"The Clue" - ny amerikansk högtalare

[paa]

Hello James,
Earl Geddes seems to say this is as good sound distribution as you need;



Could those curves be close to any goal you would want to reach?
Is it ok to have that much wider dispersion below 1 kHz than above?
Maybe the bass couples to the boundries and thus tends to narrow its dispersion a bit more than higher frequencies does?

[Jacro]

(By the way, the post of Naqref’s stopped after only 2 pages… Was there further work and measurements reported on the experimental devices?)

Hello James,

Finally I had som time (and focus) to read the whole thread and I'm very impressed as many others.

Regarding My little sweetheart it's scrapped but I'll think I have some measurement of it somewhere in one of my computers. Perhaps I'll find it and then I'll post some information here (and in the original thread).

The first ambience-module with the principle was made around 1995. A couple of years later I built the one on the pictures. I was still a poor student at the time so there was actually no money to develop the principle then. And then I found other things more promising.

I found your patent:
http://www.freepatentsonline.com/7551062.pdf
Impressive. I'll think the right way to go ist to use multiple active units just as you say. Passive damping of the sound from the back of the unit is to test intensive to get right.

I am like you intrested in what I call wall integrated speakers. It should perhaps be altered to external wall integrated speakers (EWI) or something like that to avoid beeing misstaken for in-wall-speakers. But anyhow. I'm the designer of Larsen speakers (it's a follow up of Carlsson). Two major problems with ewi-speakers is the perception of depth and excitment of roomresonances I'll think. The problem with depth is that the mind has problems to project auditory sources beyond the visual boundaries. The eye's override of other senses is making it harder to imagine a sound stage beyond the front wall. Do you think this is a big problem and if so what else is there to do about it other than removal of visual clues or to give additional visual clues to aid the sound stage?

The second problem (actually it's really a problem with almost every type of speaker), what is your preffered method of dealing with that? Acoustical lf-absorbers, parametrical eq or perhaps using multiple sound sources to counter act the resonanses?

Send my regards to Lars Erickson.

Hi Naqref,

Good to hear from you.

You mention two issues related to “EWI” loudspeakers:

(Over here we differentiate from "In-Wall" by calling them either; "On-Wall" or "Boundary Coupled"):

1) Perception of depth

2) Excitement of Room Resonances

I'm not clear on what you are referring to with #2. I think you are referring to reflections off the near boundary, but that is not actually a "resonance" per say. So, to discuss number 2, it would be good if you can clarify for me.

So, in terms of depth perception, in one way it is a problem (difficulty in imagining depth through a wall) and in another way, it is an improvement, in that it doesn't create "false depth" due to reflections off of the wall, spaced some distance behind the loudspeaker.

Loudspeakers spaced away from the front wall tend to create some degree of a false sense of depth depending on the ratio of the direct energy vs. the intensity of the energy reflected back to the listener.

So, in this regard, the boundary coupled systems are more faithful to reproducing the actual depth captured on the recording. (This is an over simplification, due to there being many additional variables determining perceived depth).

Your other observation, of the visual wall blocking our ability to imagine depth behind the wall, can be a significant limitation with some listeners. It is not a universal problem, but certainly a problem for many. Initially, I found it to be problematic for myself.

Sometimes I listen with my eyes closed, when listening to reproduced sound through loudspeakers, and also when I am at a live concert. In both cases, I find that the visual aspects alter my perception of the aural event, so closing my eyes creates a more equivalent situation for comparing live vs. reproduced sound.

If I have my eyes open at a live concert, I perceive the individual instruments as being more focused in a specific location in space than if my eyes are closed. Also, with eyes open, I "visualize" more depth of field, commensurate with the actual stage depth. If I close my eyes at a live concert, then the location of instruments is more vague, less focused, and the perception of depth of field is less.

So, if we are attempting to transport the actual focus, and scale of the sound sources to our listening room, without the actual musicians being visually present, the boundary coupled system is actually a more “accurate” facsimile of the spatial dimension of the original event.

(In a smaller second venue, boundary coupled loudspeakers and loudspeakers out in the room both distort the image compared to the original, but the loudspeakers that are out in the room (unless highly directional) will add more artificiality to the reproduced sound.

To minimize the visual barrier (for some customers that find it bothersome), but still have a boundary-coupled relationship for the loudspeakers, I have developed an alternative use-model.

Let’s see if I can explain it without a drawing.

Instead of facing the front wall of the room, the listener faces into a corner of the room. As an example, the listener might sit 3-meter’s straight out from the corner, and the left loudspeaker is coupled to the wall that extends to the left of the corner, and the other loudspeaker is coupled to the wall that extends to the right of the corner. The distance between the two loudspeakers and the distance from the loudspeakers to the listener are the same as they normally would be, and the loudspeakers are angled, or aimed, in relation to the listener, as they normally would be.

Visually, the corner extends back behind the loudspeakers, but the loudspeakers are effectively boundary coupled to each wall as they normally would be on the front wall.

I usually put extra absorption/diffusion in the corner, such as a quarter round of a half-meter radius, covered with 100 mm of foam, and I use a larger piece of absorption on the wall where the loudspeaker is coupled.

With an angled configuration, such as the Larsen 4, 6, or 8, the right loudspeaker may work better on the left side, and visa versa.

Make sense?

If properly set up, this has a few advantages, two of which are; it provides a visual expansion of depth, and it can leverage a more effective boundary coupling (less than a half-space) which allows a small loudspeaker to better power a larger room.

I use this approach for listeners that find the wall boundary mounting to be problematic for imagining depth, and/or for listeners who have rooms that are larger than what the loudspeaker is normally capable of filling.

This is a rather unorganized, non-technical discussion of the issue that you raised, but hopefully it is still somewhat useful.

I’ll address the rest later.

Best regards,

- James

PS – I’ll be sure to say hello to Lars for you.

[Jacro]

Hello James,
Earl Geddes seems to say this is as good sound distribution as you need;



Could those curves be close to any goal you would want to reach?
Is it ok to have that much wider dispersion below 1 kHz than above?
Maybe the bass couples to the boundries and thus tends to narrow its dispersion a bit more than higher frequencies does?

paa,

I have to run to a meeting, so a short answer for now.

