A Broad Discussion of Speakers with Major Audio Luminaries

[bmc0]

Any suggestions?

Nope . I agree that it's not that useful for loudspeakers; just wondering if I was missing something obvious regarding its measurement.

 

[Keith_W]

Dr. Toole, a speaker designer told me that these days, flat frequency response and directivity are solved problems given the availability and sophistication of modern modelling software and DSP. One only needs to be a competent designer and know how to use the software and you can create "any spinorama you want" (his words, not mine). About the only unsolved problems are driver resonances, power handling, and distortion - although all are improving. In your view, is he correct? And what do you think are the remaining unsolved problems in loudspeaker design?

 

[NTK]

One of the few quotes I remember from my grad acoustics class was when I asked the professor how loud 1 acoustic watt was. He just said "It'd blow your ears off!"

It isn't that complicated to calculate

Assuming an omni-directional point source in free field, at 1 m distance, it is radiating uniformly to a spherical surface of area 4π m². When the source strength is 1 W, the sound intensity (power/area) at 1 m is 1/(4π) W/m². Using this formula from Wikipedia (which is basically an alternate form of the more familiar dB SPL formula from sound pressure at standard sea level conditions), the sound intensity/pressure level at 1 m distance is 10 log10( 1/(4π) / 10⁻¹² ) = 109 dB.

 

[Floyd Toole]

Dr. Toole, a speaker designer told me that these days, flat frequency response and directivity are solved problems given the availability and sophistication of modern modelling software and DSP. One only needs to be a competent designer and know how to use the software and you can create "any spinorama you want" (his words, not mine). About the only unsolved problems are driver resonances, power handling, and distortion - although all are improving. In your view, is he correct? And what do you think are the remaining unsolved problems in loudspeaker design?

I think the large and growing collection of great looking spinoramas from several loudspeaker manufacturers in several countries around the world indicates that he is right. In the "old days" of not all that long ago, transducer design was a serious challenge, a slow process, involving trial and error, and resulting in systems that were crippled by flawed components. Now, thanks to a measurable target performance and computer aided design tools, transducers have improved enormously. Sound fields radiated by transducers can be significantly predicted before a physical prototype exists. Transducers can be custom designed to meet many different objectives.

With many sources of well engineered transducers, designing good systems is much easier, and the results are in the marketplace. It has been a long time coming. Now, the challenge is to convince the purchasing public and professionals that many of the old audio beliefs were really myths. Good sound is not the result of golden ears or unobtainium diaphragms, it is competent engineering - and measurements.

In the past loudspeakers from different manufacturers sometimes had "signature" timbres - a house "sound". More often they were just variable sounds. These differences added a dimension to audio as a hobby, and we see evidence of it in these forums. However, the notion that loudspeakers might one day sound as similar as amplifiers and wires creates anxiety. Relax, because at last we might get to hear the recorded arts as they were created. Isn't it the art that is important, not the technology? It may be a fantasy, but it is a worthy one.

Power handling and non-linear distortion are also part of transducer design, and these are factors that can force compromises in acoustical performance if very high sound levels are demanded. But, again, computer modelling and innovation are providing solutions. When it is not necessary to build a physical prototype to test an idea, the ideas can be tested quickly.

Who knows, perhaps the day will come when AI will be designing loudspeakers. It already is designing chips and algorithms. I wish your designer friend a long and prosperous career

 

[MattHooper]

I am very familiar with the Quad 57 and 63, and show anechoic measurements and subjective evaluation results in all editions of my book.

Dr Toole… out of curiosity….

I understand that you would not choose to own either of those loudspeakers for daily use. But did you ever enjoy music through Quads and could you enjoy listening to any music through them now? Or has your familiarity with more advanced speaker design made the flaws in the Quads too obvious to ignore?

 

[Newman]

One of the few quotes I remember from my grad acoustics class was when I asked the professor how loud 1 acoustic watt was. He just said "It'd blow your ears off!"

Yes...
and...

