cut the corners off so they were unequal octagons or "stop signs".

Cut a 2&3/4 inch hole at the center of each.

Obtain 2pc of 1/2 inch (nominal) open cell foam and using spray contact cement (like Super 77) , adhere it to one side of each baffle.
Cut an X in the foam where the hole in the wood is and mount the driver on the foam side so that the rear pokes through the wood and the front, squashing the foam flat around the driver and making a nice contour in the foam..

As one is trying to make a single source in time and space, a back volume is needed, not for bass tuning (although you can do that) but filled to absorb the higher frequencies that can partially reflect back and re-radiate passing through the cone body as a weak "late" radiation. The idea here is there is one simple radiation in time and space with no crossover "all pass" phase shift.

If you put these at ear height and around 6 feet away and away from your room's walls, they can produce an amazingly good stereo image, especially if the hf rise is eq'd, can essentially disappear in the face of a very strong phantom image and have the properties i described in other posts. I hope people try this, it is eye opening.

The last 20 odd years, I have been trying to make speakers that do that but are +30 or +40dB louder, with constant directivity for large spaces. While we don't advertise at work or have a marketing dept, we have sound systems in many places including about half of the 100,000+ seat stadiums and the big sound / projection venues in Mouse land Orlando etc all those based on the single source in time and space objective in a full range horn

Portfolio Archive | Danley Sound Labs

Portfolio Archive | Danley Sound Labs

www.danleysoundlabs.com

Tom

 

[Spkrdctr]

Wow! You had me at Disney. If you have your equipment at Disney it must be top shelf stuff. Disney has "money is no object" deep pockets and they usually get very good stuff. They have been for the last 30 years or maybe 40 the largest US manufacturer of fireworks, all for their own in house use. Disney is really an interesting company along with the parks. I know next to nothing about current pro-audio as I am 30 years out of date. Pre-computer controlled systems. I know that is old! I'm glad you guys at Danley go the extra mile to make good equipment, we need good American engineering and construction. Not everything has to be made in China! I tip my hat to you!

 

[markus]

Hi
A point i was trying to get across is measurements are a "View" of some phenomena and are not necessarily the phenomena itself. Describing the measurements does not normally connect strongly to exactly what one hears first hand and is usually only a pale descriptor of what is experienced.
A couple years ago, back when there were still trade shows, Doug and I presented a couple classes at Infocomm. The challenge to me was demonstrate two speaker systems in stereo that had the same response but had entirely different stereo images. One system was to be a "single source" but NOT using speakers from work.

The difference between the two systems was greater than i had expected but the "good pair" are simple and easy, makes me wonder if there are those curious enough to try it themselves and see what a single point source without crossover sounds like.

If so, do this;

Locate and obtain 2 of these drivers

3FE22 (8Ω) by FaitalPRO

FaitalPRO combined the latest R&D methods and technologies with the most advanced industrial automation techniques and all FaitalPRO drivers are made in Italy with the highest level of quality, consistency and durability.

faitalpro.com

This is an unusually good full range driver, small enough to be an actual point source up to a couple Khz and obviously no crossover phase shift. It has an on axis rise up high but that is easy to eq out if desired. It also goes surprisingly loud but is not a subwoofer haha.
Get a 2 & 3/4 inch hole saw and two pc of 1/2 inch (nominal) plywood about 2 feet square, for the class i cut the corners off so they were unequal octagons or "stop signs".

Cut a 2&3/4 inch hole at the center of each.

Obtain 2pc of 1/2 inch (nominal) open cell foam and using spray contact cement (like Super 77) , adhere it to one side of each baffle.
Cut an X in the foam where the hole in the wood is and mount the driver on the foam side so that the rear pokes through the wood and the front, squashing the foam flat around the driver and making a nice contour in the foam..

As one is trying to make a single source in time and space, a back volume is needed, not for bass tuning (although you can do that) but filled to absorb the higher frequencies that can partially reflect back and re-radiate passing through the cone body as a weak "late" radiation. The idea here is there is one simple radiation in time and space with no crossover "all pass" phase shift.

