What is the prime determinant of sound quality?

[Steve F]

Time for a new topic, since this has veered off the original.

Soundminded has pointed out something that I have noticed. Once I got my AR3a pair sounding "right" I was able to use a graphic equalizer to make a pair of DCM Timewindows (with a subwoofer) sound remarkably similar to the ARs. (Timewindows have a tweeter arrangement like an Allison 1). I can't believe all the years of frustration that could have been avoided if I had not been so adamant about not using tone controls or outboard processors.

JMTC after months of lurking. This a great board.

Thanks

Abigail

This was reported in Stereo Review magazine--

Many years back (1973 or so), a well-respected audio designer took 2 pairs of AR-7's (2 spkrs per channel) and angled them so as to closely approximate the radiation pattern of the LST. He then equalized the 7's so their far-field response was as close to the LST's as he could get it (within their LF frequency limit).

He then played material that didn't have significant bass below 50-60 Hz and A-B'd the LSTs against the paired 7's.

They were almost indistinguishable, leading this individual to conclude that FR and spectral balance is the primary determinant of a speaker's identifiable sound quality. This experiment presents a very compelling case for that viewpoint.

Aadams' post lends independent credibility to that stance.

Lot's of us like to think that other, more "subtle" and "esoteric" traits are what make speakers sound radically different from each other, but it may just be that good 'ol FR is the BIG ONE. Not the only one, but the big one.

Steve F.

[soundminded]

The sound eminated by a musical instrument is determined by the way it is constructed and the way it is played. This principle can be applied to human voices too. The sound that reaches your ears depends on your distance from the source of sound, its orientation, and the effect of room acoustics. In large public venues where concerts are often heard, the effect of acoustics is the overwhelming preponderance of sound you hear. But the importance of the first arrival whether it comes directly from the source or is an early reflection should not be minimized. It generally contains the highest percentage of high frequencies which help us characterize the timbre of sounds and differentiate them from one another.

The effort to record and recreate musical sounds by mechanical and electronic means with high fidelity has been generally disappointing. Few honest observers with normal hearing familiar with the sounds of actual music would often be unable to differentiate between the real thing and the playback of a recording. This despite enormous advances in many areas related to relevant technologies over the last 110 years or so since practical recording devices were possible.

I think this is in part due to the fact that most of those who have investigated this area are electrical engineers. While much advance in electronics has been made, advances which allow engineers to amplify, store/transmit, retrieve/receive and control electrical signals that are supposed to be the analogs of musical sounds with great precision and at very low cost, the science of acoustic is not based on electrical phenomena but on a specialized branch of mechanical engineering called fluid dynamics. The relevant fluid in acoustics is air.

Were engineers who are well versed in this area to study the problem of sound recording and reproduction seriously at its most basic levels and without any preconceptions, they'd see the obvious fatal flaws in the designs and even the underlying concepts and assumptions those designs are based on.

It was clear by the early 1970s that two channel stereophonic sound by itself was not going to progress sufficiently to advance the state of the art much further. Much money and effort was spent on the failed technology called quadraphonic sound, the predecessor of SACD and modern home theater sound also incapable of true high fidelity. The latest efforts by those not satisified to squeeze the last possible drop of blood out of stereo are those experimenting with another failed concept, binaural sound. There are multiple variants of it being experimented with but none of them really works very well even to being able to make it appear that a fly is buzzing around your head. So enjoy what you have, it isn't likely going to get much better anytime soon.

[Steve F]

Interesting info, as always.

But...you didn't answer my question.

What, in your opinion (and everyone else's) is the primary determinant of a speaker's sound quality? FR? Dispersion? THD? Something else?

This is a speaker Forum. What do think determines a speaker's sound? What is the main thing or things--in what order--that makes an AR sound different from, say a B&W?

Steve F.

[Shacky]

Curious as well. I use to get into testy exchanges with Zilch on the subject. Not being well learned on the science and nomenclature of audio I made many mistakes. Talking about "accuracy" and other things.

But why do AR 3a's sound so good so right? I have a restored pair of Theil 03a's that I'd think are in same ballpark as B$W. Even with their active EQ making bass similar to AR's, I can't listen to old Beatles. They sound thin and unwelcoming.

But on AR's I could listen to old Beatles all day - preferably my Mono Box Set.

[soundminded]

Interesting info, as always.

But...you didn't answer my question.

What, in your opinion (and everyone else's) is the primary determinant of a speaker's sound quality? FR? Dispersion? THD? Something else?

This is a speaker Forum. What do think determines a speaker's sound? What is the main thing or things--in what order--that makes an AR sound different from, say a B&W?

Steve F.

