Is harmonic accuracy and timbre important at all?

Disclaimer: I am not Richard Hardesty in disguise. But I have reached similar ground after many years of listening and equipment swapping and upgrading and would enjoy discourse from a position that is simply not discussed enough here.

I feel a strong need to get on a soap box here, albeit friendly, and I don't mind a rigorous discussion on this topic. My hope is that, increasingly, manufacturers will take notice of this important aspect of music reproduction. I also know that it takes time, talent, money and dedication to accomplish accuracy of timbre in speaker design and that "shamanism" and "snake oil," along with major bux spent on fine cabinetry that may do little to improve the sound, exists everywhere in this industry.

I fully acknowledge that Dunlavy and Meadowlark, a least for now, are gone, and that only Vandersteen and Thiel survive amidst a sea of harmonically inaccurate, and frequently far more expensive, speakers.

Can you help me understand why anyone would want to hear timbre and harmonic content that is anything but as accurate as possible upon transducing the signal fed by the partnering amplifier? It seems to me if you skew the sonic results in any direction away from the goal of timbral accuracy, then you add, or even subtract, any number of poorly understood and potentially chaotic independent and uncontrollable variables to listening enjoyment.

I mean, why would you want to hear only some of the harmonic content of a clarinet or any other instrument that is contained on the recording? Why would you not want the speaker, which we all agree is the critical motor that conveys the musical content at the final stage of music reproduction, to provide you with as much as possible by minimizing harmonic conent loss due to phase errors, intentionally imparted by the speaker designer?

Why anyone would choose a speaker that does this intentionally, by design, and that is the key issue here, is something I simply cannot fathom, unless most simply do not understand what they're missing.

By intentional, I mean inverting the midrange or other drivers in phase in an ill-fated attempt to counter the deleterious effects that inexpensive, high-order crossovers impart upon the harmonic content of timbre. This simply removes harmonic content. None of these manufacurers has ever had the cojones to say that Jim Thiel, Richard Vandersteen or John Dunlavy were wrong about this fundamental design goal. And none of them ever tries to counter the fact that they intentionally manufacture speakers they know, by their own hand, are sonically inaccurate, while all the all the same in many cases charging unsuspecting so-called audiophiles outlandish summs of money.

Also, the use of multiple drivers assigned identical function which has clearly been shown to smear phase and creates lobing, destroying essentially the point source nature of instruments played in space that give spatial, time and phasing so important to timbre rendering.

I truly belive that as we all get better at listening and enjoying all the music there is on recordings, both digital and analog, of both good and bad recording quality, these things become ever more important. If you learn to hear them, they certainly do matter. But to be fair, this also requires spending time with speakers that, by design, demonstrably present as much harmonic phase accuracy that timbre is built upon, at the current level of the state of the art.

Why would anyone want a speaker to alter that signal coming from the amp by removing some harmonics while retaining or even augmenting others?

And just why in heck does JMLab, Wilson, Pipedreams and many others have to charge such large $um$ at the top of their product lines (cabinetry with Ferrari paint jobs?) to not even care to address nor even attempt to achieve this? So, in the end I have to conclude that extremely expensive, inaccurate timbre is preferred by some hobbyists called audiophiles? I find that simply fascinating. Perhaps the process of accurate timbre appreciation is just a matter of time...but in the end, more will find, as I did, that it does matter.
I too have wondered the same thing. It has taken time and listening to understand why it is important. FWIW I agree with you competely. I do feel you may have opened Pandoras box. I also suggest a good flame retardant suit as things may get a little hot around here. :-)

Some confuse accuracy with dry and lifeless. It is all a perception of what one believes harmonic and timbre accuracy is. I look forward to Agoner's comments on the subject.
The Vandersteens and Theils may be harmonically accurate, but may also lack speed, dynamics, transient capability, and other qualities that are important to many audiophiles.

Also, once you put any loudspeaker into a room and connect it to a bunch of random components, it could be the most accurate speaker in the world, but it sure won't measure or sound that way. I think that is when some of the other performance parameters become important.

I find it interesting that many audiophiles claim they strive for accuracy, yet their room acoustics (and lack of addressing serious room problems) may never allow sound that is anywhere near accurate. I mean, would you buy any amp that measured within plus or minus 6 to 10 dBs from 30Hz to 15kHz? That's what many folks are listening to.

