 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
In Reply to: Subjective reviews and differences in hearing - any research been done? posted by John C. - Aussie on February 19, 2001 at 17:03:40:
While marked hearing losses would no doubt limit the ability of a listener to differentiate sonic differences beyond their comprehension, I believe that the factors you mentioned are not the prevalent ones. The ongoing debate really centers around the ability to adequately measure anything relevant to the reproduction of music with today's "science". There is an excellent article by Robert Harley in the latest TAS concerning "objective" measurements vs "observational" differences. He correctly notes that audio is the only field that measures its products in ways that bear no relation to the way the products are used. Does anyone listen to and enjoy steady state test tones possessing zero dynamic range?For countless years, many have struggled to mimic the valuable and "secret" formula that earns the Coca Cola Company many billions of dollars per year in sales. Why don't you just measure what's in it and duplicate the recipe? Many have tried and failed. Why don't you just take apart a Stradivarius violin, "measure" the parameters, duplicate the "specifications" and make a veritable pile of money selling them? Again, many have tried and failed.
I strongly believe that there is substantially more at work in the perception of music as with the other examples I cited. Similarly, I believe that anyone, short of a former artillery specialist, is capable of perceiving subtle differences found in the music reproduction by audio componensts that continue to elude the white coats "listening" to their single tone distortion graphs - if they regularly listen to music and have developed a long term memory of what live music sounds like.
Follow Ups:
re: "He correctly notes that audio is the only field that measures its products in ways that bear no relation to the way the products are used."If this is Hartley's interpretation of an electronics manufacturing quality test procedure, best he stay out of the electronics biz and keep pushing the 'hype'.
I don't believe he is seriously knocking the industry that supports his habits and income, nor do I think he is that clueless. He's just stirring the pot.
steve
e-state wrote:Similarly, I believe that anyone, short of a former artillery specialist, is capable of perceiving subtle differences found in the music reproduction by audio componensts that continue to elude the white coats "listening" to their single tone distortion graphs - if they regularly listen to music and have developed a long term memory of what live music sounds like.
Yeah, audio researchers only use single tone distortion graphs, blacks are all lazy, Jews are all greedy, Italians all belong to the Mafia, yadda, yadda, yadda.
Much easier to swallow simpleminded, spoon-fed gross generalizations than the vastly more complex reality of things.
se
My intention was not to state that current manufacturers do not employ more exotic kinds of measurements, but rather that I think it is fair to say that observational listening ultimately plays a big role in the voicing of high end components. Do you believe that any designer worth his/her salt solely relies on the numbers for the performance of their products?What measurements, for example, do you find directly equate to the ability of a component to reproduce natural timbres of brass or woodwinds? How do you explain the fact that there are large numbers of excellent high end products designed by talented engineers that all sound different? Are some/many/all just stupid and don't understand the test gear? I think not.
There may eventually come a time when the science catches up and is able to quantify what as yet it cannot. I would hope that those in the industry would continue to refine such testing methodologies to meet that eventual goal. Just don't rely on them solely until then. Wasn't it only recently that science was able to explain how a bumblebee fiies?
e-stat wrote:My intention was not to state that current manufacturers do not employ more exotic kinds of measurements, but rather that I think it is fair to say that observational listening ultimately plays a big role in the voicing of high end components. Do you believe that any designer worth his/her salt solely relies on the numbers for the performance of their products?
You didn't make any specific mention of manufacturers in your original post. The closest you got to any specific mention was "the white coats" which seems to imply you were referring to those who are involved in the pure research of things audio. I just don't associate "white coats" with manufacturers (save for Matthew Polk, but even he's traded in his white coat for Yuppie attire).
But my underlying point was that too often those who spout the line "there's so much we don't know" have hardly an inkling of what we do know which leads to oversimplified, gross generalizations.
What measurements, for example, do you find directly equate to the ability of a component to reproduce natural timbres of brass or woodwinds?
That's sort of putting the cart before the horse. I think a better way of putting it would be to ask what measurements equate to the INability of a component to reproduce natural timbres of brass or woodwinds. And the simple answer to that is that any nonlinearity would contribute to a component's inability to reproduce natural timbres of brass or woodwinds.