It isn't that complicated to calculate

Assuming an omni-directional point source in free field, at 1 m distance, it is radiating uniformly to a spherical surface of area 4π m². When the source strength is 1 W, the sound intensity (power/area) at 1 m is 1/(4π) W/m². Using this formula from Wikipedia (which is basically an alternate form of the more familiar dB SPL formula from sound pressure at standard sea level conditions), the sound intensity/pressure level at 1 m distance is 10 log10( 1/(4π) / 10⁻¹² ) = 109 dB.

Another way to 'visualise' it, is to think of most conventional modern loudspeakers as being roughly 1% efficient at converting electrical power into acoustic power, so if we play 100 Watts of amplifier power into it, it will be outputting 1W of acoustic power.

That's how loud! Pretty loud!

cheers

 

[Floyd Toole]

Dr Toole… out of curiosity….

I understand that you would not choose to own either of those loudspeakers for daily use. But did you ever enjoy music through Quads and could you enjoy listening to any music through them now? Or has your familiarity with more advanced speaker design made the flaws in the Quads too obvious to ignore?

You obviously have not read my books, because the Quad 57 was a candidate in my first blind, loudness balanced, randomized, multiple loudspeaker comparison way back in 1966. The most preferred loudspeaker in that test exhibited the same basic qualities we look for today: flat and smooth on-axis response and smooth and gradually changing or constant off-axis response. The Quad did not win, but it was not disgraced. It delivered a good direct sound, but the large diaphragm areas caused substantial off-axis irregularities. Anechoic measurements are shown in Figure 1.3. The much improved Quad ESL 63 was used in an elaborate stereo vs mono loudspeaker directivity test described in detail in Section 4.4.

Could I enjoy music through them now? Yes, music is very durable. The ESL 63 would be my choice; it has no gross flaws. The large diaphragm areas required of electrostatic loudspeakers is a persistent challenge, but Peter Walker struggled mightily to overcome it. I drank Scotch with him in his living room, listening to ESL 57s at the time he was designing the ESL 63. I sat in his listening chair at the factory and listened to a prototype - in mono, by the way.

 

[Floyd Toole]

Yes...
and...

Another way to 'visualise' it, is to think of most conventional modern loudspeakers as being roughly 1% efficient at converting electrical power into acoustic power, so if we play 100 Watts of amplifier power into it, it will be outputting 1W of acoustic power.

That's how loud! Pretty loud!

cheers

And also, the 109 dB was for a point - truly omnidirectional - source. When the same acoustic power is confined to a fraction of the sphere, as in typical loudspeakers, the sound level is even higher.

 

[DonH56]

Thanks for that....all makes sense.
Especially the part about staying away from ASR ! lol.... keep telling myself it's a bad habit...

Habits are hard to break. Witness my continued posting.

Anyway, I can see (and have known) that measuring efficiency is a different measurement than a simple formula conversion from sensitivity.
Often I've seen where using manufactures specs, the sensitivity and efficiency don't tie together using the general conversion formula.
I think because most of the time sensitivity is just eyeballed from the response curve, to a number the manufacturer feels they can justify.

No, it's because the test methodology and analysis is usually different.

Personally, I like to measure sensitivity as an integration across the driver's pass band. I use pink with all driver processing in place, and measure average RMS voltage at driver terminals, while simultaneously with measuring acoustic LEQ SPL. Gives real world voltage sensitivity imo. Add in measuring average current over same time interval, and real world power sensitivity too. Todays true RMS, averaging meters make it pretty easy to do i think.

Broadband pink noise is what I advocated for and from what @Floyd Toole said it sounds like things are heading that direction.

Anyway again, I think the subject gets way too much attention.....generally from just wanting to prove points...lol again

Agree with that!

 

[DonH56]

It isn't that complicated to calculate

Assuming an omni-directional point source in free field, at 1 m distance, it is radiating uniformly to a spherical surface of area 4π m². When the source strength is 1 W, the sound intensity (power/area) at 1 m is 1/(4π) W/m². Using this formula from Wikipedia (which is basically an alternate form of the more familiar dB SPL formula from sound pressure at standard sea level conditions), the sound intensity/pressure level at 1 m distance is 10 log10( 1/(4π) / 10⁻¹² ) = 109 dB.