Plus, everyone has a different idea of what sounds "correct" and accurate to themselves on an individual basis...
I got tired of reading your post the second time through but I couldn't figure out what you were saying the first time through. Is it: speaker builders intentionally invert the phase of midrange drivers? And are you saying that Vandersteen and Thiel do this or do not do this? Wilson does or does not? Thanks for the clarification.
If professional sound/recording studios use incorrect speakers to mix the albums they produce (which must be the case since they aren't using the speakers you listed, at least none of the really famous studios I know of), then what does it matter at all how correct your personal home audio speakers are? Anything you listen to was recorded in the studio to sound correct on the so called incorrect speakers.

I personally agree with Plato's points. You get rid of one fault for another when you strive for total perfection in one or two areas of speaker design, and total perfection isn't possible, which is probably why these perfect speaker companies aren't cornering the market and in a lot of cases have gone stagnant.

An acquaintance of mine was saying their teacher told them about an experiment someone did where they tried to make the perfectly "correct" system you are striving for. The teacher said it might have been correct, but it sounded so boring no one wanted to listen to it. I don't know if this was a real experiment or if his teacher was trying to make a point about the nature of sound, but I am interested in any one's opinion why his teacher said this or why this might be true.
I agree with Plato that harmonic and timbral accuracy are not all there is to compelling music reproduction. I'm sure they are the most important aspects for many people, myself included, but not for everyone, and there's nothing wrong with that.

My Harbeths do not share the design approaches of Vandersteen and Thiel. They image less well than those speaker lines, but in many ways seem to be more true to the sound of instruments. Which brand is more correct?
Steve, good post. The answer to your question is simple: Most speaker designers don't know the physics behind designing a time/phase accurate speaker. It's much easier to claim that it's not audable. I suggest that anyone owning a pair of well designed time/phase coherent speakers for 6 months would never be able to go back to what most manufacturers claim as "hi end" again. I have been listening to Green Mountain Audio speakers for the last 3 years and I can now hear the crossover in every non time/phase coherent speaker I hear.
Live sound direct from an instrument to our ears does not have delay that changes with frequency superimposed on its original response. It is an artifact of speaker physics. We would not tolerate such phase smear in our consoles, mixing boards, amplifiers, pre-amps or any other piece of gear. As speaker technology improves, the remaining clues that we are listening to speakers, such as distortion, horn signature and other artifacts, are reduced. Phase delay is a subtle but critical clue to our ears, and its reduction puts us closer to the real thing. All other things being equal, the speaker with the flattest phase response sounds the closest to being there live. Every time. To claim that our rooms cause problems therefore we should accept phase shifts on the order of one full cycle from our loudspeakers does nothing but fuel the fire for the designers that lack the knowledge to build a time/phase coherent product.
Great post songwriter!
"I suggest that anyone owning a pair of well designed time/phase coherent speakers for 6 months would never be able to go back to what most manufacturers claim as "hi end" again. I have been listening to Green Mountain Audio speakers for the last 3 years and I can now hear the crossover in every non time/phase coherent speaker I hear."

Glad I'm not the only one that notices this. I even hear it in 1st orders that aren't time/phase coherent.

Here's an article on the subject. You may become tired of reading it also. Phase, Time and Distortion in Loudspeakers .

Plato does make good points. There are certainly no perfect speakers. You pick your poison and live with it. I just hope designers don't get too far away from the time aligned and phase correct approaches.

With more and more people listening to compressed crap on Ipods. It looks dismal for the guys designing speakers without the hyped treble and pronounced midbass humps..

How come no one has anything to say about almost all recording studios using "incorrect" speakers to mix albums? I would think this would make most recordings inherently incorrect, making correct speakers moot, wouldn't it?

Isn't this also a problem with live music that isn't acoustic? Most live venues don't use "correct" speakers. Songwriter mentions "live sound direct from an instrument to our ears" but leaves out that much of that music is run through incorrect speakers in very large speaker arrays at large venues such as Madison Square Garden, or Giants Stadium, etc, which use incorrect speakers. Most live music is incorrect unless it is acoustic.

I don't understand why no one is addressing these issues. Don't we tolerate incorrect speakers all the time if you take these issues into consideration, or am I missing something? In particular B&W is used in some of the most popular and widely used recording studios around the world, and aren't those time-coherent instead of time-coincident?

I'm hoping someone will address what effects these issues have, so I can understand why they aren't significant. If all most all albums we listen to are time coherent but not time coincident, then why would it matter if we play these albums back on 100% "correct" speakers as opposed to fairly "incorrect" speakers. Please have some pity on me with the vocabulary, I do not have a lot of experience with all the terminology being used. I do not know a large amount of terminology but I am very interested in understanding the significance of this thread and all the issues therein.