The more I studed the phenomenon of FR the more complex it became. The bottom line is the sound field that reaches the listener is what counts. This relates to the sound field propagated by the speaker and the way the room interacts with it. Therefore if this answer is right, flat on axis response isn't enough and flat power radiation isn't right answer either. (Peter Snell had the same answer about the criteria of the sound that reaches the listener I learned years after I came to that conclusion.) AR speakers had a big advantage over other speakers because of their super wide HF dispersion. In live rooms more typical of those most people live in than dealer demo rooms are, AR speakers produce much more early lateral high frequency reflections and some off the ceiling too than other speakers produce (except for Roy Allison's.) This is one reason why you can equalize them to sound flat and the don't become shrill. IMO if you don't get the treble right nothing else matters. But AR's unique strength comes from its ability to produce the lowest octave of sound other speakers can't. When you get the top octaves in balance the lowest octave doesn't overwhelm everything else and make recorded music sound bass heavy. The octaves in the audible range that are least important in the perceived quality of sound are....none of them. Every one of them is important if you want acoustic instruments to sound accurate. On the other hand, if you're just looking for something pleasing, some might be satisfied with B&W I suppose :-)

[Carlspeak]

Interesting info, as always.

But...you didn't answer my question.

What, in your opinion (and everyone else's) is the primary determinant of a speaker's sound quality? FR? Dispersion? THD? Something else?

This is a speaker Forum. What do think determines a speaker's sound? What is the main thing or things--in what order--that makes an AR sound different from, say a B&W?

Steve F.

A friend of mine, after listeneing to a newly built DIY speaker project of mine coined a non-scientific term, "noodge factor". That is, if he can listen to music for an extended period without fidgeting around, he deems it a good sound reproducer.

Personally, I think Floyd Toole does a pretty good job of answering your question in his book, "Sound Reproduction". Of course, there will be those who claim Toole's 'tome' is Harman-biased. The audio society I belong to is currently in discussions about a meeting topic similar to your question. I think someone put it in perspective, saying 'everyone's opinion of what sounds 'right' is different'.

[Steve F]

A friend of mine, after listeneing to a newly built DIY speaker project of mine coined a non-scientific term, "noodge factor". That is, if he can listen to music for an extended period without fidgeting around, he deems it a good sound reproducer.

Personally, I think Floyd Toole does a pretty good job of answering your question in his book, "Sound Reproduction". Of course, there will be those who claim Toole's 'tome' is Harman-biased. The audio society I belong to is currently in discussions about a meeting topic similar to your question. I think someone put it in perspective, saying 'everyone's opinion of what sounds 'right' is different'.

Yes, but what speaker characteristics make it sound 'right'? Toole is pretty much an axial FR, first-arrival guy. Do you agree with that as being what makes a speaker sound like it sounds?

Steve F.

[Carlspeak]

Yes, but what speaker characteristics make it sound 'right'? Toole is pretty much an axial FR, first-arrival guy. Do you agree with that as being what makes a speaker sound like it sounds?

Steve F.

What sounds 'right' to me may not sound 'right' to you (as I wrote earlier). So, what's the point of arguing our opinions?

Toole did the research with human listening panels which I believe represent a resonably significant group size. Thus, Toole's conclusions should be deemed as scientifically valid - representing an 'avarage' of listening prefrences on what characteristics sound 'right'. I have no basis to argue against axial FR other than my personal opinion.

Perhaps a poll of CSP readers might be helpful. Prepare a list of characteristics, each listed individually and in pairs with other characteristics (i.e. FR, or FR+first arrival, etc.) and see how the readers respond. The characteristics have all be identified. All you need is to list them and see which 'cream rises to the top'. :-)

[soundminded]

"

Toole did the research with human listening panels which I believe represent a resonably significant group size. Thus, Toole's conclusions should be deemed as scientifically valid"

Toole's major scientific contribution if you can call it that to knowledge of acoustics was finding out where subwoofers should be placed in a typical small listening room to obtain the most uniform bass response in different parts of the room. His answer is that you need four and only four (more will not necessarily be merrier) and that they should be placed in the corners or at the midpoints of the walls. That's it.

As far as I can tell, Toole's other research is in the area of market preferences which has nothing to do with scientific knowledge but with catering to tastes to maximize sales and profits. Profits are what Sidney Harman was reportedly about, at least that what I understood from people who knew him. Ironically, Toole's findings that early lateral reflections are important is the same conclusion Leo Beranek came to about preferences for the acousitcs of concert halls, yet today's designs fly in the face of that conclusion by producing speakers with very narrow dispersion patterns especially at high frequencies. They deliberatly produce litte if any lateral HF reflections (opposite of AR's philosophy.)