While I believe that Geddes radiation pattern of constant directivity above 1 kHz is a big improvement over most free-standing loudspeakers, I disagree with Geddes suggesting that it is not important to maintain the constant radiation pattern below 1 kHz. A 12 dB transition below 1.5 kHz still impacts the “room sound”.

Geddes admits that he is not sure that it is important to improve it further, but states that, “even if it is important, it is not practical to do so”.

I humbly disagree.

But, I must also say, that Geddes has done some excellent work, his approach being superior to 99% of the loudspeakers available today, and he is one of the few actually advancing the art.

So, I agree that his configuration is an improvement over most of the other approaches, I merely feel that he has stopped short of what can be achieved on an ideal basis… definitely with his actual device, and possibly with his belief.

He is on the right track… he just needs to take it a bit further.

More later…

- James

[Jacro]

I am currently working on a new type of low frequency system.

Hell James,

This sounds exciting! Is this something you can elaborate on?

JanBanan,

I'm sorry, but that is confidential at this time.

If and when it can be disclosed publicly, I will be glad to discuss it here.

- James

[paa]

Thanks James for your answer about directivety below 1kHz.
Now imagine that I manage to make loudspeakers that maintain directivety down to say 80Hz, and want to add some subwoofers to them:
If I place two subs in the front corners, they will be boundry coupled in their frequency range and have 90+90 degrees of dispersion which should be fine. But probably we play the subs in mono so now we have 180 degrees bass dispersion which is quite a bit more more than the theoretically perfect main speakers, how do we sort this out? And if we want one or two more subs to improve on room modes, how would all that effect frequency response vs power response in such a system?

[LypsylateX]

I hope that the Google translator doesn't provide too many embarrassing mistakes. I often translate from English to Swedish, and then translate the Swedish back to English again, to see what it says, and the errors can be quite entertaining.

It helps that these errors is mostly one-way, i.e. when you're reading swedish threads. I'd say that 75% (just a guess) around here have no problems using english as primary language on a forum, and half of those should have no problem with quite advanced technical english.

Not a single lecture in my university education, in sweden, were in swedish. It was all in english. That helps. I guess this goes for most of the other people here with any form of higher technical education aswell. For me it's not the language barrier that sometimes keeps me from understanding threads like these, it's lack of physics knowledge.

My beleif is that this kind of participation from you could be alot more difficult in some other languages. (Like... well... french for instance. That would be very difficult for the majority of us swedes aswell, for the same reason.)

[Naqref]

(By the way, the post of Naqref’s stopped after only 2 pages… Was there further work and measurements reported on the experimental devices?)

Hello James,

Finally I had som time (and focus) to read the whole thread and I'm very impressed as many others.

Regarding My little sweetheart it's scrapped but I'll think I have some measurement of it somewhere in one of my computers. Perhaps I'll find it and then I'll post some information here (and in the original thread).

The first ambience-module with the principle was made around 1995. A couple of years later I built the one on the pictures. I was still a poor student at the time so there was actually no money to develop the principle then. And then I found other things more promising.

I found your patent:
http://www.freepatentsonline.com/7551062.pdf
Impressive. I'll think the right way to go ist to use multiple active units just as you say. Passive damping of the sound from the back of the unit is to test intensive to get right.

I am like you intrested in what I call wall integrated speakers. It should perhaps be altered to external wall integrated speakers (EWI) or something like that to avoid beeing misstaken for in-wall-speakers. But anyhow. I'm the designer of Larsen speakers (it's a follow up of Carlsson). Two major problems with ewi-speakers is the perception of depth and excitment of roomresonances I'll think. The problem with depth is that the mind has problems to project auditory sources beyond the visual boundaries. The eye's override of other senses is making it harder to imagine a sound stage beyond the front wall. Do you think this is a big problem and if so what else is there to do about it other than removal of visual clues or to give additional visual clues to aid the sound stage?

The second problem (actually it's really a problem with almost every type of speaker), what is your preffered method of dealing with that? Acoustical lf-absorbers, parametrical eq or perhaps using multiple sound sources to counter act the resonanses?

Send my regards to Lars Erickson.

Hi Naqref,

Good to hear from you.

You mention two issues related to “EWI” loudspeakers:

(Over here we differentiate from "In-Wall" by calling them either; "On-Wall" or "Boundary Coupled"):

1) Perception of depth

2) Excitement of Room Resonances

I'm not clear on what you are referring to with #2. I think you are referring to reflections off the near boundary, but that is not actually a "resonance" per say. So, to discuss number 2, it would be good if you can clarify for me.

So, in terms of depth perception, in one way it is a problem (difficulty in imagining depth through a wall) and in another way, it is an improvement, in that it doesn't create "false depth" due to reflections off of the wall, spaced some distance behind the loudspeaker.

Loudspeakers spaced away from the front wall tend to create some degree of a false sense of depth depending on the ratio of the direct energy vs. the intensity of the energy reflected back to the listener.

So, in this regard, the boundary coupled systems are more faithful to reproducing the actual depth captured on the recording. (This is an over simplification, due to there being many additional variables determining perceived depth).

Your other observation, of the visual wall blocking our ability to imagine depth behind the wall, can be a significant limitation with some listeners. It is not a universal problem, but certainly a problem for many. Initially, I found it to be problematic for myself.

Sometimes I listen with my eyes closed, when listening to reproduced sound through loudspeakers, and also when I am at a live concert. In both cases, I find that the visual aspects alter my perception of the aural event, so closing my eyes creates a more equivalent situation for comparing live vs. reproduced sound.

If I have my eyes open at a live concert, I perceive the individual instruments as being more focused in a specific location in space than if my eyes are closed. Also, with eyes open, I "visualize" more depth of field, commensurate with the actual stage depth. If I close my eyes at a live concert, then the location of instruments is more vague, less focused, and the perception of depth of field is less.

So, if we are attempting to transport the actual focus, and scale of the sound sources to our listening room, without the actual musicians being visually present, the boundary coupled system is actually a more “accurate” facsimile of the spatial dimension of the original event.