If you put these at ear height and around 6 feet away and away from your room's walls, they can produce an amazingly good stereo image, especially if the hf rise is eq'd, can essentially disappear in the face of a very strong phantom image and have the properties i described in other posts. I hope people try this, it is eye opening.

The last 20 odd years, I have been trying to make speakers that do that but are +30 or +40dB louder, with constant directivity for large spaces. While we don't advertise at work or have a marketing dept, we have sound systems in many places including about half of the 100,000+ seat stadiums and the big sound / projection venues in Mouse land Orlando etc all those based on the single source in time and space objective in a full range horn

Portfolio Archive | Danley Sound Labs

Portfolio Archive | Danley Sound Labs

www.danleysoundlabs.com

Tom

I've been experimenting with full range drivers for quite some time but quite honestly I'm not convinced the perceptual properties you describe are caused by the lack of crossover or being a single source in space but because the phase and magnitude response between full range drivers – and therefore L and R speakers – is inherently very similar. You also get high directivity depending on the driver size. Furthermore if you're using a bigger full ranger like an 8" there also is a crossover involved, not an electrical one but a mechanical one. Currently using a Seas FA22RCZ.
I also tried to introduce deliberate phase distortion to simulate typical multiway speaker crossover behavior but the effect on imaging was not there. Always wanted to look at this in more depth but got sidetracked by real life

 

[Tom Danley]

I've been experimenting with full range drivers for quite some time but quite honestly I'm not convinced the perceptual properties you describe are caused by the lack of crossover or being a single source in space but because the phase and magnitude response between full range drivers – and therefore L and R speakers – is inherently very similar. You also get high directivity depending on the driver size. Furthermore if you're using a bigger full ranger like an 8" there also is a crossover involved, not an electrical one but a mechanical one. Currently using a Seas FA22RCZ.
I also tried to introduce deliberate phase distortion to simulate typical multiway speaker crossover behavior but the effect on imaging was not there. Always wanted to look at this in more depth but got sidetracked by real life

Hi
There is more than one way to produce the "simple radiation" i spoke of, the Faital driver does this with it's small size, up to around 5KHz and below that on a flat foam covered baffle is a simple source with no directivity other than the baffle which preserves the expanding bubble out to the edges.
The Seas driver shows cone issues beginning at a few hundred Hz is chaotic above a couple KHz. Up real high, the whizzer cone radiates most of the highs and to a very narrow angle and all of this produces the interference pattern shown in the frequency response.
These are both full range drivers by name but other than that with little similarity.
Hey these drivers aren't that expensive and the baffle is easy to make, i am sure you will be surprised how different and good that stereo image is.
Best,
Tom

 

[markus]

Hi
There is more than one way to produce the "simple radiation" i spoke of, the Faital driver does this with it's small size, up to around 5KHz and below that on a flat foam covered baffle is a simple source with no directivity other than the baffle which preserves the expanding bubble out to the edges.
The Seas driver shows cone issues beginning at a few hundred Hz is chaotic above a couple KHz. Up real high, the whizzer cone radiates most of the highs and to a very narrow angle and all of this produces the interference pattern shown in the frequency response.
These are both full range drivers by name but other than that with little similarity.
Hey these drivers aren't that expensive and the baffle is easy to make, i am sure you will be surprised how different and good that stereo image is.
Best,
Tom

While I don't own that particular Faital driver the smaller ones I have here and experience with is a Visaton FRS8, Tymphany TC9, Tangband W3-871, Vifa 10 BGS 119/8, Dayton RS100, Markaudio Alpair 7.3.

Sorry Tom, but I really don't think being a point source is that important to stereo imaging. Don't get me wrong, source size and position is important within certain limits and depending on frequency but once you listen in a stereo interference field and put the speaker in the room other factors are more significant. One of the most neglected and ignored ones is similarity of performance. I guess it comes as a welcome byproduct with well designed speakers like yours.

By the way, haven't been on your website for a while and was pleasantly surprised to see those two new (?) studio monitors. The specs download doesn't seem to work though?

 

[Tom Danley]

While I don't own that particular Faital driver the smaller ones I have here and experience with is a Visaton FRS8, Tymphany TC9, Tangband W3-871, Vifa 10 BGS 119/8, Dayton RS100, Markaudio Alpair 7.3.