Toole's TOTL design was the Revel Salon Ultima which had a rear firing tweeter in addition to the front firing one. This helps create lateral HF reflections arriving at the listener. In version two which was produced after he retired....the rear firing tweeter was removed.

[soundminded]

"Yes, but what speaker characteristics make it sound 'right'?"

If you're referring to making recordings sound like live concerts at real and large public venues, nothing known that is commercially available can do that. I've explained why before. But if you're talking about what makes recordings of soloists or small groups sound like they are in your room and not coming out of boxes, that's a combination of the direct and indirect arriving field having flat FR all the way back to the recording mike and having a distribution of angles of arrival of sound energy comparable to what musicians themeselves would produce if they were there. This IMO is one reason why Bose 901 was so popular, it produced this type of distribution of energy uniquely even though FR wasn't even remotely close to flat. While large area radiating panel speakers including bipolar types are usually better in this regard than conventional forward firing box speakers, they aren't nearly as effective at it as 901 was. It is probably hard for any audiophile who ever heard 901 to believe that this design could be salvaged and made to work. Knowing exactly what had to be done it took me four years on my second try to figure out how, the first attempt ending in failure about 10 or 15 years earlier.

Interesting that the closer you get to getting it "right" the more you become aware of even minor flaws others would probably overlook. That's how I became a golden ears, by finding flaws and trying to fix them. And with every new recording, it's a whole different problem. BTW, the only way to make sense of it is to record the settings for all of the variables so you can get back to where you were the last time. And if you change the design even a little, you can throw all of your old settings away, you have to start all over again with every recording.

[Aadams]

I don’t know the answer and you guys sound way more informed on this topic than I but the only I reason I got into using an equalizer is because my AR3a pots (after refurbishment) were not equally smooth over the entire adjustment range. My solution was to subjectively adjust the mid and hi pots so the outputs were the “same” to my ear for both speakers and then dial them in with a ten band equalizer. The equalizer adjustments from octave to octave were so radical to me yet so pleasing that I wondered to what extent my ultimate displeasure with so many speakers over the years could have been avoided if I had used an equalizer all along.

So I pulled out the Large Advents, AR2ax’s, Cambridge Model 6’s and DCM Timewindows (The latter two speakers had a sub woofer attached.)

What I discovered, using the AR3a as the target, is that an equalizer can make all of these speakers sound better in my listening environment. They sound as though they were voiced by the same person but they do not sound identical. The DCMs however sound eerily close to the AR3as and from reading this discussion thread I may know why. The DCM horizontal dispersion is 180 degrees. Vertical is 60. When I try to closely AB the DCM against the 3a what you guys are calling the far field sound is very, very similar. The 3a has an elegant smoothness that the equalizer cannot impart to the DCM. The DCM has a certain clarity that the 3a can’t match. Go figure.

This is a great board.

Abigail.

[deek]

"Noodge factor." I love it! A measurable contemporary might be listener fatigue but that's more of a "how long can I stand it" measure, not a "how long can they keep me interested" measure.

All the attributes I can think of for determining sound quality boil down to frequency response and linearity across the spectrum. In my opinion, if you get those two things right the rest is up to the room and placement. The dependency upon the subject mater and subjectivity of the listener on rest of the measurements are inconsequential.

Example: if you're a listener of electronic, modern pop and generally "studio" type music like myself, the subject matter benefits very little from the efforts made to generate wide dispersion patterns. Likewise, a very "dead" room like my theater could probably benefit from some late HF reflection in some subject matter.

Again, this is my opinion and it's the reason I absolutely detest most of the Bose products. You can listen to any given system and practically name the "holes" in the frequency response. (Or point out the peaks) When you get "used" to a particular system, it's easy to hear the differences and as long as your system is in the neighborhood of "flat" then it's very easy.

-Deek

P.S. I've had a 10-band EQ on every system I've ever had... When I was cheap and wouldn't drop coin on decent components, I used them to compensate for poor quality and shoddy design. In my current system, the tiny nudges here and there are more to do with preference than shortcomings. I like a stronger midbass than "flat" and compensate. In my dead theater, I twist the high end slightly to bring out that "shimmer" in vocals. I would be perfectly happy with the EQ out (and I listen to most recordings with the EQ out first) but I have a preference that I can achieve with the EQ.

[speaker dave]

"

As far as I can tell, Toole's other research is in the area of market preferences which has nothing to do with scientific knowledge but with catering to tastes to maximize sales and profits. Profits are what Sidney Harman was reportedly about, at least that what I understood from people who knew him. Ironically, Toole's findings that early lateral reflections are important is the same conclusion Leo Beranek came to about preferences for the acousitcs of concert halls, yet today's designs fly in the face of that conclusion by producing speakers with very narrow dispersion patterns especially at high frequencies. They deliberatly produce litte if any lateral HF reflections (opposite of AR's philosophy.)

