The Goals for an "Ideal Loudspeaker"

[kkantor]

I think it was a clever (grounded in good science) attempt to break the "directional for focus, dispersive for spaciousness" compromise mentioned above. More directional than anything else we've talked about, with stacked arrays in foam lined recesses, yet it put more lateral sound into the room- but only after appropriate time delay.. It should have fooled the ear into thinking you were in a larger room since there would be strong lateral reflections but at a time delay apropriate to a space with wider dimensions.

But I've never heard a pair. Hey Ken, what did they sound like?? Did it work? Did you get focus + spaciousness?

David

Ah, they were terrible.... Somewhere, I have a folder of reviews from the time, but have looked at length and cannot yet find it. (Probably burned it in a fit of guilt and despair.) I do remember being surprised that TAS gave them a long and very positive write up. Meanwhile, here's something that ought to appeal to Z's sensibilities...

http://www.hometoys.com/news_detail.php?id=18213964

This (inexpensive) design employs a relatively complicated crossover on the tweeter array for pattern control. It's a bit like the "Bessel Array" that Philips tried, but with the advantage of being able to actively EQ the overall response to allow more degrees of freedom in the passive sections.

-k

[Zilch]

Meanwhile, here's something that ought to appeal to Z's sensibilities....

We'll talk.

I've been intrigued by Keele's CBT, but don't understand it yet.

I THINK this is an implementation steered via passive group delay:

http://www.jblproservice.com/pdf/CBT/CBT50LA-WH.pdf

21" x 4" x 6". Outrageous!

Now remembering there was an NHT product with a wide/narrow switch, no?

[i may have something new with defined directivity to show you soon. ]

[speaker dave]

This (inexpensive) design employs a relatively complicated crossover on the tweeter array for pattern control. It's a bit like the "Bessel Array" that Philips tried, but with the advantage of being able to actively EQ the overall response to allow more degrees of freedom in the passive sections.

-k

Ah, line arrays, here is a good general paper on the subject. (Page 7 for Bessel arrays.)

Ken, you don't get off that easy. What did the Magic speakers sound like?

David

Line_Arrays.pdf

[Carlspeak]

I am still of the opinion that a speaker's coloration is due in some measure to what the drivers are made of - all other things being equal (if that's even possible)

Art Dudley's May Stereophile piece on his re-habbing of an Advent speaker mentions a quote from one of his favorite writers, Herb Reichert:

"Most things really do sound like whatever it is they're made of"

I guess I'm not alone in this opinion.

[speaker dave]

Was there any substantial change in what JBL drivers were made of between the 50's and 70's? I heard a lot of classical music played through some tube-powered Hartsfields when I was growing up, and the sound didn't seem terribly 70's "westcoast-y" at all.

I've never liked the West Coast/East Coast generalizations. It is true that Altec and JBL had a cinema heritage and that steered them into more horn loaded designs. Beyond that all generalizations are pretty loose. What about University and EV? Were they West coast? (New York and Michigan) The big change from the 50's generation of products to the 60's was the realization that achieving flatter response by giving up considerable efficiency was the right way to go. Certainly the AR-1 was the leader here.

I grew up in a house with a JBL D130 woofer in a 10 cubic foot box plus a University horn tweeter on top. Later measurements of the woofer showed a Qt (from memory) of about .25. Lots of magnet and a very light cone meant that that woofer couldn't have bass within 10dB of the midrange level. Maybe in earlier days with amplifiers with a high source impedance it might have worked a little better. In the absence of affordable calibrated measuring systems a lot of early speakers were designed to a simple philosphy of the virtue of large magnets and high efficiency. People thought in ideal driver terms rather than system terms.

With the AR-1 Vilchur piled on the mass until the 2pi response was flat. (This was the essence of his invention, rather than high compliance and letting the cabinet set the resonance. That had been done before.) Woofer and cabinet were designed together to achieve a performance specification.

Beyond the high midband efficiency of earlier products it is hard to generalize what a "West Coast" sound is. In truth most early speakers were bad performers. Their response varied all over the place. East Coast speakers varied as well. An AR3 and a KLH 6 come from pretty different schools of thought.