View attachment 501796

Didn't say I could not calculate it, or at least look it up, and I paid for my question by having the prof assign exactly that problem for us to solve! Need(ed) to learn to keep my mouth shut... But IIRC he specified a significantly smaller radiating area, to match a typical speaker and room, which in turn led to higher SPL. Not that 109 dB SPL isn't already pretty durn loud to me.

 

[Newman]

[to Newman] And also, the 109 dB was for a point - truly omnidirectional - source. When the same acoustic power is confined to a fraction of the sphere, as in typical loudspeakers, the sound level is even higher.

Yes, I was debating whether to include that in my reply, and couldn't decide whether to apply a half-space, quarter-space, or eighth-space correction, then realised it would differ with frequency and speaker placement, so I opted not to mention it!

 

[DonH56]

Yes, I was debating whether to include that in my reply, and couldn't decide whether to apply a half-space, quarter-space, or eighth-space correction, then realised it would differ with frequency and speaker placement, so I opted not to mention it!

I have a vague memory of 120 dB for the solution to what he gave us, but the prof had the semi-vexing habit (like Richard Feynman) of creating problems that looked gnarly, with weird numbers and assumptions/givens in the problem statements, to produce a very simple number as the resulting answer. He said it made it easier to grade papers. I tried that when I was giving lectures but it was often enough hard to set up such a problem so I quit trying.

 

[MattHooper]

You obviously have not read my books, because the Quad 57 was a candidate in my first blind, loudness balanced, randomized, multiple loudspeaker comparison way back in 1966. The most preferred loudspeaker in that test exhibited the same basic qualities we look for today: flat and smooth on-axis response and smooth and gradually changing or constant off-axis response. The Quad did not win, but it was not disgraced. It delivered a good direct sound, but the large diaphragm areas caused substantial off-axis irregularities. Anechoic measurements are shown in Figure 1.3. The much improved Quad ESL 63 was used in an elaborate stereo vs mono loudspeaker directivity test described in detail in Section 4.4.

Thanks. I was aware of the quad blind tests and that they performed fairly well in some areas, but not in others, which is the reason my question came to mind. Since you have so much experience in research/helping design better speakers, I wondered if over time this had diminished your ability to enjoy a speaker like the quads. Thanks for clearing up with your answer.

Could I enjoy music through them now? Yes, music is very durable. The ESL 63 would be my choice; it has no gross flaws. The large diaphragm areas required of electrostatic loudspeakers is a persistent challenge, but Peter Walker struggled mightily to overcome it. I drank Scotch with him in his living room, listening to ESL 57s at the time he was designing the ESL 63. I sat in his listening chair at the factory and listened to a prototype - in mono, by the way.

Awesome! What a life!

 

[Newman]

The linked to diagrams attributed to Purify are complete fakes. There is absolutely no way any physical system can respond like that to a square wave.

That's not the point. Neither I nor Purify are saying they are measurements. But they do show the principle of loudspeaker operation.

The point being made is that, to many lay audiophiles, it can seem intuitively obvious that the SPL (pressure waves in the air) won't mimic the applied signal voltage as accurately if the speaker has a heavier diaphragm or a weaker magnetic field, because they think that the SPL tracks the 'shape' of the diaphragm's velocity or displacement graph. But in fact it is tracking the acceleration, and because F=ma, making a (fixed) change to 'F' (magnet strength) or 'm' (diaphragm mass) simply changes the (fixed) magnitude of 'a' (acceleration) and hence the magnitude (not shape or accuracy) of the sound waves. The sound waves just get louder or quieter, they don't adopt a different shape to the input signal.

cheers

 

[NTK]

The linked to diagrams attributed to Purify are complete fakes. There is absolutely no way any physical system can respond like that to a square wave.

It all depends on the scale of the time (horizontal) axes

But more seriously, the square pulses shown in the figures are for current/force/acceleration/pressure. The limit of how quickly acceleration or force can change is nowhere as "limiting" as for velocity or displacement since inertia plays much much less of a role. The main bottlenecks in this chain would, I think, first be current. Current is necessary for the generation of the motor force (motor force is directly proportional to current), and its rate of change is limited by the inductance of the system. And the next limit would be the rigidity of the mechanical components.