Toole's TOTL design was the Revel Salon Ultima which had a rear firing tweeter in addition to the front firing one. This helps create lateral HF reflections arriving at the listener. In version two which was produced after he retired....the rear firing tweeter was removed.

Perhaps you should get your facts correct. For many years prior to working at Harman Toole was at the Canadian NRC at Ottawa. There he worked in the acoustics laboratory with a number of other bright people such as Peter Shucks, Sean Olive and Mark Bonneville. They pioneered a rigorous approach to blind speaker testing and paralleled it with objective testing to find the correlation between listener preference and loudspeaker measurements. His landmark papers from the early 80's cover the subject very well and go extensively into which factors matter and which don't (more in a second).

He moved to Harman in the mid 90's after 27 years at the NRC. At Harman he was the VP of Research. I don't believe he ever developed products. Kevin Voecks is responsible for all Revel products, they are certainly not designed by Floyd (he is not a speaker designer). Floyds primary work was in developing uniform testing methods and improved facilities for listening tests, such as the fast speaker shuffler (a mechanical system that lets you a-b two speakers from the same room position).

I don't agree with absolutely everything that Floyd has published but he has done more to clear the air in this business than anyone else who's work I've read (Sean Olive would be second on that list). I've read most of his papers as well as his book and I see no commercial bias in it. It extends the work they began at the NRC and also pulls together and sumarizes respected works from other industry researchers. Rather than "market preferences" he proved that both sophisticated and inexperienced listeners could come to consistant ranking of speaker accuracy via methodical and carefully controlled blind tests. I would recommend Sean Olive's blog for multiple topics on listener preference and listening test methodology. (although I haven't been able to get it to open in a while)

http://seanolive.blogspot.com/

As to what is the prime determinant of sound quality, Sean Olive did a facinating study where many measurements of a group of speakers was taken as well as a ranking of the units with carefully controlled blind listening tests. Having both subjective rankings and a battery of measurement he could do factoral analysis between the two. Essentially this says: "if speaker A is well regarded and speaker B is not, in what way does A measure well and B badly?" With a good group size and a broad number of measurements, you can actually divine the weighting factor of each and every measurement such that the measurements alone would rank order the speakers the same as the listening tests did. His weighting factors gave very high correlation with the subjective rankings.

His results (from memory) were that on-axis response flatness, on-axis response smoothness, off axis response smoothness, and bass extension (-10dB point) alone would properly rank order a group of test speakers. Note that off-axis response smoothness does not require a particular power response shape, just a smoothness that verifies that major resonances are not present.

High frequency dispersion was not a factor.

David S.

[soundminded]

Perhaps you should get your facts correct. For many years prior to working at Harman Toole was at the Canadian NRC at Ottawa. There he worked in the acoustics laboratory with a number of other bright people such as Peter Shucks, Sean Olive and Mark Bonneville. They pioneered a rigorous approach to blind speaker testing and paralleled it with objective testing to find the correlation between listener preference and loudspeaker measurements. His landmark papers from the early 80's cover the subject very well and go extensively into which factors matter and which don't (more in a second).

He moved to Harman in the mid 90's after 27 years at the NRC. At Harman he was the VP of Research. I don't believe he ever developed products. Kevin Voecks is responsible for all Revel products, they are certainly not designed by Floyd (he is not a speaker designer). Floyds primary work was in developing uniform testing methods and improved facilities for listening tests, such as the fast speaker shuffler (a mechanical system that lets you a-b two speakers from the same room position).

I don't agree with absolutely everything that Floyd has published but he has done more to clear the air in this business than anyone else who's work I've read (Sean Olive would be second on that list). I've read most of his papers as well as his book and I see no commercial bias in it. It extends the work they began at the NRC and also pulls together and sumarizes respected works from other industry researchers. Rather than "market preferences" he proved that both sophisticated and inexperienced listeners could come to consistant ranking of speaker accuracy via methodical and carefully controlled blind tests. I would recommend Sean Olive's blog for multiple topics on listener preference and listening test methodology. (although I haven't been able to get it open in a while)

http://seanolive.blogspot.com/

As to what is the prime determinant of sound quality, Sean Olive did a facinating study where many measurements of a group of speakers was taken as well as a ranking of the units with carefully controlled blind listening tests. Having both subjective rankings and a battery of measurement he culd do factoral analysis between the two. Essentially this says: "if speaker A is well regarded and speaker B is not, in what way does A measure well and B badly?" With a good group size and a broad number of measurements, you can actually divine the weighting factor of each and every measurement such that the measurements alone would rank order the speakers the same as the listening tests did. His weighting factors gave very high correlation with the subjective rankings.