Sorry Tom, but I really don't think being a point source is that important to stereo imaging. Don't get me wrong, source size and position is important within certain limits and depending on frequency but once you listen in a stereo interference field and put the speaker in the room other factors are more significant. One of the most neglected and ignored ones is similarity of performance. I guess it comes as a welcome byproduct with well designed speakers like yours.

By the way, haven't been on your website for a while and was pleasantly surprised to see those two new (?) studio monitors. The specs download doesn't seem to work though?

The "up side" about the scientific method is that a valid result from an experiment, if described in sufficient detail, can be repeated exactly by anyone and get the same results.
Those who repeat the experiment as i described will have an opinion based on hearing that result. It's a free country not to experiment or stick with what you believe based on a different experiment. But having made these and demonstrated them as part of a class at a trade show, i wouldn't have suggested this if i wasn't already sure of the results.

The key is, what you don't want is complex radiation AND that I think is what your hearing system can identify as "the location in distance" of the source when your eyes are closed. If it were a mono system this property doesn't matter but does when making a phantom image where you are trying to fool your hearing into hearing something between the two speakers. There are other aspect of a loudspeaker that effect this or limit how far one can go in say a generation loss test, but this spatial identity was a new one on me until i was noticing what it was like not to have as much of that.

The web site is new and a work in progress (ugh) and they add more stuff nearly every day.
Tom

 

[markus]

The "up side" about the scientific method is that a valid result from an experiment, if described in sufficient detail, can be repeated exactly by anyone and get the same results.
Those who repeat the experiment as i described will have an opinion based on hearing that result. It's a free country not to experiment or stick with what you believe based on a different experiment. But having made these and demonstrated them as part of a class at a trade show, i wouldn't have suggested this if i wasn't already sure of the results.

The key is, what you don't want is complex radiation AND that I think is what your hearing system can identify as "the location in distance" of the source when your eyes are closed. If it were a mono system this property doesn't matter but does when making a phantom image where you are trying to fool your hearing into hearing something between the two speakers. There are other aspect of a loudspeaker that effect this or limit how far one can go in say a generation loss test, but this spatial identity was a new one on me until i was noticing what it was like not to have as much of that.

The web site is new and a work in progress (ugh) and they add more stuff nearly every day.
Tom

You still seem to think I didn't do such experiments*? Not sure why.

Anyhow, I'm all for the scientific method and I don't question the results, I'm just sceptical about the theory explaining the perception. I think the reason is much simpler, namely better interchannel tracking. But don't take my word for it. A proper experiment could shed some light on this. Unfortunately the people interested and willing to do proper testing are scattered all over this planet and can't work on things in an effective manner.

---
*One of the informal experiments I did many moons ago involved comparing a 2-way speaker and a 3" full range driver in an open baffle very similar to what you described above. Listening was fairly near. I'd say less than 6' with the walls as far away as possible.

The perception was exactly as you described it. Much better imaging with the full ranger.

Then I equalized (with Acourate) both speakers so they had a very similar frequency response. This involved limiting the low fequency response of the 2-way speaker. At the same time I optimized the phase response within each pair so it was better matched. The phase response between 2-way and full range was NOT matched.

Surprising result: both speaker types now imaged the very same.

This drives my personal theory up to this day – frequency response (magnitude AND phase) of stereo speakers should track each other as close as possible. Overall phase response is largely irrelevant as long as it matches between stereo pairs.

Then I also compared imaging of the 2-way speaker with just the low frequency limit engaged. Imaging improved as well. Second take away for me: overall bandwidth has a significant effect on imaging. Here the room dominates what is heard, so great care needs to be taken at the low end (room treatments and EQ).

 

[j_j]

This drives my personal theory up to this day – frequency response (magnitude AND phase) of stereo speakers should track each other as close as possible. Overall phase response is largely irrelevant as long as it matches between stereo pairs.

As far as imaging, I think that's a somewhat safe thing to say, assuming time delay to both speakers is spot on, too, of course.

