Audiophiles Who Don't Trust Their Ears...

[PeterA]

These Nelson Pass quotes get to the essence of what the hobby is about for many of us. The designer uses measurements as a basis for his designs and he listens to the result, going back and forth until he is satisfied with the result. Many customers seem also quite satisfied with the result. The listening panel through which the Pass designs are tested must be able to trust their ears or they would not bother which such extensive listening tests before releasing a new design.

 

[thedudeabides]

http://www.weiss.ch/assets/content/41/Can-You-Trust-Your-Ears.pdf

Maybe this article will get us back on track.

Highly unlikely. Look back at all the similar threads / discussions on this forum related to objective vs subjective opinions.

Normally they blow up, the thread is closed and nothing is resolved.

 

[amirm]

Are you saying that because Nelson Pass's opinions conflict with yours that he has no 'authority' on this subject?

It does? Why do you say that?

Also, other than your post, and this post, I didn't see the word authority used anywhere from the post (350) that first referred to Nelson Pass.

Come again?

 

[Groucho]

I read these interviews and articles by celebrated designers and I just think "Yes, I would say exactly that, too". It's like a formula: flatter the customer's discernment and taste but imply there is hidden rigour in the design process. If your product doesn't measure as well as your rivals' then you'd better make sure you have that covered with an implication that there is a trade-off between musicality and measurements: your rivals and their customers are philistines. It may even be true, but it doesn't have to be - no one can prove it either way.

And if you're a veteran designer who has done nothing but build one type of technology all your career, you can certainly spin a yarn about why you think the new rival technology doesn't work so well. Purchasers of such new fangled gizmos are philistines while your customers will always have good ears and good taste, of course.

 

[Fidach Lad]

http://www.weiss.ch/assets/content/41/Can-You-Trust-Your-Ears.pdf


Maybe this article will get us back on track. If we hear but can't measure it ,is it real or imagined?

So, there IS an absolute sound out there! But no one can hear it! I'm not sure who will be made happier by this incredible conclusion: Plato, Saint Augustine or William James, since reason, authority and auditory experience are useless? So, by default, audio engineers become our new revelators!

 

[anders]

I read these interviews and articles by celebrated designers and I just think "Yes, I would say exactly that, too". It's like a formula: flatter the customer's discernment and taste but imply there is hidden rigour in the design process. If your product doesn't measure as well as your rivals' then you'd better make sure you have that covered with an implication that there is a trade-off between musicality and measurements: your rivals and their customers are philistines. It may even be true, but it doesn't have to be - no one can prove it either way.

And if you're a veteran designer who has done nothing but build one type of technology all your career, you can certainly spin a yarn about why you think the new rival technology doesn't work so well. Purchasers of such new fangled gizmos are philistines while your customers will always have good ears and good taste, of course.

Well, yes and no... obviously he has to appeal to the ego of his customers. But if you ask him what his favorite sounding amps are they are the ones that a first time DIYer (with some online assistance) can put together. Having owned several Pass/FirstWatt amps, I am in complete agreement with him.

 

[PeterA]

I read these interviews and articles by celebrated designers and I just think "Yes, I would say exactly that, too". It's like a formula: flatter the customer's discernment and taste but imply there is hidden rigour in the design process. If your product doesn't measure as well as your rivals' then you'd better make sure you have that covered with an implication that there is a trade-off between musicality and measurements: your rivals and their customers are philistines. It may even be true, but it doesn't have to be - no one can prove it either way.

And if you're a veteran designer who has done nothing but build one type of technology all your career, you can certainly spin a yarn about why you think the new rival technology doesn't work so well. Purchasers of such new fangled gizmos are philistines while your customers will always have good ears and good taste, of course.

Wow, why so negative about such a highly regarded and successful amplifier designer? This is a small and, some would argue, challenged industry, so should we not be more positive and optimistic when discussing one of its most successful and respected brands? Amir writes that measurements are downplayed and you write that there is a hidden rigor implied. In the case of Pass Labs, are testing and measurements not discussed openly by the designer himself?

