[Steve Eddy]

I'll restate what I attempted to say earlier for clarification:

Thank you. Clarification is always a good thing.

It is possible to pick the difference between different musical instruments by ear with great reliability. It's easy to verify that we get it right here [unlike with say a capacitor] cos you can simply look at the physical object and there's no question as to what it is.

Yes.

You appear to be suggesting above [correct me if I'm wrong  ] that measuring equipment might be unable to pick such simple matters as the difference between instruments.

What I'm saying is that it's trivially easy to measure differences between two instruments. Identifying a particular instrument goes beyond measurement and on to interpretation. And what I don't understand is what is the particular relevance of identifying a particular instrument as it relates to the subject at hand?

I don't know if this is true or not and I don't know the limitations of measuring equipment. Dan is quite certain that it can be done and maybe he's right. Still, I don't know the limitations of the equipment here and "voice recognition" to name one tool he backs still seems to have a long way to go. How would it go picking a single instrument out of the midst of an orchestra for example?

Certainly voice recognition isn't perfect.

But again, I still don't see the point in this. You seem to be confusing the issue between measurement and interpretation. They're not one and the same.

The significance of the "what instrument is playing" issue seems to me to be that it's a good test in that there's no question that instruments are different and it offers a good base to compare the ability of the ear to pick a difference as against the ability of a piece of measuring gear to do the same [yeah, yeah .. it can't be "the same"  ] but you get my point?

No, I'm afraid I don't get your point.

As I see it, the point is measurement, not interpretation.

If one is able to measure everything upon which the interpretation is based on, what's the point of the interpretation itself?

se

[jules]

ah, well if you are saying that it is trivially easy to measure the difference between two different instruments, Then I have misunderstood the earlier statement I quoted.

As for your measurement/interpretation point. I hear what you're saying but I don't agree 

Not a bad discussion though ... thanks

jules

[Daygloworange]

Hey guys,

Loving this thread. Really.  

As far as the pre amp example that Danny gave, that's out of phase crosstalk causing the effect.

Very interesting. Sounds very plausible as to why one pre amp would have better soundstaging than another.

Just out of curiosity, does anybody make a mono bloc configuration pre-amp? You know, two separate units, so there's no possibility of crosstalk?

What I'm saying is that it's trivially easy to measure differences between two instruments. Identifying a particular instrument goes beyond measurement and on to interpretation. And what I don't understand is what is the particular relevance of identifying a particular instrument as it relates to the subject at hand?

Perhaps what he is getting at here Steve, is that humans can easily identify what instrument is producing a note, but electronic testing instruments cannot, or cannot as of now. Although I'm sure that you could build an FEA type of data base for sound analysis software that could, and then maybe if such a thing were powerful enough, that it might be able to shed more objective data that could reinforce or dispel a lot of the things that are true and aren't true in the field of sound reproduction.

I know that there have been great advancements in binaural and 4 way soundfield microphones that could perhaps be employed to record the sounds coming out of a pair of speakers A/B'ing different components( be it PC's , IC's, SC's, Pre-amps, CDP's, Amps) that could then be analyzed (on scopes, frequency analyzers, etc..) to perhaps quantify what subjectivists feel they are hearing. I feel that if these microphones could not pick it up, I'm afraid it's definitely not there. Microphones are much more sensitive that human ears.

Cheers

[WEEZ]

Dual mono Pre-amp?

www.canaryaudio.com/903.htm

How much money 'ya got?

[jules]

damn it dayglow .. I'd decided to end my contributions 

wot I originally thought Steve was saying was that it wasn't possible for measuring equipment to pick the difference between two different instruments.

What I now think he's saying is that it's "trivially easy" for measuring gear to pick the difference between say a flute and a recorder but that measuring gear can't pick the diff. between the sound of the 1st violin and the sound of the 3rd violin in an orchestra ... and, it doesn't matter.

?

jules

[Scott F.]

Just out of curiosity, does anybody make a mono bloc configuration pre-amp? You know, two separate units, so there's no possibility of crosstalk?

The Korato KVP 10 is a true dual mono pre with a separate power supply. Fully tubed gain stage with a tubed phono stage and tube rectification.

[Steve Eddy]

ah, well if you are saying that it is trivially easy to measure the difference between two different instruments, Then I have misunderstood the earlier statement I quoted.

Ah, ok.

As for your measurement/interpretation point. I hear what you're saying but I don't agree 

And as long as I have any say in the matter, you're perfectly free to do so. 

Not a bad discussion though ... thanks

Ditto! 

se

[Steve Eddy]

Perhaps what he is getting at here Steve, is that humans can easily identify what instrument is producing a note, but electronic testing instruments cannot, or cannot as of now.

Yes, I know what he's getting at. What I'm getting at is that I don't see what relevance that has.

This sidetrack essentially began with Ethan saying "Everything that matters with audio is well known and can be measured to orders of magnitude beyond what any human can hear."

The key word, and the context of Ethan's remark being "measured." Ethan was simply saying that any difference sufficient enough to produce an actual audible difference can be measured.