But there are a number of other elements that complicate things just a bit. For example, how does one really know whether or not those natural timbres were even accurately captured during the recording process? Perhaps they weren't. So how do you know whether to blame the component or the recording? Is it a good recording because it's a good recording or is it because nonlinearities in the reproduction chain are accentuating certin things or masking other things which produce a final result which might falsely lead us to believe that the recording is closer to perfect than it really is?
The fact is that no one can listen to JUST the recording. Or JUST the turntable or CD player. Or JUST the preamplifier. Or JUST the amplifier. Or JUST the speakers. We can only listen to the result of the entire reproduction chain. Yet reviews would have one believe it's the sound of a particular component that's being described. And while the review may be qualified with a "this is how this component sounds in my system" I think that's rather backward and should read "this is how my SYSTEM sounds with this component in the chain."
Anyway, I'm going off on a bit of a tangent.
How do you explain the fact that there are large numbers of excellent high end products designed by talented engineers that all sound different?
There are any number of reasons that can explain that. But among those reasons is because perceived differences are not always inherently due to actual audible differences. How would you explain the fact that three different products, identical but for the name on the faceplate can all "sound different"?
There may eventually come a time when the science catches up and is able to quantify what as yet it cannot.
And there may eventually come a time when the high-end catches up and realizes that there's not necessarily always something there to quantify.
I would hope that those in the industry would continue to refine such testing methodologies to meet that eventual goal. Just don't rely on them solely until then.
Well at the end of the day all that matters is one's own perceptions of the end result and that wholly transcends any tests or measurements.
But with respect to basic research and expanding our knowledge and understanding, such subjective perceptions have limited use in determining the presence or absence of actual audible differences.
se
You didn't make any specific mention of manufacturers in your original post. The closest you got to any specific mention was "the white coats" which seems to imply you were referring to those who are involved in the pure research of things audio. I just don't associate "white coats" with manufacturers (save for Matthew Polk, but even he's traded in his white coat for Yuppie attire).I used the term "white lab coats" as a euphemism for engineers. It brings to mind visions of McIntosh engineers pouring over a THD plot from the old clinics they used to sponsor.
But my underlying point was that too often those who spout the line "there's so much we don't know" have hardly an inkling of what we do know which leads to oversimplified, gross generalizations
Fine. We know a bunch. You didn't address the question I asked concering the ultimate reliance on what we do know: Do you believe that any designer worth his/her salt solely relies on the numbers for the performance of their products?
And the simple answer to that is that any nonlinearity would contribute to a component's inability to reproduce natural timbres of brass or woodwinds.
Which nonlinearities? Though I am a software engineer and not an audio engineer, it is logical to assume that if the answer were so simple, then someone could put all the other designers out of business.
For example, how does one really know whether or not those natural timbres were even accurately captured during the recording process? Perhaps they weren't.
While it is certainly true that certain sonic characteristics can be "fixed" by offsetting elements in the reproduction chain, I assert there are some that cannot. I have a number of classical recordings on both vinyl and CD which possess very nice renderings of strings and woodwinds. On any of these recordings, my ARC VT-100 sounds more like live music than my Threshold Stasis. Arguably, the Pass amp out specs the Johnson amp.
How would you explain the fact that three different products, identical but for the name on the faceplate can all "sound different"?Huh? You wanna be more specific? Define "identical". Are we talking manufacturing variations in the various active and passive components?
And there may eventually come a time when the high-end catches up and realizes that there's not necessarily always something there to quantifyAre you of the opinion that current audio components are in every way completely perfect reproducers of the musical signal with absolutely no potential for improvement?
But with respect to basic research and expanding our knowledge and understanding, such subjective perceptions have limited use in determining the presence or absence of actual audible differences.
This is a restatement of your previous point. Over ten years ago, Ray Dolby was quoting in saying something to the effect that musical reproduction could not improved upon. I respectfully suggest he must not perceive a musical event like I do! I am amazed every time I go the the ASO how incredibly challenging is the task of reproducing the sound of one hundred individuals blowing, scraping, and banging on various instruments!
I don't dispense with the value of "objective" science of evaluating audio component performance. What I do suggest is that "observational listening" reveals details not yet quantified by the science. I am confident that there will continue to be audible improvements in the audio chain in the coming years.
Ralph Wallace
e-stat wrote:I used the term "white lab coats" as a euphemism for engineers. It brings to mind visions of McIntosh engineers pouring over a THD plot from the old clinics they used to sponsor.
Ah, ok.