These quotes by Nelson Pass are great examples for a thread on the subject of trusting one's ears. He discusses that he uses measurements and he also listens. The combination is how he arrives at his designs. The designs by all measures - patents, sales, reputation, and sound - are wildly successful.

If you want to challenge the measurements and efficacy of Pass Labs' Class A topology, should you not take this discussion to the Science Forum? Here, we are trying to discuss whether or not ears can be trusted and the subjective act of listening impressions.

 

[microstrip]

(...) The message here is not that measurements are not useful or critical to design of amplifiers. They are and he spent 99% of the article talking about them. What he says at the end is a political backtrack based on faulty logic. Market success has nothing to do with best audio performance. Or else we could blame every high-end gear including Pass as a failure because they too have not dominated the market.

I (and I think almost anyone here) do not pretend that measurements are not useful or critical to design of amplifiers. (I quote my post #375) : The measurements are tools for electronic design, as well as electronic simulators and CAD facilities. Most subjectivists accept measurements - what we usually question is just the correlation with sound quality and the use of measurements as the best sound quality evaluator, with the capability of rating and even "eliminating" products.

In high-end, domination of the market is not the target. But yes, some designers are more successful than others. IMHO fortunately Nelson Pass is one of them and it seems me he trusts his and others ears. IMMV.

BTW, I have been reading Nelson Pass in DIY forums for many years. I even build a small one FET amplifier long ago. But never measured it.

 

[microstrip]

I read these interviews and articles by celebrated designers and I just think "Yes, I would say exactly that, too". It's like a formula: flatter the customer's discernment and taste but imply there is hidden rigour in the design process. If your product doesn't measure as well as your rivals' then you'd better make sure you have that covered with an implication that there is a trade-off between musicality and measurements: your rivals and their customers are philistines. It may even be true, but it doesn't have to be - no one can prove it either way.

And if you're a veteran designer who has done nothing but build one type of technology all your career, you can certainly spin a yarn about why you think the new rival technology doesn't work so well. Purchasers of such new fangled gizmos are philistines while your customers will always have good ears and good taste, of course.

Fortunately most people who trust their ears and have listened to SOTA sound reproduction using products designed by these celebrated designers have all the reasons to think differently. Some of them even created a forum called WBF to debate matters that include such excellent products and related subjects. I respect your opinion but I feel you should get information about Pass career and designs - he has used many types of circuits and technologies.

 

[amirm]

These quotes by Nelson Pass are great examples for a thread on the subject of trusting one's ears. He discusses that he uses measurements and he also listens. The combination is how he arrives at his designs.

This is how *every* amplifier designer does his work. Why do you think that is unique? If I were to design an amplifier I would do exactly what he does. The exception is that I have far more awareness in how listening tests can lead to improper conclusion especially when it comes to one's own product. Example:

I am working at Microsoft and our goal is to better our competitor's audio compression. That is, can we get the same quality at half the bit rate (or something like that). Remember, I was the "golden ear" of the company and could do no wrong at the time when it came to listening tests. I created a bunch of tests and they all "proved" that we had accomplished the goal. I proudly mention that to our marketing department and they say that we need an independent test party to duplicate the results and then we could make a lot of hay from it.

So we hire a third-party company at great expense (thousands of dollars), they recruit 100 or so listeners from general public. They review their test set up with me and all is good. Then the testing starts. Next morning at 6:00am I get this worried call from our marketing person saying something is seriously wrong and I better to talk to the test company (they were ahead of us in time zone). I get on the horn and the two guys running the tests say that the "results are not what you expected. Do you want us to continue?" I ask what that means. He said the listeners overwhelming voting that the competitor's codec is sounding better than us.

I was furious. I thought they had screwed something up. I ask them for exact copies of the files and they sent them to me. I listen and I cannot at all verify their results. No way was their codec better sounding (even though it had higher bit rate). But just before calling them and our marketing department back, I ask my wife and son to come and listen without telling them what was going on. Just played one file and then another. Darn it if they both did not say the other codec was better -- just like the test results.