That is the main gist of Purifi's blog post. Inertia (i.e. mass) does not play a direct role in determining how "fast" loudspeaker drivers can respond in terms of sound pressure generation.

[Edit] Inertia play a direct role in limiting how fast the position of the driver diaphragm can change, and how fast its velocity can change. The role of inertia in limiting how fast the driver diaphragm acceleration can change is very minor and secondary.

 

[audiofooled]

I have a vague memory of 120 dB for the solution to what he gave us

Perhaps this was the case where you have only acoustic power expressed in watts compared to the reference acoustic power, like in this equation from Leo Beranek's book:

Thinking about it, it has nothing to do with distance or radiation pattern or surface area, just power at the source. If reference power is 10 times greater (10^(-12)), that would result in 120 dB.

Sound Power

Sound power from sources like fans, jet engines, cars, humans and more.

www.engineeringtoolbox.com

 

[Arindal]

It has been a long time coming. Now, the challenge is to convince the purchasing public and professionals that many of the old audio beliefs were really myths.

Me thinks, the announcement that technically perfect sound reproduction through research is finally coming and that the ´only thing left to do is educating the public´ on good vs, bad sound or solid engineering vs. myths, is not really new. Rather like 50+ years old, maybe dating back to your first experiments. Looking back, this seems to have failed spectacularly in the past, paving the ground for all sorts of audiophile beliefs, expensive high end audio, renaissance of outdated tech like vinyl records or stereo.

What makes you think that this time things will work out? Hi-fi folks who have their beliefs, in the vast majority seemingly stick to them, tend to ignore what is presented as scientific findings, and in the end of the day believe what they experience in own listening tests.

In the past loudspeakers from different manufacturers sometimes had "signature" timbres - a house "sound". More often they were just variable sounds. These differences added a dimension to audio as a hobby, and we see evidence of it in these forums.

Apperently, nothing has changed since then. Looking at the market, I would say the importance of a ´house sound´ deviating from an acoustic ideal, to successfully market loudspeakers, has rather increased. Products which emerged in recent decades and turned out to be finding their niche for success, all have their recognizable house sound, from bass-heavy portable bluetooth speakers to big hi-fi speakers with dull tonality or either boosted or recessed brilliance. The main difference to a more random approach to sound is that today one can taylor such house sound, even hiding it behind flat frequency response graphs and excellent THD specs.

I don't know a single example of a speaker being successful just for sounding neutral and showing ideal measurements. Do you?

because at last we might get to hear the recorded arts as they were created. Isn't it the art that is important, not the technology? It may be a fantasy, but it is a worthy one.

We obviously share the same theoretical ideal, but looking at modern days´ typical home environment, room acoustics, level of care when it comes to speaker positioning and requirements for speaker size, connectivity, preference for different music and alike, me thinks we are even further away from any standardization than ever. People have bigger and bigger living rooms with close to no absorption, prefer very tiny speakers and vast listening distances, don't care for stereo triangle and other important parameters.

Even if there might be some sort of approximation to a theoretical ideal of flat anechoic response among some speaker manufacturers (which I doubt is a mass movement), I would assess this gets more than overcompensated by increasingly disadvantageous conditions within the living rooms. Sadly.

 

[DonH56]

Perhaps this was the case where you have only acoustic power expressed in watts compared to the reference acoustic power, like in this equation from Leo Beranek's book:

View attachment 501870


Thinking about it, it has nothing to do with distance or radiation pattern or surface area, just power at the source. If reference power is 10 times greater (10^(-12)), that would result in 120 dB.

Sound Power

Sound power from sources like fans, jet engines, cars, humans and more.

www.engineeringtoolbox.com

Quite possible. I took the last grad course ca. 1983 or 1984 so my memory is fuzzy. We covered a lot of ground and I may have mixed problems in my head trying to think back. I pulled my old textbook, but the index is almost worthless, and I did not see acoustic power after a quick glance. I thought it would be fun, and it had its moments, but mostly it was a grind through multidimensional differential wave equations. What I do not have, and wish I did, were my lecture notes that did a much better job of explaining things.