Fine. We know a bunch. You didn't address the question I asked concering the ultimate reliance on what we do know: Do you believe that any designer worth his/her salt solely relies on the numbers for the performance of their products?
I'm sure like anyone else it ultimately has to satisfy them at the subjective level as well. So no, I don't think any engineer relies solely on numbers.
Which nonlinearities? Though I am a software engineer and not an audio engineer, it is logical to assume that if the answer were so simple, then someone could put all the other designers out of business.
No, I didn't say the answer was simple. I only gave a simple answer. You know, someone asks "How does an amplifier work?" and instead of starting with a lecture on basic physics and working your way up from there you say "Well the simple answer is..."
While it is certainly true that certain sonic characteristics can be "fixed" by offsetting elements in the reproduction chain, I assert there are some that cannot. I have a number of classical recordings on both vinyl and CD which possess very nice renderings of strings and woodwinds. On any of these recordings, my ARC VT-100 sounds more like live music than my Threshold Stasis. Arguably, the Pass amp out specs the Johnson amp.
First, don't confuse specs with measurements. For example "0.05% THD" is just a lump specification. It tells you next to nothing about the distortion itself.
But getting back to the point, how exactly do you know just how nice the renderings of strings and woodwinds are on the recording itself since you can only experience the recording through a rather complex chain of elements. Perhaps the Threshold is really giving you a more accurate representation of the recording than the ARC but the nonlinearities of the ARC produce an end result that you find more pleasing.
Take something like the TDS "harmonic enhancement" unit. It's sole purpose is to CREATE nonlinearities (i.e. copious amounts of odd-ordered harmonic distortion, frequency response anomalies with its attendant distortion in the time domain, etc.). Yet people also say that devices such as this make their systems sound "more like live music."
So again, how can you REALLY know just how good the recording itself is when it unavoidably must pass through the entire playback chain?
Are you of the opinion that current audio components are in every way completely perfect reproducers of the musical signal with absolutely no potential for improvement?
No, I'm simply of the opinion that what we perceive is not necessarily entirely the result of what we hear. This phenomenon is so trivially demonstrable that I fail to see how it can be denied. Yet the high end by and large continues to assume that any subjectively perceived differences are entirely due to some underlying physical phenomenon taking place within the equipment itself.
But NEITHER am I of the opinion that subjectively perceived differences are entirely due to psychological bias. Only that because such biases CAN effect our subjective perceptions, subjective analysis has limitations with regard to determining with certainty actual audibility of perceived differences.
This is a restatement of your previous point. Over ten years ago, Ray Dolby was quoting in saying something to the effect that musical reproduction could not improved upon. I respectfully suggest he must not perceive a musical event like I do! I am amazed every time I go the the ASO how incredibly challenging is the task of reproducing the sound of one hundred individuals blowing, scraping, and banging on various instruments!
I don't know what Ray Dolby's quote has to do with what I said so I'll refrain from any comment on it.
I don't dispense with the value of "objective" science of evaluating audio component performance. What I do suggest is that "observational listening" reveals details not yet quantified by the science.
Certainly. And all I am saying is that "observational listening" has its limitations due to the conscious and unconscious biases inherent in the human psyche which our egos mistakenly lead us to believe don't exist or that we're able to turn them on and off at will.
I am confident that there will continue to be audible improvements in the audio chain in the coming years.
I never said or intended otherwise.
se
But getting back to the point, how exactly do you know just how nice the renderings of strings and woodwinds are on the recording itself since you can only experience the recording through a rather complex chain of elements. Perhaps the Threshold is really giving you a more accurate representation of the recording than the ARC but the nonlinearities of the ARC produce an end result that you find more pleasing.Fair question. The answer is that I have half a dozen or so recordings of the ASO, a symphony I frequent. One of which I was there during the two day recording sessions. It was for the Firebird and the Polovetsian Dances. There is a beautiful clarinet solo in the Dances that was performed several times in succession. Shaw was merciless on the poor musician! Record a take and go downstairs to monitor the master Soundstream digital tape. Critique the technique and go again. The final take is masterful. Coincidently, the sessions were monitored using an earlier version of my Threshold amp driving ADS speakers. The sound was somewhat sterile and not true to what my ears remembered hearing ten minutes before. I have subsequently heard the recording more faithfully reproduced to the performance using different gear.