Overall phase response,though, can create some "interesting" effects, and having midrange inverted, even if the two speakers are spot on, can seriously hurt some of the time coherence binaurally. Does it matter? Probably, but maybe not a lot.

 

[markus]

As far as imaging, I think that's a somewhat safe thing to say, assuming time delay to both speakers is spot on, too, of course.

Overall phase response,though, can create some "interesting" effects, and having midrange inverted, even if the two speakers are spot on, can seriously hurt some of the time coherence binaurally. Does it matter? Probably, but maybe not a lot.

Hm, I can see how that could have an effect. Any suggestion how we could A/B test this? Split the signal with a LR-2 crossover (at what frequency?), invert the higher frequency part and sum?

 

[j_j]

Hm, I can see how that could have an effect. Any suggestion how we could A/B test this? Split the signal with a LR-2 crossover (at what frequency?), invert the higher frequency part and sum?

More like do 3-way version, so that midrange only is inverted. That might do it, on the exactly wrong material. It would have to be a very pitchy kind of thing, fuzz guitar, maybe a very open vocal tract, that sort of thing.

 

[luft262]

A viewer of my videos and member he suggested that I do a video commenting on a video that Paul McGown did on audibility of phase shifts. Here is Paul's video which was really about a different question (why we need wideband amplifiers) but turned it into phase being an audible problem:

Here is my answer to him:

Of course phase is an important electrical and acoustic thing and has relevance in countless situations. It is just that it should not be used to create myths and fear in audiophiles with respect to audibility in the context Paul and others are using.

Great video and explanation Amir. Excuse my ignorance, but does this mean that the 0 to 180 phase switch on the back of subwoofers is irrelevant or unnecessary? Side question, but does this mean phase rotation is irrelevant? Thank you for your videos Amir!

 

[markus]

More like do 3-way version, so that midrange only is inverted. That might do it, on the exactly wrong material. It would have to be a very pitchy kind of thing, fuzz guitar, maybe a very open vocal tract, that sort of thing.

Do you have any such sound files we could use without running into copyright issues? I could create a virtual 3-way crossover with the mid's polarity flipped and pre-render it on some sound files so we could A/B test them.

 

[j_j]

Do you have any such sound files we could use without running into copyright issues? I could create a virtual 3-way crossover with the mid's polarity flipped and pre-render it on some sound files so we could A/B test them.

Annoyingly, nothing "natural" that's copyright-free.

 

[ernestcarl]

Overall phase response is largely irrelevant as long as it matches between stereo pairs.

How about several pairs of multichannel speakers? It would seem to me that it is at least desirable to have all channels share similar enough phase profiles. It might also make bass management a little simpler.

 

[ernestcarl]

Great video and explanation Amir. Excuse my ignorance, but does this mean that the 0 to 180 phase switch on the back of subwoofers is irrelevant or unnecessary? Side question, but does this mean phase rotation is irrelevant? Thank you for your videos Amir!

There have been at least two threads discussing "phase polarity" at length. Try searching for the quoted keywords in the forum.

 

[markus]

How about several pairs of multichannel speakers? It would seem to me that it is at least desirable to have all channels share similar enough phase profiles. It might also make bass management a little simpler.

Sure. Wouldn't make bass management simpler though. From a computational standpoint it's irrelevant if there's like 8 AP's with the same parameters vs. having differing parameters.

 

[ernestcarl]

Wouldn't make bass management simpler though.

I was thinking of those systems esp. where all satellite speakers are the same, for example. Let's say the sub and xo unit were already designed to handle the bass management of the system from the get go as opposed to when one is just mixing and matching several different set of speakers and sub(s).

 

[markus]

I was thinking of those systems esp. where all satellite speakers are the same, for example. Let's say the sub and xo unit were already designed to handle the bass management of the system from the get go as opposed to when one is just mixing and matching several different set of speakers and sub(s).

Not sure I understand. You still need an AP filter per channel.

 

[ernestcarl]

Not sure I understand. You still need an AP filter per channel.

Ah, nevermind... But I get what you mean about the desirability of having all-pass filter(s). Just so you know, I do not use any in my current implementation and am perfectly fine -- as far as I am aware, there are no cancellations between any or all channels.