Fundamentals of Acoustics, 3rd ed., Lawrence Kinsler, Austin Frey, Alan Coopens, James Sanders, Wiley and Sons, 1982.

 

[kyuu]

Apperently, nothing has changed since then. Looking at the market, I would say the importance of a ´house sound´ deviating from an acoustic ideal, to successfully market loudspeakers, has rather increased. Products which emerged in recent decades and turned out to be finding their niche for success, all have their recognizable house sound, from bass-heavy portable bluetooth speakers to big hi-fi speakers with dull tonality or either boosted or recessed brilliance. The main difference to a more random approach to sound is that today one can taylor such house sound, even hiding it behind flat frequency response graphs and excellent THD specs.

This is complete nonsense.

I don't know a single example of a speaker being successful just for sounding neutral and showing ideal measurements. Do you?

There are a lot of examples. Are you being deliberately obtuse?

On another note, kindly refrain from being overtly hostile and rude to Mr. Toole. This has nothing to do with putting anyone on a pedestal, but I'd prefer he didn't get driven off by hostile posters as one could hardly blame him if he decided it's not worth spending his retirement dealing with that.

 

[Floyd Toole]

Me thinks, the announcement that technically perfect sound reproduction through research is finally coming and that the ´only thing left to do is educating the public´ on good vs, bad sound or solid engineering vs. myths, is not really new. Rather like 50+ years old, maybe dating back to your first experiments.

Looking back, this is probably the most frustrating fact of all. The fundamental answers were evident in my 1966 experiments, and greatly reinforced with competent anechoic measurements in 1986/86 that visually correlated with double-blind subjective sound quality ratings. The answers were published for all to see, but buried in the Journal of the Audio Engineering Society, not being discussed in real time on an Internet forum or presented in lectures on YouTube. The means of communication were primitive, and the world was slow to realize that the research existed.

Then come the matters of belief, scarce measurements, and marketing. Back then loudspeaker design was a blend of art, craft and engineering, with "golden ears" playing a role, and only a small percentage of those designing loudspeakers read the journals. To this day, I explain decades old science to people who are surprised that it existed. Manufacturing loudspeakers is a business, and businesses must make profit to survive, and that relates directly to sales. It was obvious to me and others, that sales did not correlate well with sound quality. Some consumers might have thought they were looking for sonic perfection, others responded to "excitement" in sound, and still others just wanted it loud. These customers still exist, but satisfying their desires can now be delivered by the same (timbrally neutral) product + tone controls or an equalizer.

When I transitioned from being a full-time research scientist at the NRCC to being a corporate officer of a $500M audio corporation in 1991 it was a dramatic change. Harman had several consumer and pro audio brands, like JBL, Infinity, Mark Levinson, Lexicon, Crown acquired over the years, and Revel which was created while I was there. Each brand operated somewhat separately, with brand managers, marketing staff, and design engineers sharing corporate facilities. I set up a research group to continue the NRCC work for which I had been hired, and (amazingly) we were permitted to publish the research results - educating our competitors (who have learned well) - a rising tide lifts all boats.

From my perspective it was interesting to see how science was accepted among the Harman brands - some jumped at the new guidance, while others were more attached to the power of marketing to move product. The automobile engineers associated with our automotive audio business respected the science and that business grew dramatically, expanding into virtually all automobile services other than the drivetrain itself. It is now about 75% of a $7B business. Meanwhile, consumer audio plodded on, with the engineers adopting the latest knowledge and technology, only to find too often that the product may or may not sell well - for totally business management and marketing reasons. Consumers cannot differentiate between canny BS and facts, and subjective evaluations on a sales floor or at an audio show are not trustworthy experiences. Not all Harman products met our "standards", especially as the company developed off-shore design and manufacturing capabilities. In some locations attitudes towards science were decades old and my role expanded to giving tutorials on the science of audio. As forum members may have noted, it still goes on.