With that outcome in front of me, I listen again, this time attempting to put aside all of my biases, I could now hear the problem. The testing company and my family were completely right. I was blinded by something being ours and therefore being better, all of my skills as a trained listener be damned.

Even though I have been an analog designer for years and started my electronics hobby and degree in analog design, I don't hold a candle to Nelson in designing and building an amplifier for commercial sales. But likewise, he doesn't come remotely close to my experience in running hundreds of listening tests, blind and sighted to know all the pitfalls of looking yourself in the mirror and trying to judge if you are pretty . I don't know all that I need to know to design an amplifier like he does. And he doesn't now all that he needs to know about faults of casual listening tests.

The designs by all measures - patents, sales, reputation, and sound - are wildly successful.

All true. Missing in that are knowledge of psychoacoustics, and some kind of proof point we can use to know that when he decides on circuit A versus circuit B based on his listening test, that whatever conclusion he arrived at is correct.

In the above story I told, such an objective proof existed. Many other people with no stake in the test differed from my opinion of what sounded right. Had we gone to the market with my personal claims we would have had a big egg on our face. Unless you all want to show me some data on this front, I say his skills as a listener, is no better or worse than anyone else. Which means it can and likely is highly faulty. What a relief it is then to see him use measurements extensively for his design and not "trusting his ears" for vast majority of design decisions he makes.

If you want to challenge the measurements and efficacy of Pass Labs' Class A topology, should you not take this discussion to the Science Forum? Here, we are trying to discuss whether or not ears can be trusted and the subjective act of listening impressions.

He is not at all on stand, nor are we discussing the specifics of any measurements. You all put him forward as an authority that designs equipment and "trusts his ears." I am explaining that you have arrived at completely wrong conclusion. That he doesn't trust his ears like some of you do. You buy an amplifier without looking at its measurements. He makes countless measurements before finishing the design. Don't equate yourself with a designer or your case winds up with far more holes than it may have now.

And once again, as a designer, Nelson is not an authority on validity of listening tests in amplifier evaluations. That is not what he is famous for and attempting to position him as such caused me to say if you are going to appeal to an authority, it better be right. In this case it is not. I know what he is doing and he is not remotely your friend in this argument. Find a famous designer that has not post measurement after measurement in his work and maybe you have a case then. Not here.

 

[Mike Lavigne]

Amir,

I would prefer subjective listening feedback from known to me 'serious listeners' to high end audio or known to me 'serious designers' of high end audio compared (with all due respect) to you, your family, or these hundred random people. I've had enough 'random people' in my listening room to recognize a 'deer in the headlights' that have no clue what they might be listening to or how something should sound. it takes time to 'get it'. there is a learning curve.

when I visit a new system to me it takes me time to take it all in and have meaningful feedback. and I seriously listen for 30+ hours a week.....for the last 20 years.

I'm interested in listening opinions from people who have gone down the road I have gone down. who have been exposed to what I have been exposed to. done the work, put in the time, investigated the questions.

all the rest are less significant data points. not of no value, but never very important data points.

this is just for me, YMMV, everyone can decide for themselves where the most significant feedback might be found.

 

[Yuri Korzunov]

Amir,

I would prefer subjective listening feedback from known to me 'serious listeners' to high end audio or known to me 'serious designers' of high end audio compared (with all due respect) to you, your family, or these hundred random people. I've had enough 'random people' in my listening room to recognize a 'deer in the headlights' that have no clue what they might be listening to or how something should sound. it takes time to 'get it'. there is a learning curve.

I'm interested in listening opinions from people who have gone down the road I have gone down. who have been exposed to what I have been exposed to. done the work, put in the time, investigated the questions.

all the rest are less significant data points. not of no value, but never very important data points.

this is just for me, YMMV, everyone can decide for themselves where the most significant feedback might be found.

"Serious listeners" and "serious designers" it's 2 approaches to single problem.

Main "objective" issue is threshold between "what we can measure" and "what we can listen".