I think that by the late 70's companies on both coasts were designing to fairly similar philosophies. Everyone had adequate measuring systems and either anechoic chambers or outdoor measuring areas. This greatly reduced the coastal differences.

David

[speaker dave]

"Most things really do sound like whatever it is they're made of"

I guess I'm not alone in this opinion.

As a rash young engineer at JBL I made that statement about the old 075 tweeter. They had a very distnictive (generally bad) sound and I stated that it was inherent in the aluminum diaphragm, that even if they could be made to measure better the sound character would remain.

Sombody called my bluff and I had to set up a comparison between two 3-way systems, identical except one had the 075 tweeter grafted on. We used a new 1/3rd Octave equalizer to try and get the response of the ring radiator to match the smoother performing 1" dome.

The response of the typical 075 was primarily composed of 3 individual peaks. That was the good ones that the lady had creased correctly with the tip of her ball point pen (honestly!). Still, with an equalizer I could get the response to be a pretty good match to the 1" dome. Surprisingly, as the response got closer to that of the dome, the sound became the same as well. The "aluminum" character disappeared.

Material, of coarse, determines the response of a driver, but if you can fix the response that material related character will disappear.

David

[Zilch]

Art Dudley's May Stereophile piece on his re-habbing of an Advent speaker mentions a quote from one of his favorite writers, Herb Reichert:

"Most things really do sound like whatever it is they're made of"

I guess I'm not alone in this opinion.

I'm hoping Ken will weigh in on this one; he's our transducer guy.

[You forgot hemp, tho.... ]

[Carlspeak]

I'm hoping Ken will weight in on this one; he's our transducer guy.

[You forgot hemp, tho.... ]

Yeah MAN! hemp...Hmmmmmmmmm

Tone Tubby woofers are pretty popular among the axe grinders....

[Zilch]

Yeah MAN! hemp...Hmmmmmmmmm

Tone Tubby woofers are pretty popular among the axe grinders....

It's likely our Ken is well familiar with the "parameters...."

[genek]

Material, of coarse, determines the response of a driver, but if you can fix the response that material related character will disappear.

I have previously offered my non-professional opinion that the sound of older speakers is the result of someone's deliberate voicing choices rather than technological limits and that it should be possible for an expert speaker designer/builder (which would be someone other than me) to construct a speaker based on AR-type comonents that sounds "west coast" or a JBL-type that sounds "east coast." So far, nobody has taken this up, at least not deliberately (IMO, a lot of people who have passed through CSP looking for ways to increase the HF output of their ARs were trying to turn them toward the west). Anybody?

[Pete B]

I have previously offered my non-professional opinion that it should be possible for an expert speaker designer/builder (which would be someone other than me) to construct a system with similar response, dispersion, etc., to an AR-3a and sounds just like one using completely different components (for example, horn or waveguide instead of a dome tweeter). So far, nobody has taken this up. Anybody?

I am willing to do this given the time, a cabinet to work with and a good working sample AR-3a

but what would be in it for me? I'm sure those who have shown their hatred for my honest

comments here would never praise my design - thus the test would have to be blind. But still

what is in it for me? Seems I recall you offering a fee, that would be better, but you had

unrealistic expectations from my perspective, about spare parts, maintenance and so on.

[Zilch]

Carl has done some preliminary work with this, and it sounds like he'll be doing more.

Same with Pete, maybe.

I believe Dave has affirmed that replicating AR3a would be trivial.

I think LST and the divergent-baffle Allisons are the more interesting ones, as they better achieved the fundamental design thesis.

See the pic at #37, above.

But still

what is in it for me?

It doesn't have to be about the money, but much of it is around these parts.

OTOH, Toole suggests that the others may well be marketable for a particular application, 16.4.3....

[Pete B]

Carl has done some preliminary work with this, and it sounds like he'll be doing more.

Same with Pete, maybe.

My plan would be to use all new and modern drivers, and to have a switch for Classic

sound, as close as possible to the 3a, and "Revoiced" to compete with modern designs.

I would use a low diffraction cabinet in order to get the best possible sound - perhaps

something like the AR-91 cabinet.

I have done a quick prototype of the "Revoiced" version on an AR-11 using the original

drivers - just to prove to myself that there was room for improvement. This is probably

what you were thinking of, but I didn't use an AR-3a.