(In a smaller second venue, boundary coupled loudspeakers and loudspeakers out in the room both distort the image compared to the original, but the loudspeakers that are out in the room (unless highly directional) will add more artificiality to the reproduced sound.

To minimize the visual barrier (for some customers that find it bothersome), but still have a boundary-coupled relationship for the loudspeakers, I have developed an alternative use-model.

Let’s see if I can explain it without a drawing.

Instead of facing the front wall of the room, the listener faces into a corner of the room. As an example, the listener might sit 3-meter’s straight out from the corner, and the left loudspeaker is coupled to the wall that extends to the left of the corner, and the other loudspeaker is coupled to the wall that extends to the right of the corner. The distance between the two loudspeakers and the distance from the loudspeakers to the listener are the same as they normally would be, and the loudspeakers are angled, or aimed, in relation to the listener, as they normally would be.

Visually, the corner extends back behind the loudspeakers, but the loudspeakers are effectively boundary coupled to each wall as they normally would be on the front wall.

I usually put extra absorption/diffusion in the corner, such as a quarter round of a half-meter radius, covered with 100 mm of foam, and I use a larger piece of absorption on the wall where the loudspeaker is coupled.

With an angled configuration, such as the Larsen 4, 6, or 8, the right loudspeaker may work better on the left side, and visa versa.

Make sense?

If properly set up, this has a few advantages, two of which are; it provides a visual expansion of depth, and it can leverage a more effective boundary coupling (less than a half-space) which allows a small loudspeaker to better power a larger room.

I use this approach for listeners that find the wall boundary mounting to be problematic for imagining depth, and/or for listeners who have rooms that are larger than what the loudspeaker is normally capable of filling.

This is a rather unorganized, non-technical discussion of the issue that you raised, but hopefully it is still somewhat useful.

I’ll address the rest later.

Best regards,

- James

PS – I’ll be sure to say hello to Lars for you.

What you say makes a lot of sense. I hade tried to convince people of using the adjacent walls in a corner and shift right and left speaker in some cases but the look at me like they think; "Are you crazy or something" Perhaps I should work on my way of convincing people more to adress this problem.

Anyhow I think the problem is non existing when you have a picture of some sort between the boundary coupled (better word than mine at least in English) speakers. Like a TV or a projector screen.
Perhaps using mirrors (with some parts of the surface that are opaque) could do the trick?


My second question was about the normal resonances you have in a rectangular room (where you always have pressure nodes and velocity antinodes in the corners). A boundary coupled (just because it's places close to the walls) excite or trigger thes resonances more than free-standing speakers. At higher frequencys as you know the resonances have a modal density that is so high that single resonanses can't be heard. But at low frequencys (below 100 Hz or so) they can be quite problematic. As I see it there are (at least) three methods that can be used against these resonanses;
1. Using absorbtion (helmholtz, mass surface absorbers or large normal sound absorbers).
2. Using multiple or distributed sound sources. If you place a single sound source at in a place in the room where there is a pressure node for a certain frequency then it will trigger that corresponding resonance. But if you place a second correlated sound source at another pressure node for that frequency then that resonance can't be triggered. One way to use this technique is to use several subwoofers for the range of 0-100Hz and place them in a pattern that supresses these resonances.
3. Using a parametric eq.

Does my explanation makes my question more clear?

Reagards
Anders Eriksson

[Jacro]

Thanks James for your answer about directivety below 1kHz.
Now imagine that I manage to make loudspeakers that maintain directivety down to say 80Hz, and want to add some subwoofers to them:
If I place two subs in the front corners, they will be boundry coupled in their frequency range and have 90+90 degrees of dispersion which should be fine. But probably we play the subs in mono so now we have 180 degrees bass dispersion which is quite a bit more more than the theoretically perfect main speakers, how do we sort this out? And if we want one or two more subs to improve on room modes, how would all that effect frequency response vs power response in such a system?

paa,

My computer is in for repair for a few days, so I will have to wait to address this question more completely as soon as my computer is returned to me.

You ask a good question. Part of the answer is that it is most important to maintain the directivity down to the Schroeder frequency. Below the Schroeder frequency things change and to some degree, the sound source becomes the room modes themselves, as opposed to the loudspeaker, but that is a concept that needs some further explanation.

I hope to get back to you with more on this topic in a few days.

All the best,

- James

[Jacro]

(By the way, the post of Naqref’s stopped after only 2 pages… Was there further work and measurements reported on the experimental devices?)

Hello James,

Finally I had som time (and focus) to read the whole thread and I'm very impressed as many others.

Regarding My little sweetheart it's scrapped but I'll think I have some measurement of it somewhere in one of my computers. Perhaps I'll find it and then I'll post some information here (and in the original thread).

The first ambience-module with the principle was made around 1995. A couple of years later I built the one on the pictures. I was still a poor student at the time so there was actually no money to develop the principle then. And then I found other things more promising.

I found your patent:
http://www.freepatentsonline.com/7551062.pdf
Impressive. I'll think the right way to go ist to use multiple active units just as you say. Passive damping of the sound from the back of the unit is to test intensive to get right.

I am like you intrested in what I call wall integrated speakers. It should perhaps be altered to external wall integrated speakers (EWI) or something like that to avoid beeing misstaken for in-wall-speakers. But anyhow. I'm the designer of Larsen speakers (it's a follow up of Carlsson). Two major problems with ewi-speakers is the perception of depth and excitment of roomresonances I'll think. The problem with depth is that the mind has problems to project auditory sources beyond the visual boundaries. The eye's override of other senses is making it harder to imagine a sound stage beyond the front wall. Do you think this is a big problem and if so what else is there to do about it other than removal of visual clues or to give additional visual clues to aid the sound stage?

The second problem (actually it's really a problem with almost every type of speaker), what is your preffered method of dealing with that? Acoustical lf-absorbers, parametrical eq or perhaps using multiple sound sources to counter act the resonanses?

Send my regards to Lars Erickson.

Hi Naqref,

Good to hear from you.