I submit that for live unamplified instruments, there are certain sonic signatures that immediately communicate to the ear/brain mechanism "LIVE". I would only hope that there are audio designers/researchers who truly love music and are able to translate what they hear to what they measure.
What measurements, for example, do you find directly equate to the ability of a component to reproduce natural timbres of brass or woodwinds?The spectrum of distortion products, ie., the distribution and magnitude of odd and even harmonics under various stimuli, directly equate to this ability, and can be easily measured.
How do you explain the fact that there are large numbers of excellent high end products designed by talented engineers that all sound different?
Because the engineers themselves are different, with different taste and different design priorities, who balance performance and cost compromises differently.
Are some/many/all just stupid and don't understand the test gear? I think not.
Why not? You don't think it's possible that many high end amplifier designers simply lift topologies out of major semiconductor manufacturer design manuals, substitute expensive parts and place them in a nice chassis? I think you are giving too much credit to the supposed "talent" of many high end designers. Please note that I am not claiming that the resultant product does not have high quality and performance. It's just that it may not have taken much talent at all to "design" if the technology is a simple variation on public domain technology. So yes, it is possible that at least some/many don't understand the test gear, or what to test for, for that matter.
Biggleswurth
Though I do not disagree with any of the fine points you raised, here is one fact I find very interesting. There are a handful of hideously expensive cost-no-object "no compromise" amplifier designs from the worlds best designers that all sound different on a very high resolution system. Each seems to have certain strengths. Does that mean that there is no way one can combine all the "known" methods in one box? Or perhaps does it suggest that there are as yet some secrets to be revealed? I hold the optimistic view that we are still learning.
It really depends on what you mean by "no compromise." A super expensive tube amp might be considered "no compromise" because it employs some rare, Soviet NOS tubes, silver wire wound output transformers, and regulated DC for its filaments. A super expensive "no compromise" transistor may use some new fangled FET devices, a switched power supply, class A/B operation and DSP feed forward correction. These two amps will have dramatic differences both measurable and sonically, yet both are "no compromise."Have you considered the possibility that many, if not most, people (including me) do not care for the sonic result when their favorite pet recording is very, very faithfully played back? Or that a really low distortion, high performance amplifier may simply emasculate mediocre speakers? There are innumerable, fickle, subjectively driven elements that define a desired audiophile system result that have nothing to do with a lack of knowledge of the technology needed for accurate audio reproduction.
Biggleswurth
Interesting.So none of us really wants the sonic truth. Designers can easily achieve said goal but know that deep down, we really don't want accuracy after all. The constant march in improvements over the years is due to improved ways to mask the inherent limitations of the reproduction chain.
I find that position axiomatically contrary to my objective for audio equipment to reproduce the musical experience. Do I have some poor sounding recordings? You bet! Some I still cherish due to the musical quality. Others are retired.
So none of us really wants the sonic truth. Designers can easily achieve said goal but know that deep down, we really don't want accuracy after all. The constant march in improvements over the years is due to improved ways to mask the inherent limitations of the reproduction chain.No, that's not what I stated. One point I was trying to make is that many coveted "reference" recordings are not very good, and in order to serve the subjective, sentimental expectations of an enthusiast using such recordings, it must sound "good" to that person. So, we gravitate towards the sound quality that we prefer. We may try to be objective, thinking that our memories of our impressions of the live event are accurate, and we may try to intellectually govern our assessment of reproduced sound, but ultimately we usually want what makes our favorite recordings sound good. This seriously complicates things, since A.) There are few actual "reference" grade recordings that can be ruled out as a variable in the accuracy of the playback chain, and B.) We usually prefer enhancement/exaggeration of that playback anyway. So, in order to meet subjective expectations and wants, an enthusiast will use whatever combinations of gear that yield a "good sound." I truly believe that the weakest link is by far and away the loudspeaker, and my example was intended to show that highly technically flawed speakers may only sound good with technically inferior amplifiers. Such amplifiers then receive great praise via authoritative statements from the user about how realistic or accurate they sound, when instead the user should be saying that the result simply sounds good to them.