[Pete B]

It doesn't have to be about the money, but much of it is around these parts.

OTOH, Toole suggests that the others may well be marketable for a particular application....

I already know that I can clone the sound of most speakers, I have done it several times. I

don't have to prove it to myself. It would be easy for anyone to dismiss any effort that I engage

in so what is left besides money?

OTOH, Toole suggests that the others may well be marketable for a particular application....

I do not follow what you are referring to here ...

This should probably be moved to a cloning thread or whatever you want to call it since we are now off topic.

[Zilch]

I added the citation above.

It's on topic, actually, as the ideal loudspeaker might well incorporate switchable directivity and/or voicing. Horizontal vs. vertical deployment and boundary proximity are also significant "issues."

[The JBL CBT has a "wide/narrow" switch, and others handle orientation with steering options.... ]

[genek]

Seems I recall you offering a fee, that would be better, but you had

unrealistic expectations from my perspective, about spare parts, maintenance and so on.

IIRC, that was when we were discussing what I might be willing to pay if someone offered a new speaker as a product for sale. Not the same thing as building one for "scientific reasons." If we're talking about a community research exercise, I could probably kick in a hundred or so to help defray the cost of parts and materials, the cost of a leg of a round-robin shipment for trials and maybe some food and booze for a SoCal listening party. What would be in it for whoever builds the thing would be bragging rights, nothing more.

[speaker dave]

I have previously offered my non-professional opinion that the sound of older speakers is the result of someone's deliberate voicing choices rather than technological limits and that it should be possible for an expert speaker designer/builder (which would be someone other than me) to construct a speaker based on AR-type comonents that sounds "west coast" or a JBL-type that sounds "east coast."

I don't see it that way. Every 60's or earlier speaker that I have measured or seriously listened to has been highly colored, always in random ways. Who intentionally designs a system to sound bloated, honky, shrill, forward, recessed, or whatever. Many early drivers are very non-flat. Crossover design in the early days was simplistic, often relying on fixed filter values published in theory books rather than dealing with the actual driver response. There was little understandng of how the driver and cabinet interacted.

Why are we on this tangent about cloning East Coast or West Coast sound? (Rather than the ideal speaker) First you have to define what those terms mean, a task I don't think can be done in any justifiable way.

Good design is about voicing a system to sound neutral and serve all forms of music, not to mimic another system from another era or indefinable ideas about East Coast and West Coast.

David

[genek]

I don't see it that way. Every 60's or earlier speaker that I have measured or seriously listened to has been highly colored, always in random ways. Who intentionally designs a system to sound bloated, honky, shrill, forward, recessed, or whatever. Many early drivers are very non-flat. Crossover design in the early days was simplistic, often relying on fixed filter values published in theory books rather than dealing with the actual driver response. There was little understandng of how the driver and cabinet interacted.

Why are we on this tangent about cloning East Coast or West Coast sound? (Rather than the ideal speaker) First you have to define what those terms mean, a task I don't think can be done in any justifiable way.

Good design is about voicing a system to sound neutral and serve all forms of music, not to mimic another system from another era or indefinable ideas about East Coast and West Coast.

Possibly a conceptual disconnect between people who are trying to talk about the possible development of a new "ideal speaker" goal and others who inhabit a site devoted to classic New England speakers because they think their "ideal speakers" for all forms of music were already designed 40 years ago.

Maybe this discussion would be better served on DIY Audio, Audioforums or some other site where it is not based on the assumption that the design philosophy that a large number of the site's readers are still fans of is invalid.

[kkantor]

We'll talk.

I've been intrigued by Keele's CBT, but don't understand it yet.

I THINK this is an implementation steered via passive group delay:

http://www.jblproservice.com/pdf/CBT/CBT50LA-WH.pdf

21" x 4" x 6". Outrageous!

Now remembering there was an NHT product with a wide/narrow switch, no?

[i may have something new with defined directivity to show you soon. ]

The NHT model, (VT-2, maybe others), had a switch intended to alter the stereo image, by changing the overlap of the midrange output between drivers, and thus altering the impulse response. The idea was to minimize localization image-blur for audio, but increase it somewhat for video program use. I didn't really think about this in terms of directivity, but I suppose it had an effect.