You mention two issues related to “EWI” loudspeakers:

(Over here we differentiate from "In-Wall" by calling them either; "On-Wall" or "Boundary Coupled"):

1) Perception of depth

2) Excitement of Room Resonances

I'm not clear on what you are referring to with #2. I think you are referring to reflections off the near boundary, but that is not actually a "resonance" per say. So, to discuss number 2, it would be good if you can clarify for me.

So, in terms of depth perception, in one way it is a problem (difficulty in imagining depth through a wall) and in another way, it is an improvement, in that it doesn't create "false depth" due to reflections off of the wall, spaced some distance behind the loudspeaker.

Loudspeakers spaced away from the front wall tend to create some degree of a false sense of depth depending on the ratio of the direct energy vs. the intensity of the energy reflected back to the listener.

So, in this regard, the boundary coupled systems are more faithful to reproducing the actual depth captured on the recording. (This is an over simplification, due to there being many additional variables determining perceived depth).

Your other observation, of the visual wall blocking our ability to imagine depth behind the wall, can be a significant limitation with some listeners. It is not a universal problem, but certainly a problem for many. Initially, I found it to be problematic for myself.

Sometimes I listen with my eyes closed, when listening to reproduced sound through loudspeakers, and also when I am at a live concert. In both cases, I find that the visual aspects alter my perception of the aural event, so closing my eyes creates a more equivalent situation for comparing live vs. reproduced sound.

If I have my eyes open at a live concert, I perceive the individual instruments as being more focused in a specific location in space than if my eyes are closed. Also, with eyes open, I "visualize" more depth of field, commensurate with the actual stage depth. If I close my eyes at a live concert, then the location of instruments is more vague, less focused, and the perception of depth of field is less.

So, if we are attempting to transport the actual focus, and scale of the sound sources to our listening room, without the actual musicians being visually present, the boundary coupled system is actually a more “accurate” facsimile of the spatial dimension of the original event.

(In a smaller second venue, boundary coupled loudspeakers and loudspeakers out in the room both distort the image compared to the original, but the loudspeakers that are out in the room (unless highly directional) will add more artificiality to the reproduced sound.

To minimize the visual barrier (for some customers that find it bothersome), but still have a boundary-coupled relationship for the loudspeakers, I have developed an alternative use-model.

Let’s see if I can explain it without a drawing.

Instead of facing the front wall of the room, the listener faces into a corner of the room. As an example, the listener might sit 3-meter’s straight out from the corner, and the left loudspeaker is coupled to the wall that extends to the left of the corner, and the other loudspeaker is coupled to the wall that extends to the right of the corner. The distance between the two loudspeakers and the distance from the loudspeakers to the listener are the same as they normally would be, and the loudspeakers are angled, or aimed, in relation to the listener, as they normally would be.

Visually, the corner extends back behind the loudspeakers, but the loudspeakers are effectively boundary coupled to each wall as they normally would be on the front wall.

I usually put extra absorption/diffusion in the corner, such as a quarter round of a half-meter radius, covered with 100 mm of foam, and I use a larger piece of absorption on the wall where the loudspeaker is coupled.

With an angled configuration, such as the Larsen 4, 6, or 8, the right loudspeaker may work better on the left side, and visa versa.

Make sense?

If properly set up, this has a few advantages, two of which are; it provides a visual expansion of depth, and it can leverage a more effective boundary coupling (less than a half-space) which allows a small loudspeaker to better power a larger room.

I use this approach for listeners that find the wall boundary mounting to be problematic for imagining depth, and/or for listeners who have rooms that are larger than what the loudspeaker is normally capable of filling.

This is a rather unorganized, non-technical discussion of the issue that you raised, but hopefully it is still somewhat useful.

I’ll address the rest later.

Best regards,

- James

PS – I’ll be sure to say hello to Lars for you.

What you say makes a lot of sense. I hade tried to convince people of using the adjacent walls in a corner and shift right and left speaker in some cases but the look at me like they think; "Are you crazy or something" Perhaps I should work on my way of convincing people more to adress this problem.

Anyhow I think the problem is non existing when you have a picture of some sort between the boundary coupled (better word than mine at least in English) speakers. Like a TV or a projector screen.
Perhaps using mirrors (with some parts of the surface that are opaque) could do the trick?


My second question was about the normal resonances you have in a rectangular room (where you always have pressure nodes and velocity antinodes in the corners). A boundary coupled (just because it's places close to the walls) excite or trigger thes resonances more than free-standing speakers. At higher frequencys as you know the resonances have a modal density that is so high that single resonanses can't be heard. But at low frequencys (below 100 Hz or so) they can be quite problematic. As I see it there are (at least) three methods that can be used against these resonanses;
1. Using absorbtion (helmholtz, mass surface absorbers or large normal sound absorbers).
2. Using multiple or distributed sound sources. If you place a single sound source at in a place in the room where there is a pressure node for a certain frequency then it will trigger that corresponding resonance. But if you place a second correlated sound source at another pressure node for that frequency then that resonance can't be triggered. One way to use this technique is to use several subwoofers for the range of 0-100Hz and place them in a pattern that supresses these resonances.
3. Using a parametric eq.

Does my explanation makes my question more clear?

Reagards
Anders Eriksson

Anders,

As I mentioned in the last post, I am without my computer this week, so I will have to get back to you with a complete answer in a few days.

Yes, I do understand your question now... thank you for clarifying.

There are a number of issues around the topic that you have raised.

The solutions that you mention, or variations on them, tend to be some of the best we have available today.

To make a couple statements until I can get back with complete answers, I would say that I am generally an advocate of "Global", preemptive, solutions (such as modifying the room, loudspeaker configuration and location, and using multiple woofer systems), as opposed to "local", post-corrective solutions (such as digital room correction).

There are a number of ways to implement multiple woofers. I have a new couple of new ways that I have found that exhibit interesting results when compared to some of the current conventional approaches of Welti and Geddes. I'll speak more about them when I get back to complete my answers.

More later...

Best regards,

- James

[Jacro]

What is according to you the best length to width ratio?
Is that ratio different depending of speakers used?

Hi sportbilsentusiasten,

My computer is repaired and I am back online. I apologize for the delay.

I have many questions to respond to here.