But, to address your last inferred claim: I think many endorse an ideal about achieving "sonic truth," but when it comes to the typical listening room and gear, most simply want what sounds good to them, and unfortunately usually equate this with sonic truth. My experience is that what people often hear and like with any range of acoustic recordings has very little to do with sonic truth. This can be illustrated by evaluating what kind of sound quality people prefer with non-acoustic recordings. If you use those as your reference, you will find that there is absolutely no consensus on what is the preferred or "best" sounding speaker for such recordings. But the fascinating point is that for a given person, the speaker on which he/she prefers the electronic recording is often the same speaker that is preferred for acoustic music! For me, this is very strong evidence that people simply want sound quality aspects that they prefer, not what is accurate.
I think it is much too complicated for designers to intentionally design for some kind of preferred inaccuracy. The "constant march" for improvements is driven primarily by the use of trickle down materials from the latest greatest electronic parts quality advancements and material science advancements and industry standard practices. And I hope you understand that product merit has very little to do with its success in the market place.
I find that position axiomatically contrary to my objective for audio equipment to reproduce the musical experience. Do I have some poor sounding recordings? You bet! Some I still cherish due to the musical quality. Others are retired.
Your sentiments not withstanding, your reference to the "musical experience" clearly shows that you pursue the result that pleases you. This is far removed from being integrally related to extant technical knowledge that can be used to design highly accurate playback systems.
Biggleswurth
One point I was trying to make is that many coveted "reference" recordings are not very good, and in order to serve the subjective, sentimental expectations of an enthusiast using such recordings, it must sound "good" to that person.Fair enough. I am the first to admit that the majority of recordings have significant flaws. Each to his own. I have two categories of reference recordings: one for musical content, and the other for sonic content. My favorite rendition of the Holst Planets is an old RCA Ormandy LP that has gotten quite noisy over the years. Kinda thin sounding and imaging is typical mult-mike fare. Nevertheless, I still cherish it. Two "live" sonic references are the Reference Recording of Symphonie Fantastique with Keith Johnson at the helm. Remarkable in every way. For a change of pace, I have a number of Windham Hill "new age" recordings. Many of them are gems. Liz Story's Solid Colors is a wonderful solo piano recording. Recorded direct to two track on a 30 ips Studer with no compression. Very natural and dynamic.
I truly believe that the weakest link is by far and away the loudspeaker, and my example was intended to show that highly technically flawed speakers may only sound good with technically inferior amplifiers.
I agree with the first point entirely. Transducers are far more challenging that amplification to get right. On your other point, however, I don't share that sentiment. In my office, I have a pair of inexpensive Polk RT-55 speakers. Cost $300 at Circuit City for the pair. While hardly SOTA, they are not at all boxy sounding (one of my subjective biases). The source is a tuner driving a Threshold class A amp directly. I provide this detail because it is not typical to use a $2000 amp (1981 cost) with such equipment. I believe that they are capable of even better sound given the chance. Unless you may consider the Threshold "technically inferior", I think that the Polks are thinking "wow, listen to me now with this amp!"
My experience is that what people often hear and like with any range of acoustic recordings has very little to do with sonic truth. This can be illustrated by evaluating what kind of sound quality people prefer with non-acoustic recordings...
You got me here. Hey, I represent that remark!
I should have qualified my statement to that of unamplified music as I did with one of my responses to Mr. Eddy. What's the real sound of a Madonna recording? Do I want the PA sound quality experienced live? Not really. I find that I hear more "sonic truth" with my acoustically tuned system than with a bank of Crowns and Marshalls driving massive horns. I chose Madonna because I have some or her very nice 45 RPM "dance singles" with lots of low end and dynamics absent from the standard recordings fare. I find it ironic that there are no sonic equivalent CD versions of these available. Your sentiments not withstanding, your reference to the "musical experience" clearly shows that you pursue the result that pleases you. This is far removed from being integrally related to extant technical knowledge that can be used to design highly accurate playback systems.
I am perplexed by your closing comments. Not so much the first part because I confess that what pleases me most is hearing musicians live without mikes, mixing boards, cables, amps, and speakers. I am not elitist - I enjoy DJ Rap along with Debussy. I would hope, however, that in order to "design highly accurate playback systems" that you would necessarily have to employ "real live music" as the basis for doing so.
Returning to the original question, I believe that psychological studies as to an individual's mood, etc or to the hearing acuity of the public at large are not very valuable to the design process of accurate sound systems.
rw
RW,Your post inspired me to write a very lengthy reply, which I may just post as an original to the General asylum. Stand by.
Biggleswurth
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