Thank you,

I use to own the Dunlavy SC-II among many others and agree with Plato that there are many of ways to skin a cat. That's why I enjoy listening to my recently acquired Spendor S8e loudspeakers which are not time/phase coherent but they do the tone/timbre and texture thing quite well!*>)
Steve, in your listening experience, what speakers have you found to be harmonically accurate? Anyone else?
and some people claim they love coffee, but add so much cream to turn the color pale, and so much sugar to kill the taste altogether. Does this mean they don't really like coffee or don't know how to taste...

I happen to agree with the original poster, but believe a desire for harmoic accuracy depends upon a number of things. For many, distortion is a much sought-after effect to help with some inherent problems with the music:

1)The music doesn't benefit from accuracy (ie. electronically generated music, pop, rock, country, etc).

2)The musicians are not accurate (this includes 99% of recording musicians).

3) The engineering is not accurate (50 ft drums, voice filters, and voice that emanates from the entire soundstage).

4) The listener places a higher priority on other musical features (dynamics, smoothness, listenability, etc).

For acoustically generated music and professional singers and engineers that know their art, accurate music reproduction is a must. However, not everyone likes this music.

If I liked jazz, my rig would probably be quite different. Most of those performances are from the 50s, have a rawness about their sound, and don't always benefit at the hands of the engineer. It makes sense in these situations to add a little sugar to the cup. You lose some of the emotional content and the musical context, but you gain listenability.
Jkalman--mixing monitors are not part of the signal chain. With amplified music, the microphones are generally capturing either unamplified sound (a vocalist or drum kit) or the sound directly off a crossoverless reinforcement device, such as a guitar amp. Yes, there are plenty of upstream problems and some recordings sound better than others. But that doesn't negate the importance of getting it right at the end of the line.

Great post, songwriter. Not sure I completely agree, but well said.
Jkalman ,

Recording engineers may use different speakers for listening to their music.This isn't how the sound is recorded. It is recorded through microphones.

Songwriter is refering to live unamplified music. It's up to the person to tolerate listening to Concert or Club type amplified music systems. Nothing wrong with it if you like listening to this stuff on a daily basis. Personally If I had to listen to the club amplified type music all the time it would give me a headache.

Several years back..I thought of it as you do. It wasn't until I sat down and really listened to some time/phase coherent speakers did I understand what the fuss was all about.

We all have different taste in music. This is one of the factors when buying speakers. Take for example highly compressed music. You listen to it in your car or on some big Cerwin Vegas it sounds good. Take the same recording and play it on a speaker without the raised treble or midbass will sound like shit!! LOL

The speaker allowed you to hear the recording as it was recorded and mixed with no help from the speaker to fill in the rest of the sore spots. The same goes for time/phase coherent speakers. You hear the timing in the recording as it was originally recorded through the mics and other recording equipment, not a mock up of what it should sound like.This stuff is tough to explain. Putting it in words isn't easy for me..but I understand the concept completely.

There are other important factors to loudspeaker design.These are just two of them.
I have always thought that "Hi end" audio was about reproducing a given source as accurately as possible without requards to the quality of this source. Crap in, crap out. Good in, good out, etc. With this in mind, I would think that a pair of speakers should be able to reproduce an excellent recording correctly and make a bad recording sound bad.
I have noticed the number of people stating that the room influence negates accuracy, or this effects it and that effects it and therefore you really don't need accurate speakers. I consider this pure BS!
If something in the chain (source) starts out messed up, it will end up even worse after the cummulative effects of the chain ending with inaccurate speakers. You stand no chance of getting annything remotely correct.
I just think you stand a better chance with as accurate of components as possible. If inaccuracies are the rule of the day then why spend all this money. Are we buying looks or sound?
I feel speakers MUST start off as flat in frequency as possible and then get the other parameters as close as you can.
I agree with the original poster on the fact that once you use time and phase correct speakers, it's hard to go back. I too hear abnormalities in the sound of a lot of so called high end speakers that sell for a lot of bucks.
As for dynamics and other so called drawbacks of time and phase speakers, how do you know they aren't correct and you are listening to exaggerations of the original source with high slope speakers?
Amplified music in a live venue is an exaggeration of what is actually going on. Colorations are added through the electronic and speaker chain.
I played sax for many years and I can say without a doubt that time and phase speakers reproduce live sax better than any speaker I have heard. They do get the harmonic structure correct. I have used a live feed to test this.
The consumer is certainly free to purchase and use any speaker that he or she likes and sounds "Correct" to them. I certainly have no problem with that.
However, I do believe that from what "High end" once stood for, it has been transposed to what costs more. This becomes evident when arguments start over why should a speaker be accurate. If it's not to be technically correct, then what the hell is all this fuss over. Buy what you want and let it go! It reminds me of what is better, a Chevy or a Ford? They both get you from point "A" to point "B" same as a Lexus. I think "Status" has become the rule of the day.
Yes harmonic accuracy and correct timbre are important, but it's up to the individual listener to determine if they are the end all and be all of audio reproduction. Sometimes and in some situations other reproduction factors can be more important.