Then some tried arguing against that statement by invoking what I've been referring to as "interpretation," i.e. the "what instrument is playing and who's singing."

Perhaps I put too much emphasis on measurement and interpretation not being one and the same. The salient point is that even if measurement couldn't directly tell you "what instrument is playing and who's singing," that doesn't invalidate Ethan's statement.

And as a side note, a human couldn't tell you what instrument is playing and who's singing either unless they alrady have some familiarity with the instrument being played and the person singing. And I don't see any reason why this sort of interpretation couldn't be done by "machine" if enough measurements were taken to adequately quantify the various characteristics of various instruments and various singers such that an appropriate algorithm could produce accurate interpretations of what's fed to it.

se

[miklorsmith]

Man, you guys are cyborg argument mongers.

My final thoughts on the matter are this:  Whether Ethan's statement is true or not is irrelevant to any non-manufacturer.  The masses are concerned with getting better sound in-room.  These folks cannot or will not go through the scientific rigor needed to pursue such things whether or not absolutes are possible.

So, the engineers/builders should keep on chasing and the buyers shouldn't really care what business the former are up to.  If they can quantify every element of sound, great.  I'll be expecting my perfect preamp, perfect amp, perfect CDP, and perfect speakers, what, next week?  Sweet, thanks fellas.

[Steve Eddy]

Man, you guys are cyborg argument mongers.

I disagree. Wanna make something of it? 

se

[Dan Banquer]

Hey guys,

Loving this thread. Really.  

As far as the pre amp example that Danny gave, that's out of phase crosstalk causing the effect.

Very interesting. Sounds very plausible as to why one pre amp would have better soundstaging than another.

Just out of curiosity, does anybody make a mono bloc configuration pre-amp? You know, two separate units, so there's no possibility of crosstalk?

Actually I really think there is no need for dual mono pre amps. Idiots like myself  managed to get  approximately -100 db at 1 kHz and about - 70 db  at 20 Khz for crosstalk which I found slightly exceeded my threshold for audibility. BTW: If you think about it for a moment you can readily argue the case that high stereo separation inside these boxes is also an strong indicator of noise immunity.
              d.b.

[Daygloworange]

Perhaps what he is getting at here Steve, is that humans can easily identify what instrument is producing a note, but electronic testing instruments cannot, or cannot as of now.

Yes, I know what he's getting at. What I'm getting at is that I don't see what relevance that has.

Again, perhaps what he might be getting at, or at least, what I'm getting at , is that humans can hear between certain component swaps better imaging where they might be able to hear the 1st and 3rd violins more clearly in the soundstage, but would you be able to measure the differences with a piece of equipment commonly used to quantify electronic design.

Some components when A/B'd have a similar "overall sound" but differ in that some have a more forward presentation, while others will have a more laid back presentation. Some have better separation between individual instruments in the soundstage, while others have a soft aura around each player/instrument, as opposed to sharp delineations.

So the relevance would be that, are current electronic measuring instruments( or software algorithms) able to clearly plot, somehow, the factors involved in making these types of different sonic presentations. Could they be able to show how the localization cues are different between the those two types of sonic presentations .

Plenty of people report very similar findings to my examples to describe the differences they hear when A/B'ing components.

As it stands, this CLC gizmo is what I would consider a faith based talisman, relying on the holistic faith of the end user. That is very different than a tweek. It is not connected to the system in any way, shape or form than I can easily see. It's claimed effect is on the test subject. This is a paranormal study, not an audio one. The fact that this thing won an Positive Feedback award for audio is amazing. If it actually works, it should have been submitted for a Nobel prize, it would be a lot more useful in other uses than to such a small demographic as "audiophiles"

The inventers of this thing should probably have a better person in charge of their marketing...

Cheers

[Steve Eddy]

Again, perhaps what he might be getting at, or at least, what I'm getting at , is that humans can hear between certain component swaps better imaging where they might be able to hear the 1st and 3rd violins more clearly in the soundstage, but would you be able to measure the differences with a piece of equipment commonly used to quantify electronic design.

Some components when A/B'd have a similar "overall sound" but differ in that some have a more forward presentation, while others will have a more laid back presentation. Some have better separation between individual instruments in the soundstage, while others have a soft aura around each player/instrument, as opposed to sharp delineations.

So the relevance would be that, are current electronic measuring instruments( or software algorithms) able to clearly plot, somehow, the factors involved in making these types of different sonic presentations. Could they be able to show how the localization cues are different between the those two types of sonic presentations .

I can't think of any reason why this couldn't be done.

Plenty of people report very similar findings to my examples to describe the differences they hear when A/B'ing components.

Yes. But the first order of business would be to establish that what they're perceiving is actually due to the components. Because if it's not, then you could end up wasting a hell of a lot of time and effort trying to chase down a phantom.

As it stands, this CLC gizmo is what I would consider a faith based talisman, relying on the holistic faith of the end user. That is very different than a tweek. It is not connected to the system in any way, shape or form than I can easily see. It's claimed effect is on the test subject. This is a paranormal study, not an audio one. The fact that this thing won an Positive Feedback award for audio is amazing. If it actually works, it should have been submitted for a Nobel prize, it would be a lot more useful in other uses than to such a small demographic as "audiophiles"

The inventers of this thing should probably have a better person in charge of their marketing...

