Objectivist or Subjectivist? Give Me a Break

[Phelonious Ponk]

Big thanks! I am happy enough if someone understands me. It is clear that we can not communicate properly - we diverge in the very basics and the signal to noise ratio of this thread is very poor.

The noise floor of this thread is high. Answer this question with substance and you'll pull the signal way up:

If parts and components can measure exactly the same and sound completely different, how do manufacturers produce a consistent product?

Tim

 

[Gregadd]

Tim[/QUOTE]

If parts and components can measure exactly the same and sound completely different

,

how do manufacturers produce a consistent product?

I read that as a hypothetical not as an assertion of fact. The author can clear up what they meant.

 

[KeithR]

Do two of the same model Krell sound quite different? Do some sound like Levinsons, or Spectrals? If not, without sound predictable by measurable, controllable specifications, how does Krell achieve consistency? Do they listen to every one and compare it to a control? If a run of caps sounds different than the last bunch, or a fuse is reversed, do they flip fuses, change caps? Without reliable measurements and standards for the amps and their component parts, how do they even know what changes to make? Trial an error? By ear?

Tim

Some audio gear manufacturers listen to every piece before leaving the factory (CAT comes to mind as I recall). more often than not, audio designers listen to every component while creating a design--in order to choose every resistor, material, cap, etc.--in conjunction with measurements

But you missed my point---read reviews for Levinson, Krell, and Spectral (if they existed) and you likely would think all three designs measure "wonderfully" although subtlety different of course.

Which one is best? Only your subjective ears can make that determination. IMO.

 

[KeithR]

Another more real world example is Halcro (now defunct)--- they had vanishing levels of distortion compared to every single amp on the market. That was the entire marketing pitch.

So why did they go bk? Did customers not agree they sounded the best despite their measurements? Shouldn't all of us owned Halcro because it measured so superb and better than anything out there?

Fact is I know a ton of folks that didn't like the sound (and I almost bought a DM38, so not a Halcro basher at all).

 

[Gregadd]

Yeah Keith. Since it measured so well the high end press should of hated it.

 

[rbbert]

...Shouldn't all of us owned Halcro because it measured so superb and better than anything out there?

Fact is I know a ton of folks that didn't like the sound (and I almost bought a DM38, so not a Halcro basher at all).

I think far more people liked them than not, but their price limited their sales potential; OTOH, maybe they didn't do well because the price was too low (i.e., not enough unit profit?)

 

[Phelonious Ponk]

Some audio gear manufacturers listen to every piece before leaving the factory (CAT comes to mind as I recall).

I'm not doubting your word, Keith, but I may be doubting theirs. For this to be practical, they would have to be a very small operation building components one at a time. And even then, if what is being proposed -- that parts of the same value can sound completely different, that components of exactly the same measurements can sound completely different, such testing would be meaningless unless they were to listen very carefully against a control and be willing to send an errant component back for testing and revision. And what would they test? Exactly the same measurements can sound completely different? They'd have to test everything by ear.

more often than not, audio designers listen to every component while creating a design--in order to choose every resistor, material, cap, etc.--in conjunction with measurements

Yes, and they measure those components, or trust the measurements that come from the manufacturer, then they specify a value for that component within the design. Unless things that measure exactly the same can sound completely different, in which case all of the above would be a waste of time.

But you missed my point---read reviews for Levinson, Krell, and Spectral (if they existed) and you likely would think all three designs measure "wonderfully" although subtlety different of course.

I didn't miss your point at all. I'm just following a logical path to a logical conclusion. If things can measure exactly the same and sound completely different, Krell and Levinson and Spectral will never quite know what they're getting, they won't be able to develop a differentiated sound. They certainly wouldn't be able to control it and manufacture it consistently. And of course that is not the case. Either one of two things (or both) are going on 1) The manufacturers are working with measurements they're not publishing, or 2) These things that measure wonderfully but subtley different also sound subtley different. Not completely different. The third possibility is, of course, that they actually sound the same. But I suspect there are subtle differences.