Until the day I see the two of you side by side in a room, I will fervently believe that Stevecham is Richard Hardesty.
when Plato stated

"Also, once you put any loudspeaker into a room and connect it to a bunch of random components, it could be the most accurate speaker in the world, but it sure won't measure or sound that way. I think that is when some of the other performance parameters become important"

i think he hit a vital point but i would interpret it a bit differently. With all the random variables in a system, room (probably the limiting factor in almost everyones set up admit it or not)and recording environment/process i would want my speakers to be as accurate as possible so as not to add coloration at the most important point in the signal chain. If your speakers are right you can begin to address the other issues such as room to get the whole thing right. Having inaccurate speakers, especially those whcih may have been designed in part by ear in the design environment, just adds one more thing to work around. Even as the rest of the system and room comes up to snuff that problem will remain.
Since we're talking about hamonic accuracy how come there are (not a one) no takers on my question? What speakers have you listened to that are harmonically accurate? Particulary the man himself: Richard Hardesty.
Alot of this is over my head but since I've had my Decware 1.5s, ALL other traditional speakers (as suggested in the opening and 2nd by songwriter) just sound wrong (plain awful?), no matter how they've done the crossovers. My Decwares aren't perfect: far from it. What they do get right is the sound, timber, call it what you will. With no crossovers, save for a capcitor on the tweeter, the sound seems so pure and unadulterated and whole. It may be off in some aspect; not have enough bass; etc. but it's VERY convincing. Making up for a poorly designed network by lavishing attention to cabinetry is just another way to market your product and seems rampant in this industry. By the way, 'The Audio Critic' used to point this out all the time and wasn't afraid to name names. I'm not saying he was right all the time (he wasn't) but he stood on his soapbox, as we all should, especially when it comes to such a great hobby and cost is such an issue.
The only speakers I've heard so far and spent any real time with that are close to harmonically accurate for me have been GMA Europas and my speakers the Brines Acoustic FTA-2000s. Some of the Thiel models do this well also. The Vandies have one of the flattest FRs I've seen. I thought the Meadowlarks also did Harmonics and Timbre well. They all may sound boring when playing mass market compressed crap however.

All are different designs but share similar traits. Well damped bass with no midbass humps, minimal phase shift, and no exaggrated treble. Also their FR responses are all extremely flat with minimal variations through the frequency range. I'm sure there are others I just haven't heard them .

I'm talking unamplified acoustic music without the hyping in the frequency extremes.
Recreating an exact acoustic analog of the original performance is a very noble-sounding goal, but is not possible at the current state of the art for most performances.

A much more practical goal would be this: To recreate the same PERCEPTION as would be experienced by a listener at the original performance. To do this requires the application of acoustic and psychoacoustic principles, some of which are well established, some relatively new discoveries, and perhaps some as yet undiscovered. Yet creating a perceptual replica is much more feasible than creating an actual acoustic replica.

I would say that tonal balance is the most important issue in recreating the perceptual replica, noting that the on-axis anechoic frequency response curve does not reliably predict perceived tonal balance. Tonal balance is of primary importance to timbral accuracy. I am not saying it's the only thing that matters, but it is I believe among the top few.

I have lived for years with time-and-phase correct loudspeakers, but based on my listening experiences would not claim that time-and-phase correctness is of primary importance in recreating the perception of a live performance. My understanding is that the current state-of-the-art in hearing mechanism theory discounts the audibility of phase above roughly 1000 Hz.
The problem I have with this idea is we are not reproducing a live event. We are reproducing what is on a disc that someone has decided what it should sound like for us. Since we ARE reproducing an electronic waveform from the source, I feel it should be as accurate as possible. Otherwise, what do you have? Anything different is a distortion is it not?
I will repeat, I think the goal of high end audio should be to reproduce the source as accurate as possible. Anything else makes all of this a moot point and high end means nothing except making people feel it's all about high prices.
I have lived for years with time-and-phase correct loudspeakers, but based on my listening experiences would not claim that time-and-phase correctness is of primary importance in recreating the perception of a live performance.
So Duke, what then is of primary importance?