Relative to your question, the most effective, angle and length-to-width ratio depends on system type and program material. While there are a numerous optimal angles, due to the wide variety of system configurations, I will just list three useful angles that tend to maximize the performance of many of the most common program sources and loudspeaker types.

1) Standard Stereo:
For a simple two-channel system with standard stereo program material, plus and minus 21 degree angles for the left and right loudspeakers tend to work best.

This is not so much due to recording type, which may vary quite a bit based on microphone usage. Instead, the angle is defined placement that provides the least amount of inter-aural crosstalk. This form of “passive” crosstalk cancellation partially mimics binaural or crosstalk cancellation systems. By doing so, it can recreate the perception of the most effective three-dimensional sound field, with “balanced” depth and spatial characteristics.

2) Three Channel:
For systems with a center channel loudspeaker, the purpose is to anchor the centermost information, with low coloration and spatial correctness in multiple listening seats (unfortunately it often introduces a problem of monophonic scale on center channel signals) but with this approach there are some spatial errors that force a different placement for the left and right loudspeaker if one wishes to maintain the full dimensionality of the sound field. One of the reasons is that a hard left signal is normally heard as placed at the left loudspeaker, in a two-channel arrangement. But, with a center channel system, the hard left signal will be placed between the center and left loudspeaker as a phantom image. If the left and right loudspeakers are kept at the same 21-degree angles, the stage width will be limited to about +/- 10.5 degrees.

So, with the center channel addition, we need to change the angular staging of the left and right channel loudspeakers. It turns out that an angle of approximately +/- 67.5-degrees is required to sustain the full stage width when using a center channel device. This wide of an optimal angle is due to a number of issues, including the fact the image spread now being limited by the phantom source width limit, and, also because the speakers can no longer be positioned at the 21-degree inter-aural cancellation angle and therefore they loose the ability enhance spaciousness due to a lack of optimal IAC.

3)Two channel program material that is Cross talk cancellation processed:
As a third angular specification, is that of optimizing for recordings that have been programmed with spatial enhancing crosstalk cancellation or are being post processed with a crosstalk cancellation system. For these recordings it is recommended that the loudspeakers place with angles that are less than plus/minus 15 degrees, and preferably on the order of plus/minus 5 degrees. This is implemented partially to minimize the variation of crosstalk cancellation and head related amplitude response vs. frequency.

What this means is that ideally, one moves their loudspeakers to have different angles relative to the listener depending on the nature of the program source.

As with most all of the issues that we are discussing, this discussion is not at all comprehensive due to both the wide range of program material, channel count and type, and also the myriad of other factors (frequency response, room interaction, etc.) that are altered as one changes the relationship of the two loudspeakers to each other and the listener.

****
To everyone:
I have lost track of which issues remain unanswered. If there are any questions that someone has submitted for me to answer that I have not responded to, or if there any that I have not answered completely of which there is still interest in further discussion, please let me know.

All the best,

- James

[paa]

I have lost track of which issues remain unanswered. If there are any questions that someone has submitted for me to answer that I have not responded to, or if there any that I have not answered completely of which there is still interest in further discussion, please let me know.?

I believe this is long lost:

...
OR,

2) Arrange for the first four boundary distances to be cross-relational such that they are distributed and/or cross-cancelling. In the second case, the system would ideally, have a smooth, but complex transitional power response that is calibrated to match the predetermined boundary interaction.

Both are rather complex to develop, and the second approach is the basis for the operation of The Clue.
...
- James

How can this be made to work for different room sizes, and different listening distances, things that also would demand different distances between the speakers?
Could this cross-relational and cross-cancelling function be simulated in some known software?

Then there was the discussion about power response with multiple subwoofers, where you said you had something more coming.

[Harryup]

To make a couple statements until I can get back with complete answers, I would say that I am generally an advocate of "Global", preemptive, solutions (such as modifying the room, loudspeaker configuration and location, and using multiple woofer systems), as opposed to "local", post-corrective solutions (such as digital room correction).

There are a number of ways to implement multiple woofers. I have a new couple of new ways that I have found that exhibit interesting results when compared to some of the current conventional approaches of Welti and Geddes. I'll speak more about them when I get back to complete my answers.

More later...

Best regards,

- James

Jacro,
I have followed the thread with great interest and you are really a good asset to have here.
I have made some test with multiple subwoofers and active digital crossovers (BSS FDS-366T) and I have found rather huge aural differences when using the time delay function. Do you recommend using multiple subbs without delay adjustments? From my small test with multiple subbs using Smaart I found that adjusting the delay for a rather straight frequency response will give a better transient response overall compared to rather straight frequency response using the peq. Are my observations correlated to using an active digital crossover which has an already built-in delay of at least 2ms? Could it be the case that just using an analog active crossover and using several subbs will "cure" all timing problems?
Unfortunately I have not the possibility to test multiple subbs with analog crossover except at home where I just have 2 Paradigm Servo-15 directly placed under the midrange and treble speakers. There I have compared an analogue JBL553 with the BSS and if I sit just a bit off center I could never prefer the analogue time uncompensated filter. In the middle of the sweet spot the difference is less but a dentist chair is preferable.
What is the risk of getting a rather straight frequency response using multiple subbs and losing some transient response due to have the subbs playing a bit in and out of phase with each other and the rest of the system?
I have looked into jevans111 small movies at youtube
http://www.youtube.com/results?search_query=subwoofer+alignment&aq=f about subwoofer alignment and I'm curious how to find out when the setup is fully optimized. What tool would you recommend or is it even a correct approach for home environments.

Regards
Harryup

[avr7000]

Hello Jacro

Is there a possibly somewhere for me to listen to, and maybe buy, the clue speakers in either:
New York
Grand Rapids
Richmond
-and maybe Chicago.

Regards
Stefan

[Jacro]

Hello Jacro

Is there a possibly somewhere for me to listen to, and maybe buy, the clue speakers in either:
New York
Grand Rapids
Richmond
-and maybe Chicago.

Regards
Stefan

Hi Stefan,

We have distribution that is ready to be set up in New York as soon as we are fully able to supply our other distributors, hopefully in the next two to three months.

We have nothing available in Grand Rapids or Richmond at this time.