His results (from memory) were that on-axis response flatness, on-axis response smoothness, off axis response smoothness, and bass extension (-10dB point) alone would properly rank order a group of test speakers. Note that off-axis response smoothness does not require a particular power response shape, just a smoothness that verifies that major resonances are not present.

Hig frequency dispersion was not a factor.

David S.

"Perhaps you should get your facts correct."

Your condescending and insulting attitude does not make you right. I will attempt give you a civil reply which will not descend to the kind of flame war this subject often degenerates into for some irrational reason I have never been able to understand.

Once upon a time there was an actual thing called music. At its best, it was written by people who had remarkable insight into how the connecting of sounds could be interesting, pleasing, completely absorbing even to the most contemplative minds. It was sometimes performed by people who spent a lifetime perfecting their art, occasionally performing individually or in groups playing instruments that had been crafted to produce the most beautiful sounds those intstruments were capable of. Special buildings were constructed not just so that large numbers of people could enjoy it at the same time but to further enhance the sounds the musicians produced. The technological pursuit of duplicating those sounds so that at least that aspect of it could be enjoyed by a wider number of people at times and places they didn't have access to music became what was called high fidelity sound. It was a noble and worthwhile pursuit and that is what Acoustic Research Inc and a handful of other companies were about, developing the technology that would one day close the gap between what people heard when they listened to real music and what they heard when they listened to facsimilies of music through machines playing recordings of music. That effort advanced to a point and then stopped. Ultimately the pursuit of that goal was a technological failure and for all practical intents and purposes no longer exists. In fact that industry has become so perverted, so distorted that the notion of what is music and what are facsimiles is confused to the point where common usage now calls the facsimile music as though it were the real thing itself.

For those of us who still have this quaint notion of how things should be, Acoustic Research was at the pinnacle of this effort because that was their goal once upon a time and their LvR demonstrations proved that at least under very special circumstances they could nearly achieve that goal if even on rare occasion. It was not only a demonstration that they were on the right track but that there was hope that if they continued, the joy real music brings, the way it enriches people's lives would expand. Unfortunately that golden era some of us look back on with fond memories ended rather abruptly. That is in part why what this industry has evolved into instead, is now dying. It deserves to die, it no longer serves any valid purpose, it's once hoped for continued achievement has reached a dead end and now it lives on an endless parade of variants of the same products offered at perpetually escallating prices for its inadequate technology. This is the reason some of us go to the trouble to still search for those products that were the best that were ever offered or who go their own route like I do starting with acquiring and restoring what was the best efforts when the goal was clear and trying to build on the knowledge those pioneering efforts left us as their legacy.

Floyd Toole represents the trend that followed this golden era. You said it yourself; "They pioneered a rigorous approach to blind speaker testing and paralleled it with objective testing to find the correlation between listener preference and loudspeaker measurements."

That is not what high fidelity sound recording and reproduction is about. That is what market research to maximize profits is about. The only science there is the science of how to make the most money. That is what Sidney Harman is and always was about. It was only natural for them to join forces. There are far more people in the world who know and like the flavor of Coca Cola than ever tasted Chateau Laffitte Rothchild. In a laboratory if you can't duplicate the taste of Chateau Laffite Rothchild by other means than planting a vinyard, harvesting grapes, fermenting and aging it into great wine, you mix chemicals to invent Coca Cola and tell people it is just as good. Adding grain alcohol to grape juice doesn't do it. For those who never even tasted Chateau Laffitte Rothschild, they wouldn't know the difference and could be convinced if the substitute were advertised heavily enough. There are no appologies or excuses, no explanations that can convince those who have tasted the real thing frequently enough to remember it clearly that will ever successfully pull off that sham. That's how I see it.

As for Floyd Toole and the original Revel Salon Ultima, whether he performed the R&D himself or supervised others who did it for him, it was done under his aegis following his directives and was for all practical intents and purposes his brainchild, the fruits of his thinking.

[Carlspeak]

Perhaps you should get your facts correct. For many years prior to working at Harman Toole was at the Canadian NRC at Ottawa. There he worked in the acoustics laboratory with a number of other bright people such as Peter Shucks, Sean Olive and Mark Bonneville. They pioneered a rigorous approach to blind speaker testing and paralleled it with objective testing to find the correlation between listener preference and loudspeaker measurements. His landmark papers from the early 80's cover the subject very well and go extensively into which factors matter and which don't (more in a second).