-k

[kkantor]

I don't see it that way. Every 60's or earlier speaker that I have measured or seriously listened to has been highly colored, always in random ways. Who intentionally designs a system to sound bloated, honky, shrill, forward, recessed, or whatever. Many early drivers are very non-flat. Crossover design in the early days was simplistic, often relying on fixed filter values published in theory books rather than dealing with the actual driver response. There was little understandng of how the driver and cabinet interacted.

Why are we on this tangent about cloning East Coast or West Coast sound? (Rather than the ideal speaker) First you have to define what those terms mean, a task I don't think can be done in any justifiable way.

Good design is about voicing a system to sound neutral and serve all forms of music, not to mimic another system from another era or indefinable ideas about East Coast and West Coast.

David

I certainly agree with you that the terms "East Coast Sound" and "West Coast Sound" are meaningless and over-used. Designers on both coasts, and points in between, have been laughing about this alledged dichotomy for as long as I have been in the biz.

I don't agree that "good design" is always about neutrality; it can also be about engagement. In fact, good design can chase many goals, including a designer's interest in interpreting and improving the definitions and approaches of others. I have tried to do that with some designs, like the 303; isn't that approach at least part of what you were after at McIntosh?

-k

[kkantor]

I'm hoping Ken will weigh in on this one; he's our transducer guy.

[You forgot hemp, tho.... ]

Nah, I'm more of a system guy.

But, I don't at all think that materials have characteristic, inherent sounds.

-k

[kkantor]

Ah, line arrays, here is a good general paper on the subject. (Page 7 for Bessel arrays.)

Ken, you don't get off that easy. What did the Magic speakers sound like?

David

Thanks. I >certainly< should have remembered to mention your paper, since I consulted it frequently in the design of that multimedia system.

I feel funny about discussing the sound of the MGC-1, since that brings me into the dual danger zones of reviewing an AR product and, worse, reviewing one of my own products. I would prefer to confine my comments to the research and experimentation which preceded the MGC-1. About that, I will say, yes, it is possible to simultaneously achieve stable and precise localization, along with spaciousness. But, not from any one single-drive speaker.

I will scan some pages from my 70's-era thesis on the topic. I do know where that is!

-k

BTW- in the Audio mag article on the Magic's, if I remember correctly, there was a picture of the 25-element array of 4.5" speakers that I used to play with midrange directivity. I definitely gained an appreciation for pro-sound guys trying to get that to work...)

[Zilch]

Maybe this discussion would be better served on DIY Audio, Audioforums or some other site where it is not based on the assumption that the design philosophy that a large number of the site's readers are still fans of is invalid.

Irrelevant, largely. Folks like 'em; the question is why, from a design perspective?

Are there elements of the "Ideal Loudspeaker" in these classics? If so, there's no better place to discover them, and we've already identified a major one -- constant directivity. They didn't call it that, and its reign was brief, but whether the purpose is maintenance, update, upgrade, clone, or modernization, the knowledge is essential.

So what if the fundamental design thesis was erroneous? What matters is the result, and we are in a far better position to analyze that retrospectively than they were, obviously.

[individual interests extend well beyond resale here.... ]

[genek]

Irrelevant, largely. Folks like 'em; the question is why, from a design perspective?

Are there elements of the "Ideal Loudspeaker" in these classics? If so, there's no better place to discover them, and we've already identified a major one -- constant directivity. They didn't call it that, and its reign was brief, but whether the purpose is maintenance, update, upgrade, clone, or modernize, the knowledge is essential.

Individual interests extend well beyond resale here....

It is relevant as a reminder to those engaged in a discussion of "idealness" in speakers that there are many different possible criteria for it, and that they are carrying this discussion on in a place where many readers' criteria for an "ideal" or even "good" speaker is one capable of sounding like speakers they already prefer and not what experts think they ought to prefer. In a more neutral setting it might not be necessary to explain how the various proposals are going to appeal to fans of certain old speaker models, but here there will likely always be someone nagging you all about it.

[Zilch]

I'm seeing nothing but interested contributors in this thread, actually, and 8.8 views/post.

[Average in The Kitchen is 84,184/2729 = 30.85.... ]