Currently, we DO have an outlet in Chicago. The contact is Paul Lawless ( paul@sjofnhifi.com )

I hope this can work for you. If not, let me know and I'll alert you when we activate our distribution in New York.

Best regards,

- James

[Jacro]

I have lost track of which issues remain unanswered. If there are any questions that someone has submitted for me to answer that I have not responded to, or if there any that I have not answered completely of which there is still interest in further discussion, please let me know.?

I believe this is long lost:

...
OR,

2) Arrange for the first four boundary distances to be cross-relational such that they are distributed and/or cross-cancelling. In the second case, the system would ideally, have a smooth, but complex transitional power response that is calibrated to match the predetermined boundary interaction.

Both are rather complex to develop, and the second approach is the basis for the operation of The Clue.
...
- James

How can this be made to work for different room sizes, and different listening distances, things that also would demand different distances between the speakers?
Could this cross-relational and cross-cancelling function be simulated in some known software?

Then there was the discussion about power response with multiple subwoofers, where you said you had something more coming.

paa,

I can’t go into all the subtleties and complexity of applying the cross-relational techniques without applying at least ten pages of writing, but I’ll state a few useful starting points to allow one to engage the general concept and get somewhat of a feel for variation relating to room dimension differences.

If the loudspeaker low frequency system is calibrated to the appropriate alignment with the correct high pass slope, and the placement of transducers and cabinet dimensions are properly integrated with the total scheme, then the desirable impact of the upper-most sensitive boundary frequencies (less than 0.75 wL) will be dominated from correct distances being established to the floor and front wall. The secondary relationships will be the sidewalls, ceiling and mutual coupling interactions. This relationship has implications for both the low frequency extension, AND, the smoothness of the response due to direct and the multiple reflected sounds being coordinated for diverse, distributed relationships. Without resorting to the optimal complexity, one approach that can work well is to apply “golden rule” relationships (1.618 x) between each of the specular sources.

The secondary sources have the greatest impacts below the Schroeder frequency, in the modal region. While the optimization of these dimensions are important (and may be interchangeable or reordered, if one of them must change outside an optimum range) have a certain amount of inherent flexibility due to modal dominance.

Yes, the optimization can be simulated and arrived at by way of software, and is currently the subject of a patent application.

There is much more to these relationships, but I would add that a two-loudspeaker system, in most cases, will not provide smooth response in the modal region. The application of multiple diversity woofers will most often be required to optimize the room/system response.

These are complex subjects, the room relationship of the loudspeaker, and also the application of multiple subwoofers.

There are a number of complex, simulation and measurement-based calibration of multiple subwoofers, but a few years ago I developed a simplified system for my customers to effectively apply multiple subwoofers to achieve excellent results with minimal analysis.

It is based on the use of a subwoofer on each of the four walls, with the relationship of each woofer being that of an increasing 1.618 relationship of each woofer to a corner. The first woofer is placed at a distance from a first corner that is based on a specific value relating to a constant relative to room dimensions. The second woofer is placed 1.618 times that distance from the second corner. Third woofer placed, relative to the third corner, 1.618 times the distance of the second woofer, and the fourth, 1.618 of the third. The phase of woofers 2, 3, and 4 are tested individually with reversed phase (usually woofer 2 or 3 work best with phase reversed)
This simple approach produces surprisingly consistent results.

Okay, back to work… I will try to write more later on system power response issues above and below the Schroeder frequency.

Cheers,

- James

[paa]

Hi James,
Once again, thanks for your patience with all our questions.
However much I await more about these things I have a follow up question here.

...
It is based on the use of a subwoofer on each of the four walls, with the relationship of each woofer being that of an increasing 1.618 relationship of each woofer to a corner. The first woofer is placed at a distance from a first corner that is based on a specific value relating to a constant relative to room dimensions. The second woofer is placed 1.618 times that distance from the second corner. Third woofer placed, relative to the third corner, 1.618 times the distance of the second woofer, and the fourth, 1.618 of the third. The phase of woofers 2, 3, and 4 are tested individually with reversed phase (usually woofer 2 or 3 work best with phase reversed)
This simple approach produces surprisingly consistent results.
...
Cheers,
- James

Are there no problems when every subwoofer stand at different distances from the listener? How much difference can be tolerated?

[Kraniet]

Toole also argues against* placing bass around the room. One in every corner och one in the middle of every wall are some of the tips. Arguing that more bass speaker are better but above four is hitting diminish return.

Ingvar Öhman argues that the bass should preferably come from the front. That this would bring whats a more "pshycoacoustical correct wavefront".

Any comment in regard to this?

edit: sorry meant to say against not "for"

[Jacro]

Toole also argues against* placing bass around the room. One in every corner och one in the middle of every wall are some of the tips. Arguing that more bass speaker are better but above four is hitting diminish return.

Ingvar Öhman argues that the bass should preferably come from the front. That this would bring whats a more "pshycoacoustical correct wavefront".

Any comment in regard to this?

edit: sorry meant to say against not "for"

Hi Kraniet,

I believe that these two viewpoints, of Toole and Öhman, are both valid, each within a limited, optimized context.

The importance of maintaining a frontal derived wavefront is dependent on a number of parameters. One parameter set is the low pass corner frequency and/or low pass slope.

It works best if the low pass corner frequency is reduced and/or the low pass slope is increased, progressively for each woofer as the placement of the woofers is moved from front to beside the listener and then behind the listener. Also, the woofers are progressively reduced in level as they are place farther from the frontal launch region.

Each additional woofer is utilized mainly to provide a sequentially smaller and smaller correction to the modal reform. If this is done properly, there will be no audible difference in the perception of the appropriate psycho-acoustical wavefront direction.

All the parameters must be included before determining the ideal low pass frequencies, but in one example of an optimized room, the frontal woofers were crossed in at 122 Hz, the side woofers were crossed at 83 Hz and the rear woofers were crossed at 61 Hz. There were some additional optimized attributes but the system was carefully tested for tonal and spatial distortions relative to the addition of the side and rear woofers and was found to be transparent in this regard. Of course, the room parameters must support this type of low frequency architecture.