I imagine that if one thing is of primary importance, it is just as important that several other dimensions, such as dynamics, at least meet some minimal level of performance.
One of the issues Audiokinesis brings up is if a speaker is phase/time accurate, but there is a difference between the on axis and off axis frequency response, then the in-room tonal balance will be compromised and the speaker will sound artificial. So it is possible for a speaker to accurately reproduce an input waveform, yet still not sound very good.
Hi Drubin,

I'd say the main attributes a loudspeaker should have for quality reproduction are: Natural-sounding tonal balance (with radiation pattern playing a critical but oft-overlooked role here); lack of audible distortion (which covers a huge territory and includes imaging and clarity and detail); lack of power compression within the required loudness range; and adequate extension at the frequency extremes.

Onhwy61, bingo. That's exactly what I was saying. You might enjoy taking a look at this site:

A fascinating book is being written by Earl Geddes, and he's posting the chapters online as he goes. You might find the second and third chapters, on "Psychoacoustics" and "Measurements" respectively, especially interesting.

I agree tonal balance is a major part of the equation. I also agree all time/phase coherent speakers don't guarantee a pleasing sound. At the same time Bigtee does have a very good point. There should be a standard in creating a waveform as close to what the speaker is fed from the source. If you didn't have this what will you have in the end? Don't get me wrong..I like some flavor in the music. But I want to add it myself upstream with tubes ..etc. Having an extremely coloured speaker from the start isn't a good idea IMHO.

If I'd known this a few years ago, I could have save myself a lot of money and heartache . There are post all the time about room issues like fat bloated bass etc. If the consumer knew from the start that the speakers he's about to purchase have a 6dB hump from 50 hz through 200Hz by its very design. It may save him from some of the headaches audiophiles post about almost every day. Of course this is ignored...and it is all the rooms fault. This doesn't make sense too me. This is good for the room treament manufactures as the money just keeps rolling in. I'm not saying their not needed. What I'm saying is you may need less of them with a properly designed speaker.
So many great points raised here, where do I start?

I asked Kathy at Thiel once if she or Jim knew of studios that monitored or mixed with Thiels let alone time and phase coherent speakers. They told me they had no knowledge. And even if the speaker is not in the chain, it is true that many recordings were made with the use of mono mix to ensure that ambient send/returns did not create cancelations so you can see how small imperfections in the subjective correction in that area might be caused by inaccurate monitors.

Believe me, sometimes I wish I were Richard Hardesty, or at least had the luxury of writing all day and listening to music and pissing people off. Instead, I do product management for a biotech firm. I'll bet Richard doesn't know what a polymorphonucelotide sequence is.

Yes mass market compressed crap is just that.

I'm not in the least concerned that we are not reproducing a live event but that we are accurately reproducing a recorded one.

Warrenh: I have three speakers that I believe are accurate. Meadowlark Kestrel2, questionable only because they are only two way and I don't think it is possible with anything less than three way. I also have Thiel CS6s and Vandersteen 2Ces, both of which I do believe offer accuracy that others I've owned, such as Paradigm Ref 100s V1, Dynaudio Contour 3.0, KEF Q15 never did. I used to have an original pair of Thiel CS7s and regret selling them for the Dynaudios. This was where my eventual revelation led me back to the sonic accuracy of the Thiel design, and now to Vandersteen. Hope this answers your question.

Hey I will stand on one unmoveable soapbox from the measurement angle in all of this: if the speaker can't do the triangular step response, then there is no way it can ever be timbrally accurate, even if it is flat FR wise; the time domain will eventually be realized as essential to recorded music reproduction in speaker design. Mark my words.

Thank you all again for such intelligent and stimulating discourse.

But even if a speaker can do a good step response it doesn't mean the speaker will sound accurate in a typical listening room. It's a lot more complex than you're making it out to be.
I think sounding accurate and being accurate are 2 different things. The speaker is either accurate or it's not. If the room is messed up, changing the speaker is not going to help. Treating the room is the only option.
If a speaker comes off the technical development path and its basic characteristics are flawed, it will never get that back. It may "Sound" good to some but sounding good and accurate are 2 different things.
It has been my experience over the years that most "Audiophiles" will not like an accurate speaker. We have all been exposed to inaccurate reproduction and developed what is probably a skewed look at things. We are looking for what "Sounds" good to us and not what is correct. I think we can blame a lot on poor sources. The CD for the most part stinks. Sure, they are a few good sounding disks around but most are pretty bad. A lot of the old remastered CD's to SACD sound even worse on that format. It exposes their flaws and I think accurate speakers do this also. Nobody wants 3/4 of their collection sounding bad. So we tune it to "Sound" good most of the time. Nothing wrong with that but if you just accept the fact, you can sure buy great audio for a lot less money.
I just think a lot of us, including myself are chasing our tails on a never ending quest that will never be fulfilled.
2 channel audio is slowly dying away being replaced by multi channel which IMO is even a bigger pile of crap. Here, inaccurancies are the rule of the day and I'm not sure I would want a "Accurate" home theater. I want a home theater that sounds good! Stark contrast to 2 channel.
In conclusion, people are going to buy what sounds good to them no matter what the technical specifications. There will always be the dissenting few that demand accuracy, quality and value for their dollar. But just look at what most of the population buys. Can I say Best Buy!
It is a real shame that audio is going the direction it is. More and more pricey equipment and more and more of the same with the same arguments.
When you have no real standards, things get pretty chaotic.
I think, "If it makes you happy, then you have done good!"
Get off the trail and buy more music. That's what it is all about---right?
With all the technology up to now: if you were before a $250,000 rig, in a perfect acoustic setting, and a live acoustic instrument, (you pick it) how long would it take you to know (eyes closed of course) which was which? Under 10 seconds? Less?
Bigtee, I agree with you completely but I am firmly in the "sounds good" camp all around. I think of my system more as a musical instrument than a surgical one.