Tim

 

[microstrip]

The noise floor of this thread is high. Answer this question with substance and you'll pull the signal way up:
Tim

Tim,

You are a master in never answering direct questions, sometimes reversing them a few posts later, with the happy exception of when recommending good music. But this time the answer to your question is not in electronics or audio but any basic book on Logic or in wikipedia entry syllogism :

http://en.wikipedia.org/wiki/Syllogism

I would answer something like not all parts in all circuits sound different. Would you be pleased with it?

Component selection is a must in audio design. Even professional / consumer brands we both respect and enjoy, such as JBL and Dynaudio claim they select their film capacitors for a certain sound - do you think they do not use cheap polyester because they feel it is too cheap? And we are speaking about high level signals - low level signals in electronics are much more sensitive.

If you were a DIY you would know that you can buy kits that are exact electronic copies of famous known circuits manufactured in China at eBay. A friend of mine has assembled some of them, and they sounded terrible. Also some manufacturers who have their circuits assembled in China suffered from the same problem - cheaper resistors and capacitors were used in the amplifiers and they sounded inferior. All of them refer that this happened to some one they know well, but never to them!

Do you know the classic papers on capacitor sound of Jung and Curl?

 

[mep]

This sentence is a red herring: If parts and components can measure exactly the same and sound completely different, how do manufacturers produce a consistent product?

First of all, parts and components never measure exactly the same. The more expensive the gear, the tighter parts are matched. But, they are never ‘perfectly’ matched or ‘exactly’ matched. Matching transistors is expensive and time consuming depending on how tight of a match the designer is going for. Even if you buy 1% resistors, that’s plus or minus one percent. If you have one resistor in the right channel that’s minus one percent from the stated value and the one in the left channel is plus one percent from the stated value, now you have a two percent difference. How many companies that buy 1% resistors are taking the time and expense to make the match tighter? This is just one example. Capacitors are the same as resistors except it is common for capacitors to be rated at 10% tolerance from the stated value with some rated at 20%. If you are buying 10% tolerance caps, you will have to buy and test lots of caps to get them close, let alone ‘perfect.’

Another misuse of the word “exactly” in this thread was the statement that two pieces of gear can measure “exactly” the same and yet sound different. And the context of this statement was that amplifiers made by different companies that measure “exactly” the same sound different. Again, they don’t measure “exactly” the same. They may have close to the same power output, close to the same frequency response, close to the same current delivery capability, close to the same THD, close to the same input and output impedance etc, but they are never going to be “exactly” the same. One or more of those parameters will be much different. The topologies of the amps will probably be different, the parts will be different, and the skill of the designers in laying out their circuit boards will be different and thus the board layouts will be different. And it goes on and on.

To answer Tim’s question of how do manufacturers produce a consistent product, the answer is fairly simple. You have to have mature production practices. You have to use the same parts from the same vendors and match all parts the same for each piece of gear. And given the speed of obsolescence, this might only hold true for the first production run unless the manufacturer lays in huge parts quantities for anticipated future builds. And don’t think that manufacturers don’t make all types of changes from one production run to another because of parts changes. Just because there is no “MKI” or “MKII” or “MKIII” designation after the model number doesn't mean that significant changes occurred from different production runs.

 

[Phelonious Ponk]

Tim,

You are a master in never answering direct questions, sometimes reversing them a few posts later, with the happy exception of when recommending good music. But this time the answer to your question is not in electronics or audio but any basic book on Logic or in wikipedia entry syllogism :

http://en.wikipedia.org/wiki/Syllogism

I would answer something like not all parts in all circuits sound different. Would you be pleased with it?

Component selection is a must in audio design. Even professional / consumer brands we both respect and enjoy, such as JBL and Dynaudio claim they select their film capacitors for a certain sound - do you think they do not use cheap polyester because they feel it is too cheap? And we are speaking about high level signals - low level signals in electronics are much more sensitive.