You said: "I don't know a single example of a speaker being successful just for sounding neutral and showing ideal measurements. Do you?"

I see you only recently joined this forum, but surely you have become aware of the obvious trend within the industry - at least the parts of it that respect science and competent engineering. The answer is of course, Yes.

Thinking back the first were startup Canadian manufacturers who were attracted to me and the NRC for what was offered to them by their government: knowledge and acoustical measurement facilities. In the '80s PSB, Paradigm, Axiom Energy, Mirage and several other branded loudspeakers were designed using NRCC facilities I had created for the research. From the outset their products were technically as good as or better than their international competition, which dealers and distributors realized when they toured our facilities and experienced double-blind listening tests. The results of such tests were published in the then well respected Canadian audio magazines. Some examples of these early reviews, showing measurements can be found in the website material for the 4th edition. They sold well and as a result the companies grew and prospered. Anechoic measurements of several of their products are in my books, all very respectable for the time. PSB continues to design products using the NRCC facilities while Paradigm and Axiom built their own anechoic and listening facilities. All respect the contribution of the National Research Council of Canada on their websites.

Another that comes to mind was Genelec. Ilpo Martikainen owned the company, designed the loudspeakers, and we started meeting at AES conventions in the '80s when I delivered my first papers. We connected regularly, each time Ilpo would take me aside, close the door of their demo suite and we would listen through all of his products from the smallest to the largest, demonstrating how similar and how good they sounded - as was confirmed by anechoic measurements. Genelec was one of the first manufacturers to publish honest anechoic data on its products, from which it was clear that they were competently engineered. Meanwhile, most other pro monitor manufacturers relied on "you can trust us, we're professionals" approach. Genelec is now among the ones to beat when it comes to accurate sound reproduction.

There are now several others who unashamedly display comprehensive anechoic data including spinoramas, KEF, Dutch and Dutch, Theory and a growing list of others who believe that displaying facts has value and virtue. Their products may or may not be in the minds of everyday audio enthusiasts, but they exist. Both facts and unsubstantiated marketing claims are in the faces of consumers, and time will tell which is the more powerful motivator for purchasers.

Meanwhile, audiophiles trust their ears and that is a problem, because subjective evaluations, opinions, about sound quality are not trustworthy under normal listening conditions. I posted this earlier in an ASR forum: " the real "weak link"in audio is human nature. We want to believe that what we think we hear at the moment is an absolute truth. Opinions matter more than facts." Add in some good marketing, a persuasive salesman and memories of forum chatter and opinions can take many forms only loosely connected to the sound actually heard. Anything can be sold in the absence of a controlled, blind, listening test or a belief in measurements.

Listening habits have changed. My own have changed. I rarely now listen in the stereo triangle, although I have one. Most of the music I enjoy, and I do enjoy it, is background music, delivered throughout our home using a combination of superb floor standers and well designed smart speakers, all of which deliver near enough to neutral timbre that the music is communicated, but lacking the soundstage and imaging. My imagination seems to fill in the blanks. I listen through noise cancelling headphones when travelling and occasionally at home when I don't wish to pollute the space with my musical choices of the moment or podcasts. The home theater is used daily for (depressing) news and (escapist) movies and streaming, with multichannel audio accompaniment with multiple subs. Nowhere in these experiences are resonances an apparent distraction - because all the manufacturers cared. For which thanks.

So, whether it is necessary or not, whether it benefits sales or not, whether the customer even knows it exists or not, many significant manufacturers have adopted good looking spinormas as their performance target. Some portables add too much bass, probably as differentiation, or acknowledging that playback may be outdoors or at low sound levels. But most of these devices have apps that allow tone controls or EQ. I have experienced TV sound bars that are impressively good for watching movies on a standard flat screen in a small room or even listening to music. I was in a friend's home recently, and the low level background music came from a fist-sized portable, sitting on the counter. It wasn't impressive, but it wasn't offensive either, being well balanced and resonance free, although bass limited and mono. It served a purpose very well. Over most of my lifetime, such acoustical performance was impossible.

It is a better audio world out there.