Even for man/woman with non-trained hearing, measurements give ability distinguish subtlest details.

 

[Bill Hart]

Amir,

I would prefer subjective listening feedback from known to me 'serious listeners' to high end audio or known to me 'serious designers' of high end audio compared (with all due respect) to you, your family, or these hundred random people. I've had enough 'random people' in my listening room to recognize a 'deer in the headlights' that have no clue what they might be listening to or how something should sound. it takes time to 'get it'. there is a learning curve.

when I visit a new system to me it takes me time to take it all in and have meaningful feedback. and I seriously listen for 30+ hours a week.....for the last 20 years.

I'm interested in listening opinions from people who have gone down the road I have gone down. who have been exposed to what I have been exposed to. done the work, put in the time, investigated the questions.

all the rest are less significant data points. not of no value, but never very important data points.

this is just for me, YMMV, everyone can decide for themselves where the most significant feedback might be found.

Mike: I just referred to you on another board, where people were talking about 'dream systems' and the degree to which it is the gear and the room. I used you as an example of someone who has gone the distance on uber gear, paid the money and went to the length of a dedicated building, and you have still had to spend time over the years changing parameters, fine tuning, etc.
Although I didn't mention it my post on the other board, my experience in the car world taught me much the same; having had a lot of those uber exotics over the years, there's nothing like a well-set up Porsche. But, I think you can only say that after you have been down the road. Me, my system is in fine fettle for what it is, strengths and weaknesses such as they are. I'm far happier spending time on all the records I accumulated over the years, learning about how they were made, and why some of them have the 'magic.' My conclusion so far: a happy coincidence of artistry and engineering.

 

[amirm]

Amir,

I would prefer subjective listening feedback from known to me 'serious listeners' to high end audio or known to me 'serious designers' of high end audio compared (with all due respect) to you, your family, or these hundred random people. I've had enough 'random people' in my listening room to recognize a 'deer in the headlights' that have no clue what they might be listening to or how something should sound. it takes time to 'get it'. there is a learning curve.

That is an approach you can take as a listener Mike. If you arrive at the wrong conclusion, oh well, it is just one of you wasting money on the wrong gear.

When you *design* a system as my team did and Nelson does for his customers, it is an entirely different matter. Vast majority of customers for such high-end system are not, let me repeat, are not audiophiles. It is wealthy independent people who just want a great system and price is no object. I once asked one of the most respected DAC companies where they sold their most products and he said Hong Kong! New York was second and the rest did not count!!! Yet we have discussed their products frequently in this forum and elsewhere. So it is critical that audio products be designed to sound good to everyone, including family members.

By the same token, the technology out of my team was part of windows and could be use by anyone. Decisions we made in its design therefore had to be *durable.* It had to sound good to audiophiles and non-audiophiles. On that topic, the person that finally showed me I should pack my bags and no longer do listening test was a non-audiophile. He could hear artifacts that I could not hear and we hired him and that was that. Audiophiles we tested sadly could do no better than non-audiophiles in many listening tests we performed on them. As you say, you need to learn to listen and scant few audiophiles have been in controlled situations to learn non-linear distortions and be able to spot them.

Back to Nelson, it is imperative that you confirm what you ears tell you with measurements. Who knows if Nelson has the ears that you have or ones made out of tin? We, as customers for a brand, need to demand that something other than his hearing which has not been verified to be as good as ours be used to create audio products. Otherwise, it is a crap shoot of whether we have just as good or broken ear as he has.

I should also note that I absolutely believe in proper listening tests for comparing amplifiers. What makes them proper would be bias controlled of course (no look at your pretty system and room ) but also include extensive training to know what amplifier artifacts sound like. I would also test amplifiers by comparing their input to their output. That is all that matters. Do they act like "wire with gain?" If they are coloring the sound, then they are not going to get my business. That is what speakers for if you want coloration .