If you were a DIY you would know that you can buy kits that are exact electronic copies of famous known circuits manufactured in China at eBay. A friend of mine has assembled some of them, and they sounded terrible. Also some manufacturers who have their circuits assembled in China suffered from the same problem - cheaper resistors and capacitors were used in the amplifiers and they sounded inferior. All of them refer that this happened to some one they know well, but never to them!

Do you know the classic papers on capacitor sound of Yung and Kurl?

The fact that you opened this post by accusing me of never answering direct questions is rich with irony, my friend. This is very simple: Two components can measure exactly the same and sound completely different. Your "suite of measurements" has not been defined by those who are making this claim (and it is their job to do so, I won't play that game), so the only safe assumption to make is that the statement applies to any and all measurements available to the designers and manufacturers of the components in question.

So how do they differentiate the sound of their product from others with the same measurements (by ear)? How do they manufacture that sound consistently? I understand that you don't have an answer, but that is the question.

Tim

 

[Phelonious Ponk]

This sentence is a red herring: If parts and components can measure exactly the same and sound completely different, how do manufacturers produce a consistent product?

First of all, parts and components never measure exactly the same.

Agreed (though good ones operate within tolerances that allow for consistency or audio would be total chaos). Then how can components made of those parts measure exactly the same and sound completely different? It's not my sentence, Mark. You think it is absurd? We're on the same page. I think the whole concept is absurd. When it appears to be true, you haven't measured properly or you have an odd notion of completely different.

Tim

 

[rbbert]

I can't exactly think of how a test such as I describe would be done, but...

A null test between two amplifiers playing music (or something comparable) into a real world load (i.e., significant impedance variation over a wide frequency range).

 

[rbbert]

...Do you know the classic papers on capacitor sound of Yung and Kurl?

It helps to spell their names correctly (Jung and Curl). I actually suspect that high-level (or at least high power) signals are more susceptible to effects of different capacitors and resistors, for example in speaker crossovers.

 

[microstrip]

The fact that you opened this post by accusing me of never answering direct questions is rich with irony, my friend. This is very simple: Two components can measure exactly the same and sound completely different. Your "suite of measurements" has not been defined by those who are making this claim (and it is their job to do so, I won't play that game), so the only safe assumption to make is that the statement applies to any and all measurements available to the designers and manufacturers of the components in question.

So how do they differentiate the sound of their product from others with the same measurements (by ear)? How do they manufacture that sound consistently? I understand that you don't have an answer, but that is the question.

Tim

Tim,

Your safe assumption puts you completely out of the game. To be brief, your answer shows you have never looked at a complete data sheet of component or never modeled a circuit in a modern electronics CAD program. No one can use all the measurements or specifications available - the complexity of the task would be enormous!

 

[microstrip]

It helps to spell their names correctly (Jung and Curl). I actually suspect that high-level (or at least high power) signals are more susceptible to effects of different capacitors and resistors, for example in speaker crossovers.

Thanks! Just changed them.

 

[Phelonious Ponk]

Tim,

Your safe assumption puts you completely out of the game. To be brief, your answer shows you have never looked at a complete data sheet of component or never modeled a circuit in a modern electronics CAD program. No one can use all the measurements or specifications available - the complexity of the task would be enormous!

My answer was deliberately absurd. But someone out there who is supposing that components can measure exactly the same and sound completely different needs to tell us what measurements we are not privy to that manufacturers are using to differentiate their products from others and mass produce them consistently. Because the guy who is turning out hundreds of amps that sound the same? He's not doing it by ear. He can measure what makes them sound that way and control it in the manufacturing process. If he couldn't, the ultimate absurdity suggested here would be absolutely true -- even two amps of the same manufacture, model and run would sound completely different. There would be no Krell sound, no Levinson sound. There would only be the sound of each amp. The ultimate in subjectivity.

Tim

 

[terryj]

Well, well well.

Firstly, let me start with a big apology.