Steve, how far do you extend your quest for accuracy? To the sources, the electronics and cables? Signal in signal out sort of stuff? Do measurements determine accuracy? If so, which ones, how many and who decides? Doesn't decision involve subjectivity?

Accuracy, however you define it, is but one component of preference; the former maybe not all that applicable to audio given the complexities of the latter.

Which is more accurate, Stadivarius or Guanarius? Maybe which do you prefer is the more appropriate question.
Warrenh, About as long as took the first string of the guitar to be plucked or the first horn to blow! I don't believe you can buy close to live at any price and never will. That's why I think we chase our tails so much.
Actually, it's kind of depressing. What's an old audiophile to do?
To mangle a phrase: Accuracy is in the ear of the beholder.

I've listened to a lot of "accurate" speakers, including some mentioned above, which were time-aligned for tremendous sound staging or which were detailed to where you could follow every nuance of the musician's fret work, but which missed the whole musical point. The sax doesn't sound like a sax. The piano doesn't sound like a piano. My first criteria for accuracy is that the instruments sound real. I admit this requires timing alignment, flat frequency response, adequate rate of attack, high level of detail, etc. Somehow the whole needs to be more than the sum of the parts. To me, this is the art of speaker design.
Khrys: I'm done. With the two systems I have, one tube, one solid state, I'm happy for the long haul. I like both Thiel and Vandersteen's designs equally, even though they are different, because regardless of the system, I can hear the depth of the music and at least be able to enjoy all that the musicians and producers were trying to convey. That alone let's me forget about the whole equipment thing, just kick back and get into the music without distraction, except to log on here once in a while of course ;-)
I really can't agree that accuracy is "In the ear of the beholder." As mentioned above, accuracy means different things to different people. When dealing with a technical product, you must adhere to some sort of standards defining accuracy. Accurate signal reproduction is the goal. (Maybe you should change the source and not the speaker!) It is a measurable quantity even if all of the measures are not complete due to parameters not fully defined by science at this point.
Your ears are not really what I would call a accurate measurement. If you had a "True" hearing test, you would find that as for frequency, our ears are no more accurate than some of these so called accurate speakers (especially as we age.) It's kind of like designing a bearing based on noise instead of tolerance. Most assume our hearing just rolls off with time but that is not the real case. We have "Response deviations also."
What disturbs me about the whole audio thing is we have speakers coming out that offer "Better accurancy" better this, better that and increase in prices. The so called "High end" is killing itself. If we debate accuracy as we do and then go back and say, "Well it sounds good so we buy it," then what are we really doing. We DON"T need all this debate because everyone will buy what sounds best to them and be done with it. Checking A'gon's listings tell me everyone is not exactly satisfied with their choices. And "Moving up" doesn't cut it because a lot of the equipment being sold is accurate stuff.
I am a firm believer that "Truly measurable accurate speakers" based on current science offer a better chance at sounding good most of the time. You can really look at todays "Full" set of measurements and get an idea to how a speaker will sound. Manufacturers do it all the time! Yea, you can tweak a little by ear to have a flavor but this is "The designers" idea of good sound that may or may not compare with a buyers thoughts.
I also think that frequency response is a starting point. Big deviations here and everything else becomes a moot point.
Crap, this could go on forever.
My bottom line is why do we pay so much for products that sound good which sort of negates the engineering aspect. I mean, where does cost fit in this. It's like designing a better match and charging 3 times for it! It still lights the same fire.
Accuracy (especially timbre) is important to me for the obvious reasons stated above. To a large degree, overall frequency response is easily corrected upstream if needed - but the subtleties of timbre are not. If a speaker can't even reproduce a moderately flat frequency response, than it's patently impossible that it could be timbrally accurate. So it's not a speaker I would want personally.