My more universal preference is to utilize a multiple woofer set that has all the individual woofers placed in the frontal hemisphere forward of the listening position, more inline with Mr. Öhman’s preference, but with a few additional enhancements.

My goal with the front hemisphere woofer arrangement, is to still address the most significant modal characteristics of the room, and this can be accomplished in the frontal portion of the room, but one must think three dimensionally.

This requires placing woofers in an arrangement that is not only horizontal, x-axis, but also to diversify in the y and z-axis, with at least one woofer above the half-height point of the room. With this approach, one can substantially accomplish the modal control of woofers placed on each of the four walls, while still maintaining 100% frontal wavelaunch.

In all of these approaches, one or more of the woofer units may contribute most effectively when operated with altered or reversed phase.

If one uses substantially full range (deep bass) left and right channel loudspeakers, then they can be included as one or two of the multi-sub elements.

I prefer to step down the gain of the left and right loudspeakers by 3 to 6 dB below the approximate frequency of the highest subwoofer crossover, and operate in a power-sharing manner below this frequency, both smoothing room modes AND increasing system maximum output capability.

Hopefully this provides at least a partial answer to your question.

Best regards,

- James

[paa]

Thanks James for your answer about directivety below 1kHz.
Now imagine that I manage to make loudspeakers that maintain directivety down to say 80Hz, and want to add some subwoofers to them:
If I place two subs in the front corners, they will be boundry coupled in their frequency range and have 90+90 degrees of dispersion which should be fine. But probably we play the subs in mono so now we have 180 degrees bass dispersion which is quite a bit more more than the theoretically perfect main speakers, how do we sort this out? And if we want one or two more subs to improve on room modes, how would all that effect frequency response vs power response in such a system?

paa,

My computer is in for repair for a few days, so I will have to wait to address this question more completely as soon as my computer is returned to me.

You ask a good question. Part of the answer is that it is most important to maintain the directivity down to the Schroeder frequency. Below the Schroeder frequency things change and to some degree, the sound source becomes the room modes themselves, as opposed to the loudspeaker, but that is a concept that needs some further explanation.

I hope to get back to you with more on this topic in a few days.

All the best,

- James

Hello James,
Do you have any more information about the Schroeder frequency, and its importance in small rooms?

[Jacro]

Hi James,
Once again, thanks for your patience with all our questions.
However much I await more about these things I have a follow up question here.

Are there no problems when every subwoofer stand at different distances from the listener? How much difference can be tolerated?

Hello James,
Do you have any more information about the Schroeder frequency, and its importance in small rooms?

Hi paa,

The nature of sound sources when operating in the modal frequency region seems to be an issue of great confusion. It is fundamentally different than how systems operate, and what parameters are important, above the Schroeder frequency.

In the modal range, it may be more appropriate to think of the dominant source of the sound reaching your ears as being that of the modal output of the room, not the loudspeaker.

In most cases, the woofer drives the room and the modal nature of the room is what delivers the sound to your ears.

Our hearing requires a couple of cycles before it can determine a sound's timbre and when operating below the Schröeder frequency, reflections become an integral part of the sound almost instantly because the wavelengths are comparable to, or greater than, the room's dimensions.

So what we hear is no longer dominated by the loudspeaker but instead is created by the room boundaries and the sense of a perceived transient response is dominated by the amplitude smoothness and harmonics of the bass fundamentals.

There are at least two arrangements that can substantially override this characteristic of the modal frequency range;

1) a woofer system that maintains high directivity at modal frequencies (a rarity, but we have built, tested, and characterized these types of systems to determine their effectiveness)

2) using a near field woofer close to the listening seat, such that the dominant arrival is the direct arrival from the woofer itself (this is a superior technique not commonly used in the industry as of yet. I believe this approach is ahead of its time and will become a standard in another 5 to 10 years. We have developed systems using this approach that consistently outperform all other approaches)

In terms of differential distances between different woofers and the listener, this is not an audible issue from a phase/transient standpoint. If there are phase/timing errors, it is the amplitude response change due to phase differentials that is audible, not the phase difference itself. In fact, when distributing multiple subwoofers around the room, one of the useful techniques to create maximum smoothness in the frequency response, is to alter the phase of at least one of the woofers to adjust the modal development and amplitude summation among the woofers as they drive the room modes.

When setting up distributed woofer systems I tend to use lower frequency low-pass filters for woofers placed behind and/or farther from the listener than the frontal woofers, attempting to maintain less than a quarter wavelength differential between all woofers when measuring their distances to the listener.

As usual, this is a very limited explanation relative to all of the acoustic and psycho-acoustics effects in play, but hopefully this at least gives you some useful information on the topic.

Cheers,

- James

[Nattlorden]

My goal with the front hemisphere woofer arrangement, is to still address the most significant modal characteristics of the room, and this can be accomplished in the frontal portion of the room, but one must think three dimensionally.

This requires placing woofers in an arrangement that is not only horizontal, x-axis, but also to diversify in the y and z-axis, with at least one woofer above the half-height point of the room. With this approach, one can substantially accomplish the modal control of woofers placed on each of the four walls, while still maintaining 100% frontal wavelaunch.

I'm stacking my woofers in the front coners - floor to ceiling. Does this conform, or would you also say I'd need some inbetween?

And what's your view about stereo vs. mono for (sub)woofers, say crossed at 80Hz?

And also something completely different - as I take it you've been around a lot over the years - have you crossed paths with Albert Von Schweikert and if so, have you any stories or such to tell about it?

[Jacro]

My goal with the front hemisphere woofer arrangement, is to still address the most significant modal characteristics of the room, and this can be accomplished in the frontal portion of the room, but one must think three dimensionally.

This requires placing woofers in an arrangement that is not only horizontal, x-axis, but also to diversify in the y and z-axis, with at least one woofer above the half-height point of the room. With this approach, one can substantially accomplish the modal control of woofers placed on each of the four walls, while still maintaining 100% frontal wavelaunch.

I'm stacking my woofers in the front coners - floor to ceiling. Does this conform, or would you also say I'd need some inbetween?

And what's your view about stereo vs. mono for (sub)woofers, say crossed at 80Hz?