I was a bit confounded-yet pleased- with the sudden switch, suddenly there ARE measurable differences which account for the perceived sound variations, a good point to reach BTW so we seem to have reached a better level of debate or at least a far greater base of agreement. In fact, if that is the new position then there is much less to disagree with (simply the little matter of audibility, that is one that will never be answered).

Anyways, here is the apology and my great **** up.

I misquoted the burmeister quote from micro! Only just found that out when I went waaay back in the thread. Just to clarify and show how I led it astray, here is what was quoted by micro Proof is given by the fact that it is possible to build two devices, which have exactly the same technical data but a completely different sound.. (I highlighted what it was that I inadvertantly changed, as you will see next)

I read that first thing in the morning and five pages later when I got to the end of that day I posted my response with the following error (from memory at that stage) IDENTICAL measurements give COMPLETELY DIFFERENT sound.

Ha, even worse is that as I typed that I did not even have the excuse of poor memory cause I actually quoted micros post (oh the shame) in my response so it was there to reread at the time (as I did just this second)...just goes to show beware of your assumptions. That that is one that has always bitten me on the bum in life it galls me that I made that error again.

Anyways, that is the apology and has had a lot to do with the curve the thread took for which I have to take a lot of responsibility.

I am not completely responsible thank heavens as even tho I started it with that error the baton was competently taken up by micro who indeed insisted that a dac chip from the same batch can sound completely different and hence from that I have managed to salvage a little bit of pride from my debacle!

That still leaves us with the idea that measurements-specifically manufacturers quality control-are insufficient to ensure consistency of sound across the board. All the implications that stems from that still apply such as a dac chip from the same batch sounding as different as two highend amplifiers supposedly do.

Is that the thought of all or is it a minority opinion. It looks increasingly like a minority one as the tenor seems to have changed since I went to bed. It now seems that measurements are likely able to show differences between gear, which in a very roundabout way (as long as I am not again misquoting five pages of posts from memory!) brings us back to a conundrum...when we started it was 'measurements don't tell us all and cannot predict or explain sound', then due to my error-which seems to have at least flushed a few birds into the air-it turned out that in order to extricate ourselves from that mess we find that measurements DO show/explain perceived sonic differences (we need that to get away from the logical fallacy that things which measure exactly sound different ya see) yet I doubt we would now agree that measurements show and explain sonics, we only needed that prop temporarily I'd say.

So it seems it is a moveable stance, when needed (to dig ourselves out of a hole) we can embrace measurements yet then revert no doubt to the default that measurements don't explain anything (a reminder that was the impetus to start the thread, Jeff disagreed with the basic 'damn measurements' stance of the article writer....does this mean BTW that the majority of the posters here now agree with Jeff that the article was hopelessly flawed? that we do indeed need measurements? Good, now go back and find more of the many strawmen she conjured)

Would it get me off the hook if I pretended I made my error deliberately so we could flush those birds out??? It would? Beauty. So as you can now see, I had a cunning plan to get subjectivists to agree with us on the value and utility of measurements!

 

[rbbert]

It's not all that unusual for review samples of different "identical" components to appear to sound different to the reviewers. In most (almost all?) cases, measurements actually show that at least one sample (sometimes two) is "defective". Nevertheless, even granting that review samples are more likely to have been damaged somewhere along the way to the current reviewer's venue, such defects may in fact be present in anyone's component, not necessarily appreciated by the listener as "defective". Just food for thought.

 

[Robh3606]

Do you know the classic papers on capacitor sound of Jung and Curl?

I posted a link to the Jung paper back at 346

http://www.reliablecapacitors.com/pickcap.htm

Rob

 

[Phelonious Ponk]

Don't be too hard on yourself, Terry. Burmeister may not have said that two components can measure the same and sound completely different, or that measurements are not useful in predicting sound, but these are core Audiophile positions.

And speaking of positions, did anyone actually change theirs or did I just get louder while you were sleeping? Sorry if I woke you...

Tim