Would you care to volunteer? Who knows? With enough effort, you might be able to get to meet the King of Sweden someday. 

se

[Ethan Winer]

Ethan, are you saying the only two parameters of quality sound reproduction are frequency response and distortion?

There are four audio parameters that define everything needed: frequency response, noise, distortion, and time-based effects. There are also subsets, and I defined them all in detail in my article last year in Skeptic Magazine:

http://www.ethanwiner.com/audiophoolery.html

> Following this line of thought then, measurements should be able to predict sound quality. <

Affirmative!

--Ethan

[miklorsmith]

This seems to have morphed into a theoretical discussion of what might be possible to do.  If it were possible, somewhat cost effective, and useful, wouldn't it be happening in practice already?

I want to see you guys quantify all this stuff, really, and in practice.  You will revolutionize the science of design and leave all others in the dust.  You will make millions and millions and millions of dollars.  I will be your first customer.  Get off the web, into your design software, and get crackin' dammit!

[Ethan Winer]

John,

> Can anyone look at these measurements and tell me what instruments and people are on the recording? <

No, but that's irrelevant to this discussion! What you are asking about is basically artificial intelligence. What I'm talking about is assessing audio equipment to determine if it changes the sound passing through it. It doesn't matter if the singer is John, Paul, George, or Ringo. Measurements absolutely can tell if a piece of gear (or your room) is changing the sound in any way that is audible.

--Ethan

[Daygloworange]

Yes. But the first order of business would be to establish that what they're perceiving is actually due to the components. Because if it's not, then you could end up wasting a hell of a lot of time and effort trying to chase down a phantom.

Agreed. That's why I think if you were to use two (4 head) soundstage microphones and implement them properly(possibly in a binaural, or dummy head configuration) to record a pair of speakers A/B'ing various components, and were to do overlays of the 8 separate tracks of info of A/B comparisons, you could possibly see very easily if there are indeed audible differences.

 I believe this would be a comprehensive test, and quite repeatable. It would not cost millions of dollars either, or require a room full of scientists. It could be put together with even perhaps one knowledgeable person with access to these microphones and access to, and knowledge of the software that I'm sure currently exists to yield successful results with a test of this nature.

Would you care to volunteer? Who knows? With enough effort, you might be able to get to meet the King of Sweden someday.

King of Sweden huh? Does he have a daughter? I have a thing for Scandinavian women.  

Cheers

[rajacat]

John,

> Can anyone look at these measurements and tell me what instruments and people are on the recording? <

No, but that's irrelevant to this discussion! What you are asking about is basically artificial intelligence. What I'm talking about is assessing audio equipment to determine if it changes the sound passing through it. It doesn't matter if the singer is John, Paul, George, or Ringo. Measurements absolutely can tell if a piece of gear (or your room) is changing the sound in any way that is audible.

--Ethan

So do you advocate that a person buy speakers or other audio components solely on the basis of how they measure with SOTA instruments instead of buying what they perceive to have the best sound quality. IOW buy what measures best regardless how it sounds. BTW, isn't the SOTA always changing?

Raj

[Ethan Winer]

Danny,

> The human ear is much more sensitive than you think. <

No, it's much less sensitive - or at least much less accurate and repeatable - than you think!

> How about spacial cues? Why does one pre-amp allow for a deep sound stage and then another one jumbles the sound stage up in a 2 dimensional way. <

This is exactly what I mean. There is no way a preamp can have any affect at all on spatial cues, unless it has a rolled off high end (easy to measure), or has DSP reverb effects. Please read my Believe article which explains why imaging etc can seem to change even when nothing at all has changed. Here's the link again:

www.ethanwiner.com/believe.html

> My Clio measuring system can very accurately measure the amplitude of a note, but it won't tell me what played it. In fact I can make two different instruments play the same note and look exactly the same on a measured frequency response, but they very sound different. <

See my reply to John Cassler above. Also, an FFT analysis can easily distinguish middle C played on a clarinet from middle C played on a flute or cello!

> Most "real" audio engineers know that if it measures good but sounds bad it is still bad. <

I can't think of one example of gear that measures good but sounds bad, assuming everything that matters is measured of course. (Not just response or just harmonic distortion.) Indeed, if all the parameters measure good, what aspect of the sound could be bad?

> A manufacturer hands me two capacitors to evaluate and compare wanting my feedback. This really happened. They are the exact same value and measure the same in every way, but the dielectric material is different. Guess what, they sound different too. <

I'm not saying that all capacitors are the same. But if they sound different, then that difference can be measured. And the difference can be measured more reliably, and more repeatedly, and to a much finer resolution, than could be done by ear. If you believe otherwise, I'd love to hear why!

--Ethan

[miklorsmith]

I heard Wilson Alexandrias with Halcro amplification and and commensurate digital source professionally set up in a treated room that sounded like crap.  My buddy and I both thought so.

My SET amp measures like crap but sounds great.