And also something completely different - as I take it you've been around a lot over the years - have you crossed paths with Albert Von Schweikert and if so, have you any stories or such to tell about it?

Hello Night Lord,

Your arrangement of creating a line source of woofers in the front corners is an excellent approach that can work very well in many types of rooms. It is one of the best at providing both smooth response, due to a diversity line wave room interface, AND, great large signal capability due to corner re-enforcement.

It is unlikely, but if for some reason the shape, dimensions or rigidity of your room still results in non-uniform amplitude response, you may try adding one more woofer, placed midway on the front wall or midway on a sidewall, to further smooth the response, but in most rooms this would not be necessary with your configuration.

As an alternative, the Welti approach of placing 4 woofers on the floor, centered on each of the walls, can be transposed to the front wall, with a woofer centered on the floor, the sidewalls, and the ceiling, all pressed against the front wall.

In terms of maintaining stereo signals in the woofer systems operating below 80 Hz, it depends on room size, modal conditions, and program characteristics.

It is rare that recordings will contain audible separation below 80 Hz but one can generate “artificial” signals that may exhibit slight audibility under controlled conditions in very large rooms, but this does not translate to meaningful differences in the modal range in standard living rooms/studios because the modal nature of the room is to cause the reflected signals to act as a left/right channel-summing network.

My preferred system is to use full range main left and right channel loudspeakers (down to at least 50 Hz, preferably down to below 40 Hz) and have them maintain the separate, stereo, signals, while further incorporating a monophonic, multiple woofer system architecture as a shared bass system below 100 Hz.

Relative to Albert Von Schweikert, I only have a vague memory of him calling a number of years ago to ask me for advise about some patent issues and questions about planar magnetic/ribbon transducers. I have never met him, so I apologize for not having something more intriguing to convey.

All the best,

- James

[paa]

James, how do you manage to have all these different crossover frequencies and shared frequency ranges to sum at something close to phase coherence?

[Jacro]

James, how do you manage to have all these different crossover frequencies and shared frequency ranges to sum at something close to phase coherence?

paa,

I believe I already addressed this in my previous answers, but maybe I wasn’t clear enough.

To summarize, first arrival phase is relevant to the extent it has summation effects on amplitude (either amplitude interaction with the main L/R loudspeaker and/or amplitude summation of the multiple woofers and modal interactions), but (assuming standard room size and crossover frequency well below the Schröeder frequency) the direct audibility of the phase of the direct sound from the woofer is not significant.

As I suggested, our hearing requires a couple of cycles before it can determine a sound's timbre and when operating below the Schröeder frequency and reflections are nearly instantaneous in becoming an integral part of the sound because the wavelengths are comparable to, or greater than, the room's dimensions.

It is difficult to understand intuitively, but the source of audible low frequency arrivals is dominated by the room itself, not the loudspeaker.

A useful analogy may be that of a bass reflex loudspeaker operating at the tuning frequency; the woofer has a direct output to the listener, but the sound heard by the listener is dominated by vent output due to the Helmholtz resonance transferring energy from the rear-side of the woofer diaphragm coupling through the resonant mode of the enclosure. The phase of these two sources is not ideal from the standpoint of the first arrival from the front side of the woofer diaphragm, but the vent dominates in a manner that overrides the direct sound of the woofer, making the phase relationship insignificant.

Similarly, below the Schröeder frequency the woofer output drives the modes of the room and the modal output of the room is the dominant arrival perceived by the ear-brain.

Again, the phase plays a role in this case, but the significance of the phase is its affect on amplitude. If you hear any anomaly related to the phase relationships not being ideal, in the normal sense, that anomaly is an amplitude error caused by the phase relationship, not the audibility of the phase error itself.

The above is not a perfect analogy, but possibly it provides a useful example of another system that has a dominant source that has an ‘indirect’ relationship to the woofer output reaching the listener.


All of this is why I started my first answer with this statement:
“The nature of sound sources when operating in the modal frequency region seems to be an issue of great confusion. It is fundamentally different than how systems operate, and what parameters are important, above the Schroeder frequency.”

The rules we are used to applying above the Schröeder frequency are not necessarily valid below the Schröeder frequency. It takes a lot of effort to sort these things out, partially because the belief systems we have established from rules that we apply to certain aspects of audio and therefore assume that we can intuitively transfer those rules to all other aspects of the system.

Also, as I stated previously, if one ‘decouples’ the woofers from the modal conditions of the room, such as using high directivity or near-field woofers, then the rules that apply above the Schroeder frequency can still have validity, but these two approaches are not usually applied.

Any clearer?

- James

[MagnusÖstberg]

Hi James!

Such low crossover frequency as 40-50hz puts large demands on the main to be able to move substantial amount of air.

Don´t you see the advantage of having the larger woofers operate up to 80-100hz as that definetly makes things easier for the mains who can concentrate on upper bass and above? Why put that extra stress on them?

[Jacro]

Hi James!

Such low crossover frequency as 40-50hz puts large demands on the main to be able to move substantial amount of air.

Don´t you see the advantage of having the larger woofers operate up to 80-100hz as that definetly makes things easier for the mains who can concentrate on upper bass and above? Why put that extra stress on them?

Hi MagnusÖstberg,

I agree with you.

What I had stated it previously, but I forgot to mention again in the last post, is that I always "shelf down" my mains by approximately 6 to 12 dB to reduce the load on the main loudspeakers, but keep them active at that reduced level from 80 or 100 Hz down to their lower limit, to provide a more effective integration with the subwoofers and to create a greater number of diverse sources of bass in the room for further improvements in smoothness.

The greater the number of subwoofers that are added, the greater the amount of LF amplitude reduction can be applied to the mains, reducing their output requirements.

Make sense?

- James

[paa]

I womder if this concept is similar to the products called "basstöd" by ino?

[Nattlorden]

James, thank you very much your reply.

[MagnusÖstberg]

I womder if this concept is similar to the products called "basstöd" by ino?

Not exactly as "Basstöd" is limmited to one woofer per channel, passively connected to the mains.

[Jacro]

I wonder if this concept is similar to the products called "basstöd" by ino?

paa,

I am not familiar with "basstöd".

Can you explain the concept?

Thank you,

- James