He moved to Harman in the mid 90's after 27 years at the NRC. At Harman he was the VP of Research. I don't believe he ever developed products. Kevin Voecks is responsible for all Revel products, they are certainly not designed by Floyd (he is not a speaker designer). Floyds primary work was in developing uniform testing methods and improved facilities for listening tests, such as the fast speaker shuffler (a mechanical system that lets you a-b two speakers from the same room position).

I don't agree with absolutely everything that Floyd has published but he has done more to clear the air in this business than anyone else who's work I've read (Sean Olive would be second on that list). I've read most of his papers as well as his book and I see no commercial bias in it. It extends the work they began at the NRC and also pulls together and sumarizes respected works from other industry researchers. Rather than "market preferences" he proved that both sophisticated and inexperienced listeners could come to consistant ranking of speaker accuracy via methodical and carefully controlled blind tests. I would recommend Sean Olive's blog for multiple topics on listener preference and listening test methodology. (although I haven't been able to get it open in a while)

http://seanolive.blogspot.com/

As to what is the prime determinant of sound quality, Sean Olive did a facinating study where many measurements of a group of speakers was taken as well as a ranking of the units with carefully controlled blind listening tests. Having both subjective rankings and a battery of measurement he culd do factoral analysis between the two. Essentially this says: "if speaker A is well regarded and speaker B is not, in what way does A measure well and B badly?" With a good group size and a broad number of measurements, you can actually divine the weighting factor of each and every measurement such that the measurements alone would rank order the speakers the same as the listening tests did. His weighting factors gave very high correlation with the subjective rankings.

His results (from memory) were that on-axis response flatness, on-axis response smoothness, off axis response smoothness, and bass extension (-10dB point) alone would properly rank order a group of test speakers. Note that off-axis response smoothness does not require a particular power response shape, just a smoothness that verifies that major resonances are not present.

Hig frequency dispersion was not a factor.

David S.

Thanks Dave for expounding and clarifying on what I was trying to say previously regarding Toole's work.

I still feel that whatever the reason was that drove the research effort, it's inconsequential because the work will stand on it's own regardless.

[Gerry S]

For me, it's definitely "tonal balance" as the MOST important single determinent for sound qualty if you want a "simple" answer. But what determines "good" tonal balance can (and has) filled volumes of books. Not a "simple" question".. like asking "what factor determines a perfect marriage". OR how do you even DEFINE a "perfect marriage" ????

If one is a "true audiophile", one has an "emotionally intimate connection" with "good" sound quality (and the gear associated with obtaining it).

For "audiophiles" and "music lovers", the difference in "sound quality" between am and fm radio is pretty obvious. But music lovers who are NOT audiophiles could care less about such obviously profound differences in "quality".

[ligs]

I like speakers which allow you to hear separate instruments and details. This is above and beyond smooth frequency response and low coloration. I wonder what measurements can point to this quality. My guess is it may have something to do with the transient response?

[Gerry S]

I like speakers which allow you to hear separate instruments and details. This is above and beyond smooth frequency response and low coloration. I wonder what measurements can point to this quality. My guess is it may have something to do with the transient response?

What you desire depends more on what's done at the "recording end"... the techniques that recording engineers use for accurately capturing a particular sonic event. At the "playback end", the very best speakers can do this because of the variables you mentioned and several dozen other technical parameters loudspeaker designers must tackle.

The "total listening experiance" results from the very complicated interaction between the source material, the electronics, the loudspeaker, and the room properties in which the speaker is being used( listened too).

[Mach3]

Soundminded,

Floyd Toole had NOTHING to do with the design of any Revel speakers. Toole was brought in as, Speaker Dave noted, to oversee R&D for Harman - not to design speakers. Those are two totally different areas of work - and the Revel speakers were NOT designed under Dr. Toole's aegis.

As Speaker Dave noted Toole's work at NRC - which was continued while at Harman - established a positive correlation between certain speaker measurements and subjective appraisals (BTW the Harman work found that Martin Logan and B&W were typically ranked much lower than a cheap Infinity model - you can read all about this on the Whats Best Forum in a sticky titled "Its a preference").

Harman is one of the few companies that has actually spent the time and money to try and determine what objective characteristics of speaker performance align with subjective preferences.

Please take the time to read Toole's book, his and Sean Olives various papers and the results of the Harman work BEFORE passing any subjective - and totally incorrect - appraisals of Toole's accomplishments.