However, it depends what you mean by your question - none of this is important to the typical buyer of expensive stereo equipment. As Hardesty points out, what sells best in the showrooms is "Boom and Sizzle". So if you manufacture and sell high end speakers are you going to strive for accuracy - unless that is your particular market segment (like Vandersteen or GMA)? Not if you want to stay in business. Remember the "Loudness" button" on your old receiver? Well, some guys would leave that sucker pushed in all the time - and others, even as teens, understood what it was for and when to unengage it.

(Any photographers out there remember what Velvia did to Ektachrome? Same thing - wildly inaccurate, oversaturated colors - but it looked great and became the standard if you were doing stock or advertising work.)

And that's OK - each to his own. Reproduction accuracy is not an ethical standard like "truth", it's a preference. And it's not a very popular one. Even "truth" itself isn't exactly a popular standard anymore. Not to get into politics, but look around! So why on earth would audio accuracy be important?

Wilson, to use a hackneyed example, are inaccurate speakers - but they don't advertise them as "accurate". They advertise them as sounding great and looking great. And to most people they do. They have a built in "Loudness button" and look like Lamborghini airscoops with woofers. The average doctor, lawyer, or hedge fund guy that picks up a copy of stereophile and orders a pair of Maxx's wouldn't in a million years consider having a pair of Vandersteens sitting in their living room. (BTW, I'm a hedge fund guy and my wife's a doctor, so I feel comfortable dissing them here :)

Certain speaker manufacturers have simply identified a target market that spends big bucks and they cater to it. It's called "high end".

Perhaps the problem is the casual interchangeability and confusion by marketers, reviewers, and consumers of the two terms "audiophile" and "high end". They are not (to us audiophiles) the same thing. If anyone wants to go on a crusade it should be to get everyone to define and use the two terms differently.

Audiophile equipment - designed for the pursuit, obviously not 100% attainable, of totally faithful reproduction of the information contained in the original source media. Wilson does not really qualify.

High End - designed to be great sounding and impressive looking with superb cabinetry/finish, utilizing high quality electromechanical components. Vandersteen does not really qualify.
Opalchip, Good post and what I was actually trying to convey in my post. I like yours better! Maybe we should just come up with a new buzz word. High end sure lost it's original definition but so have a lot of other words in todays society.
As usual Opalchip conveys what I'm thinking better than I can convey myself ! LOL

Great post
Though at times I needed to keep that loudness button pushed in, while in college I remember the day I realized it sounded much better with it out and the treble and bass controls set flat. The revelation happened on a day when no one was around and I decided to crank it up and find the best setting.

Well said.


I don't think it is as grim a picture as you paint. Speaker design is necessarily an exercise in compromise. Phase alignement should be critical in producing a speaker that gives accurate timbre. I agree that this is necessarily compromised in many passive speaker designs (meaning that certain instruments may not sound quite right). However dynamic range, sensitivity, bandwidth, frequency linearity to name a few are also important. In achieveing a balance, a speaker designer will compromise in some areas to gain in others.

Active speakers, for example, should have much better phase characteristics simply because each driver is paired with active filters including phase adjustment. Therefore active speakers should have a better timbral response....well not necessarily as this link below shows that professionals in a shootout were divided on which speaker they preferred and most found the timbre of the Dunlavys (except in the LF) to be preferred. In the end, classical listeners preferred the Dunlavys and meanwhile "rockers" preferred ATC's.

Considering that the Dunlavys also use multiple cones for mid range and that this is known to increase phase issues (a design no no), the conclusion seems surprising. Until you realize that the perception of timbre is also closely linked to harmonics, perhaps Dunlavy makes up for phase issues in other ways (harmonics) to eventually score higher than ATC's in the timbre category.

So speaker design is all about balance and not a single pursuit of only one or two factors.

In the end, two vastly different engineering approaches (ATC vs Dunlavy) have produced two great speakers.

...maybe there is more then one way to skin a cat.
Stevecham: if you were listening at low volume, you probably were better off leaving the loudness button pushed in. At higher volumes - approaching the "right" level for the recording in question (every recording has one), of course you want to turn it off.

I think you need to go back to the posts by Audiokinesis. Take a look at the Earl Geddes articles he recommends. Go to the Linkwitz Labs web site, and read what Siegfried Linkwitz has to say. It may well be that there are more important factors to our perception of audio reproduction than time/phase coherence. I say this while owning a pair of Green Mountain Audio Callistos, and having owned Meadowlarks in the past. I also own active speakers with fourth-order crossovers, which have benefits of their own.