 

[PeterA]

This is how *every* amplifier designer does his work. Why do you think that is unique? If I were to design an amplifier I would do exactly what he does. The exception is that I have far more awareness in how listening tests can lead to improper conclusion especially when it comes to one's own product. Example:

I am working at Microsoft and our goal is to better our competitor's audio compression. That is, can we get the same quality at half the bit rate (or something like that). Remember, I was the "golden ear" of the company and could do no wrong at the time when it came to listening tests. I created a bunch of tests and they all "proved" that we had accomplished the goal. I proudly mention that to our marketing department and they say that we need an independent test party to duplicate the results and then we could make a lot of hay from it.

So we hire a third-party company at great expense (thousands of dollars), they recruit 100 or so listeners from general public. They review their test set up with me and all is good. Then the testing starts. Next morning at 6:00am I get this worried call from our marketing person saying something is seriously wrong and I better to talk to the test company (they were ahead of us in time zone). I get on the horn and the two guys running the tests say that the "results are not what you expected. Do you want us to continue?" I ask what that means. He said the listeners overwhelming voting that the competitor's codec is sounding better than us.

I was furious. I thought they had screwed something up. I ask them for exact copies of the files and they sent them to me. I listen and I cannot at all verify their results. No way was their codec better sounding (even though it had higher bit rate). But just before calling them and our marketing department back, I ask my wife and son to come and listen without telling them what was going on. Just played one file and then another. Darn it if they both did not say the other codec was better -- just like the test results.

With that outcome in front of me, I listen again, this time attempting to put aside all of my biases, I could now hear the problem. The testing company and my family were completely right. I was blinded by something being ours and therefore being better, all of my skills as a trained listener be damned.

Even though I have been an analog designer for years and started my electronics hobby and degree in analog design, I don't hold a candle to Nelson in designing and building an amplifier for commercial sales. But likewise, he doesn't come remotely close to my experience in running hundreds of listening tests, blind and sighted to know all the pitfalls of looking yourself in the mirror and trying to judge if you are pretty . I don't know all that I need to know to design an amplifier like he does. And he doesn't now all that he needs to know about faults of casual listening tests.


All true. Missing in that are knowledge of psychoacoustics, and some kind of proof point we can use to know that when he decides on circuit A versus circuit B based on his listening test, that whatever conclusion he arrived at is correct.

In the above story I told, such an objective proof existed. Many other people with no stake in the test differed from my opinion of what sounded right. Had we gone to the market with my personal claims we would have had a big egg on our face. Unless you all want to show me some data on this front, I say his skills as a listener, is no better or worse than anyone else. Which means it can and likely is highly faulty. What a relief it is then to see him use measurements extensively for his design and not "trusting his ears" for vast majority of design decisions he makes.


He is not at all on stand, nor are we discussing the specifics of any measurements. You all put him forward as an authority that designs equipment and "trusts his ears." I am explaining that you have arrived at completely wrong conclusion. That he doesn't trust his ears like some of you do. You buy an amplifier without looking at its measurements. He makes countless measurements before finishing the design. Don't equate yourself with a designer or your case winds up with far more holes than it may have now.

And once again, as a designer, Nelson is not an authority on validity of listening tests in amplifier evaluations. That is not what he is famous for and attempting to position him as such caused me to say if you are going to appeal to an authority, it better be right. In this case it is not. I know what he is doing and he is not remotely your friend in this argument. Find a famous designer that has not post measurement after measurement in his work and maybe you have a case then. Not here.

Amir, thank you for your post. I will respond to specific statements that are highlighted in bold.

1. I never claimed Nelson Pass' method is unique. Where did you get that idea?
2. I don't know if Nelson Pass comes remotely close to your experience in running hundreds of listening tests. That is your assumption, but why does this matter in the least? No one but you is discussing this.
3. The conclusion he arrives at is his preference. I do not need proof. Customers are then free to evaluate his design and choose for themselves.
4. I don't think anyone is making any claims about the quality of Nelson Pass' skills as a listener. Why would we need to "show you some data on this front"? You are assuming it is likely "highly faulty".
5. I bought my first Pass amplifier, the Aleph 3, in part by looking at the measurements because they were included in the package which I saw before buying it and the individual test bench results for that particular amplifier were included in the manual and signed by Nelson Pass. This is another incorrect assumption.
6. I am not equating myself with a designer. That is another incorrect assumption.
7. I think the observation is that Nelson Pass is an authority on amplifier design. I do not think anyone has claimed that he is an authority on the "validity of listening tests in amplifier evaluations." You point out that "all" designers use a similar testing/measuring/listening methodology to develop their amplifier designs. I am not aware that anyone has claimed that any of them are authorities on listening tests in amplifier evaluations.