Speaker Dave,

While I appreciate your timely and accurate reply - and I thought it was extremely courteous, on point and most importantly ACCURATE and not just a rambling irrelevant rant about Coke and Champagne - I do believe that the Harman research found that a highly dispersive driver (per mid-range or tweeter) was the most significant contributor the achieving the wide, smooth off-axis response that hundreds of subjective appraisals found the most entertaining to listen to (the exhaustive objective tests in an anechoic chamber also determined that widely dispersive drivers were most likely to provide that smooth, wide off-axis response).

Perhaps such is what most AR speakers do so very well - for their domes are, to this day, the most widely and accurately dispersive drivers ever created. I think Villchur discovered this and simply didn't make any public note of it - for Villchur was extremely creative and insightful AND had his own anechoic measurement facility.

[Gerry S]

Soundminded,

Floyd Toole had NOTHING to do with the design of any Revel speakers. Toole was brought in as, Speaker Dave noted, to oversee R&D for Harman - not to design speakers. Those are two totally different areas of work - and the Revel speakers were NOT designed under Dr. Toole's aegis.

As Speaker Dave noted Toole's work at NRC - which was continued while at Harman - established a positive correlation between certain speaker measurements and subjective appraisals (BTW the Harman work found that Martin Logan and B&W were typically ranked much lower than a cheap Infinity model - you can read all about this on the Whats Best Forum in a sticky titled "Its a preference").

Harman is one of the few companies that has actually spent the time and money to try and determine what objective characteristics of speaker performance align with subjective preferences.

Please take the time to read Toole's book, his and Sean Olives various papers and the results of the Harman work BEFORE passing any subjective - and totally incorrect - appraisals of Toole's accomplishments.

Speaker Dave,

While I appreciate your timely and accurate reply - and I thought it was extremely courteous, on point and most importantly ACCURATE and not just a rambling irrelevant rant about Coke and Champagne - I do believe that the Harman research found that a highly dispersive driver (per mid-range or tweeter) was the most significant contributor the achieving the wide, smooth off-axis response that hundreds of subjective appraisals found the most entertaining to listen to (the exhaustive objective tests in an anechoic chamber also determined that widely dispersive drivers were most likely to provide that smooth, wide off-axis response).

Perhaps such is what most AR speakers do so very well - for their domes are, to this day, the most widely and accurately dispersive drivers ever created. I think Villchur discovered this and simply didn't make any public note of it - for Villchur was extremely creative and insightful AND had his own anechoic measurement facility.

I have to agree with Mach 3 on both points. As mentioned in some other blog/bost/topic, I was THERE when Sean Olive gave his presentation @ the BAS. He seemed as proffessional, stoic and non-judgemental as any human being can doing true objective scientific research on a topic that's is IMO inherently extremely difficult to quantify. Other researchers try to objectively quantify "beauty", "love", "faith", etc. Floyd Toole And Mr Olive is IMO doing the same kind of research , and doing it without any "self promotion".

As to AR, I agree as well; they INVENTED "acoustic suspension" after all. Though I prefer "controlled dispersion" for "near field imaging", it's been demonstrated repeatedly over decades that smooth power response is critical to the total listening experiance. Unless one listens in an anechonic chamber as part of their "natural routine", dispersion plays a critical role in percieved sound quality whether in the near OR far-field. For that reason, even "the best headphones in the world" is no match for a even a "modest" but competant pair of loudspeakers! Why? Listening to headphones is listening to "first arrival " sounds ONLY; there is NO "reverberant field".

[speaker dave]

Although Edgar Villchur comercialized and popularized the Acoustic Suspension speaker he didn't really invent it and was predated by Olson and Preston of RCA. The RCA patent makes interesting reading. They discuss high compliance woofers, allowing the cabinet to provide most of the restoring force, and increasing linearity by relying on cabinet stiffness. Their interests were primarily in mass market radios but the example is acoustic suspension.

http://www.freepaten...com/2490466.pdf

I'm not sure that the AR domes have wider dispersion than any other manufacturers domes of the same diameter. For the most part a 1" dome is a 1" dome (and a 3/4" a 3/4") as far as dispersion goes. If you have any curves that show otherwise I would like to see them.

As to Toole and wide dispersion, he does seem to advocate a fair amount of room involvement but his tests don't show that wide top Octave dispersion is needed for a speaker to do well in his tests. Look at the data in his original AES paper and see that his top ranked speakers average a directivity index of about 9dB at 10kHz. (Hemispherical dispersion would be a d.i. of 3dB.) His tests, more than anything, show that on axis flatness and smoothnes, plus off axis smoothness (free from power response peaks that would indicate resonances) lead to high test rankings. Especially wide dispersion is not a positive factor in his test results.

David S.