There is also the HUGE factor of the implementation. Vandersteen, Meadowlark, Thiel, Green Mountain, Dunlavy all aimed for first order, time/phase coherent crossovers. But their speakers are totally different from one another! Ported, sealed, transmission line, simple crossovers versus highly complex (my god, take a look at a Thiel crossover - what is happening to the electrical signal as it goes through all those passive components???). The signal may pass the triangle test in the lab, and sound not very good in a real room. As an aside, last year after reading all of Hardesty's raves about the Vandersteens, I made the one hour drive to the shop run by his former partner. Heard the 5A's, presumably set up properly, powered by all the gear that Richard Hardesty likes. The resulting sound was muffled, muddied, utterly lacking in life compared either to live music or to the same recordings (NOT compressed crap) played on other systems I've heard. I'm still open to hearing the Vandersteens again - there's no way they can have such a following and sound like that! - but I think most humans would find the sound of other systems closer to "real" rather than what I heard that particular day. On the other hand, the Green Mountain Audio Callistos have tremendous "jump" to them. Anyone who understands that dynamics are part of music, not just flat frequency response, should give these speakers a listen.

Flat frequency response, on axis, in an anechoic chamber is not going to guarantee that your bass response, in room, is going to be free of major peaks and nulls - which can wipe out any sense of time/phase coherence when listening. That's why Linkwitz designs his speakers the way he does: to have good, even power response in real rooms, without needing to pile on bass traps and the like, which can end up sucking the life out of the music in other ways.

Sorry to run on at the keyboard now, but I can't resist. On the Zu Cable website you can see the rave review their Druid MkIV received from HiFi World magazine in the UK. The photo image of the magazine pages is too fuzzy to read, but the frequency response chart is plain enough. Yeah, those speakers may do away with any crossover nasties, but that response plot is... well, "flat" would not be found anywhere near it.
"...maybe there is more then one way to skin a cat."

Yes 4th order crossover by Green Mountain that intentionally present 360 degrees out of phase, yeah that'll fix things.

Look, if the speaker demonstrates that wave fronts are leaving the planes of the drivers out of phase and out of time relative to one another, then nothing can be done to correct it before it reaches your ear. This is a matter of timing, and timbre accuracy, regardless of the other so called parameters, cannot be corrected or compensated. If it's damaged, then nothing can correct it.

And nothing can be done before it reaches your ears.

Steve, are you sure of " Yes 4th order crossover by Green Mountain that intentionally present 360 degrees of phase, yeah, that'll fix things." is correct? Roy of Green Mountain has been a very generous contributor here on Audiogon and has always been a rather strong supporter of 1st order cross-overs.
Yeah, that sounds completely wrong.

Also, how can something be 360 degrees out of phase?

I agree with you. A phase irregularity or sharp slope on the phase vs frequency plot is very likely to end up with something that is very wrong; but all I am saying is that even if wrong, it may sound pleasant.

Nothing could create sharper phase irregularities than having two cones side by side producing the same output (since the cones will never match perfectly...just like a singer can never sound exactly the same every time they are asked to repeat something)

The case of the Dunlavy's dual mid ranges having a preferred timbral response, by audio professionals, over what in theory should be a superior design (ATC) demonstrates that some phase misalignment may be preferred by some listeners!!! (see my previous post in this thread with a link)

How can this be? How can worse or wrong sound better?

I suspect that having two mid ranges is akin to the often used "voice-over" in the recording studio => the result is that the voice sounds thicker and more resonant due to very slight but sharp phase misalignments (it is a bit like having more vocal chords). This effect will also tend to give an orchestra a bigger and richer sound by "virtualy" doubling the number of instruments playing, hence classical listeners liked the Dunlavy's more.

Bose used this same technique of multiple drivers in 901's and other designs to produce a sound that many liked and was called "spacious" ( phase misalignments from many multiple cones ouputing the same frequencies might indeed be expected to sound spacious )

All of which suggests that there are two extremes to making speakers;

1) making accurate speakers
2) speakers that sound pleasant,

and this is why I liked Opalchip's comments. Of course, manufacturers will most often position themselves somewhere in between the two extremes. ATC is closer to 1) than Dunlavy but not by much. Wilson I believe is slightly closer to 2) than 1). Bose 901 might be very close to the extreme of 2).

This is not to say that 1) is better than 2) or 2) is better than 1). There are advantages and disadvantages to speakers that give a rich pleasant sound. There are advantages and disadvantages to speakers that are highly accurate. It is nice that we have a choice!

4th order Butterworth is not 1st order.

sorry 1st crossovers are used, my mistake

thanks for setting this straight