I read many assumptions in your post, and I don't understand the basis by which you make them. Is your work at Microsoft the basis of this?

Nelson Pass measures and listens back and forth. He then releases a product. It seems that this is how others designers do it also. Customers either like it or not, I presume based on listening and other criteria, and they then decide to buy it or not. I don't think there is anything controversial or unusual about any of this. We are fortunate that Nelson Pass is very willing to discuss and share his designs openly and freely on DIY discussion forums and in literature. Some may consider this brilliant marketing, but it comes across as genuine and I, for one, respect and appreciate his approach. There are certainly other wonderful designers who also are worthy of admiration and respect.

 

[amirm]

Nelson Pass measures and listens back and forth. He then releases a product.

Say that and I would give you high five as much as the next guy. It shows that he doesn't "trust his ears" fully or he would have no need for measurements. Instead we were given this:

"Appreciation of audio is a completely subjective human experience. Measurements can provide a measure of insight, but are no substitute for human judgment. Why are we looking to reduce a subjective experience to objective criteria anyway? The subtleties of music and audio reproduction are for those who appreciate it. Differentiation by numbers is for those who do not."


A number of people gave this high five saying implying that they are vindicated in trusting theirs because that is what Nelson does. Glad to see confirmation from you at least that this quote did not at all represent his design philosophy. Let's see that from others and we will be converging on one position.

Some may consider this brilliant marketing, but it comes across as genuine and I, for one, respect and appreciate his approach.

I am not at all concerned about marketing aspects of his work or Nelson himself or products. Just want to get *us* on the same page that there is no imaginary great designer that trusts his ears and ignores measurements as we do.

 

[DaveyF]

Back to Nelson, it is imperative that you confirm what you ears tell you with measurements. Who knows if Nelson has the ears that you have or ones made out of tin? We, as customers for a brand, need to demand that something other than his hearing which has not been verified to be as good as ours be used to create audio products. Otherwise, it is a crap shoot of whether we have just as good or broken ear as he has.

I should also note that I absolutely believe in proper listening tests for comparing amplifiers. What makes them proper would be bias controlled of course (no look at your pretty system and room ) but also include extensive training to know what amplifier artifacts sound like. I would also test amplifiers by comparing their input to their output. That is all that matters. Do they act like "wire with gain?" If they are coloring the sound, then they are not going to get my business. That is what speakers for if you want coloration .

+1.
I absolutely hope that any amp designer or for that matter any designer of high end audio does exactly that...listen first and then verify with measurements. What I would also hope, is that IF the measurements do not verify the results for the listening test; that said designer not scrap the design but seek out other listeners to try and verify the results. In the rare instance that the measurements are superior to the listening test, I would believe that more extensive listening and a larger listening sample be in order. A possible example of this were the old Halcro amps from the past; measured superbly, but upon listening...well!
Although, I imagine the designer thought they sounded great, but as great as the measurements?? Perhaps this is an example of the designer NOT having a great ear, and the measurements backing him up, or not???

 

[Gregadd]

I for one am not saying that ears are better than measurements. Indeed the two are incompatible. Measurements are merely arbitrary sytems developed by our sensory perceptions(In this case hearing)
Nelson Pass said if may paraphrase Nelson Pass that there is little to gained from adding more zeros to distortion figures . If we are to move forward listening has to be a vital part

 

[andromedaaudio]

Well , i would never buy a speaker designed by ear only

 

[Gregadd]

Well , i would never buy a speaker designed by ear only

How would you know?