[Gerry S]

Although Edgar Villchur comercialized and popularized the Acoustic Suspension speaker he didn't really invent it and was predated by Olson and Preston of RCA. The RCA patent makes interesting reading. They discuss high compliance woofers, allowing the cabinet to provide most of the restoring force, and increasing linearity by relying on cabinet stiffness. Their interests were primarily in mass market radios but the example is acoustic suspension.

http://www.freepaten...com/2490466.pdf

I'm not sure that the AR domes have wider dispersion than any other manufacturers domes of the same diameter. For the most part a 1" dome is a 1" dome (and a 3/4" a 3/4") as far as dispersion goes. If you have any curves that show otherwise I would like to see them.

As to Toole and wide dispersion, he does seem to advocate a fair amount of room involvement but his tests don't show that wide top Octave dispersion is needed for a speaker to do well in his tests. Look at the data in his original AES paper and see that his top ranked speakers average a directivity index of about 9dB at 10kHz. (Hemispherical dispersion would be a d.i. of 3dB.) His tests, more than anything, show that on axis flatness and smoothnes, plus off axis smoothness (free from power response peaks that would indicate resonances) lead to high test rankings. Especially wide dispersion is not a positive factor in his test results.

David S.

You're right about Olson. It's still hard to NOT to equate AR w/ "acoustic suspension". As to "domes", I wonder who truly "invented" the first dome tweeter. AR sure made it popular.

Did AR also make the first dome mid-range ? Aside from ADS/ Braun products, I don't know how many other companies chose the dome for midrange applications.

[ironlake]

What you like to hear and what you hear very seldom occur. Being a high school band director and listening to recordings of my concerts never sound the same to me. What I hear at the conductors podium never gets recorded with the mics on the poduim, so why should they sound the same. Mics are all different in their sound. There are too many variables for sound to be 100% of the original. speakers sound different, you just cant do any better than a somewhat reasonable account of the original. Ar 3a,s do the best for me of any speaker I have ever heard.

I prefer recordings recorded up close as that is how I hear it. I very seldom hear music from the 20 th row back so do not like that sound, but that is just me. People who go to concerts usually pick a certain seat section because the like the sound there. Someone else will go elsewhere. I prefer the front row of the balcony section but many others will not.

[speaker dave]

You're right about Olson. It's still hard to NOT to equate AR w/ "acoustic suspension". As to "domes", I wonder who truly "invented" the first dome tweeter. AR sure made it popular.

Did AR also make the first dome mid-range ? Aside from ADS/ Braun products, I don't know how many other companies chose the dome for midrange applications.

The AR 1 was certainly the first speaker in the modern mold. It was the first I know of that was designed, woofer and box together, to have flat response even at the expense of efficiency. In the end, adding mass until the response was flat, was the real invention. "Acoustic suspension" is more the marketing hook.

Not sure about the first domes. University did some early dome tweeters. Bill Hecht claims to have invented the soft dome. Philips did mid domes in molded paper. There must have been plenty of designers that tried the back radiation of a compression driver with no phase plug.

It is Hard to find anything in speakers with no precedence.

David S.

[Mach3]

I guess we have different information. To wit in regards air suspension:

Over the next two decades, almost all major loudspeaker manufacturers gradually changed from mechanical to acoustic suspension. At first they did so under license to AR, paying royalties to use the principles of Villchur’s patent. When the Electro-Voice Company refused to pay the royalties, AR sued them for patent infringement. Electro-Voice countersued, claiming prior art in the form of a mention of an air spring in a different system. The ensuing lawsuit resulted in the loss of the patent for Acoustic Research, a decision which Villchur chose not to appeal. In an interview about the case, Villchur says that he knew the judge’s decision to void the patent was incorrect, but that he felt he had better things to do than to spend his life in litigation. He cited the example of Edwin Howard Armstrong, the inventor of FM radio, whose patent was rendered unprofitable through the actions of RCA. Armstrong spent years unsuccessfully fighting that injustice, and eventually committed suicide. Villchur decided not to contest the loss of his loudspeaker patent, but rather to move on and continue improving the quality of high fidelity equipment.

And in regards the dome tweeter (or the direct radiating dome driver)

Villchur continued to research improvements in sound reproduction, turning his attention to the tweeter. He received U.S. Patent 3,033,945 for his invention of the direct-radiator dome tweeter. This greatly improved high-frequency fidelity by its smooth response and wide dispersion of sound,

As for Hecht and United Speakers and his soft dome technology?

All Hecht did was come up with a soft dome - one made with a rubber coating. He did NOT invent the dome radiator - Villchur did.

As for the wide spread belief that the AR dome radiators had better dispersion - well Dave - you come up with some data that shows they do not.