[Shike]

Quote:

You omitted a word. I said the aesthetic and perceptual effects of DISTORTION can only be evaluated by listening. Therefore accuracy is a subjective phenomenon.

 
No, it really isn't.  You can judge whether you like distortion, but distortion can still be measured as whether it's accurate or inaccurate quantifiably by percent (THD+N, etc).
 
While you may LIKE or DISLIKE a certain distortion, it doesn't change the fact that it's changing the signal (hence distortion).

 

[mike1127]

Quote:

Quote:

You omitted a word. I said the aesthetic and perceptual effects of DISTORTION can only be evaluated by listening. Therefore accuracy is a subjective phenomenon.

 
No, it really isn't.  You can judge whether you like distortion, but distortion can still be measured as whether it's accurate or inaccurate quantifiably by percent (THD+N, etc).
 
While you may LIKE or DISLIKE a certain distortion, it doesn't change the fact that it's changing the signal (hence distortion).

I understand that different audiophiles come from very different perspectives.
 
I'm an amateur musician and decently familiar with the sound of live acoustic music. So I evaluate distortion by evaluating how similar the reproduction is to live acoustic music.
 
Sometimes a system sounds unlike live music in some way (so it has a distortion) and is also "pretty sounding." In that case, I may say that the distortion has some kind of appeal, but I would probably not say I "like" it because, well, it's a distortion, and robs the reproduction of its full potential.
 
What you seem to be saying is a kind of tautology. Any given distortion measurement gives you a number, and like all numbers, it can be larger or smaller. That's not really saying anything. It doesn't tell you how you experience that distortion (or whether you don't notice it at all). You can't even say that a larger number is bad, first because it might have no relevance to perception, and second because of a "third rail" possibility. See my post earlier in this thread about that. (About "euphonic distortion")

 

[wavoman]

Quote:

... Your bodybuilder friend, and all the fine and honest gentlemen selling goods that time and again were proven by scientifically sound tests not to work as advertised, maybe ought to ask themselves - who is that I am really working for?

 
Excellent points really.  Let me add this: I know in the case of the bodybuilder, and believe in the case of the cable-maker, that they honestly think the methods they advocate (sell) work.  They have convinced themselves.  Initially perhaps they were duped by the upstream manufacturers, but at this point they honestly -- that's the key word again, honestly -- are convinced.  Perhaps by palcebo, or flawed observation, or whatever.
 
No fraud.
 
In the case of audio cables, I don't think the proof is final yet.  The public DBTs are very flawed.

 

[Shike]

Quote:

I understand that different audiophiles come from very different perspectives.  
I'm an amateur musician and decently familiar with the sound of live acoustic music. So I evaluate distortion by evaluating how similar the reproduction is to live acoustic music.
 
Sometimes a system sounds unlike live music in some way (so it has a distortion) and is also "pretty sounding." In that case, I may say that the distortion has some kind of appeal, but I would probably not say I "like" it because, well, it's a distortion, and robs the reproduction of its full potential.
 
What you seem to be saying is a kind of tautology. Any given distortion measurement gives you a number, and like all numbers, it can be larger or smaller. That's not really saying anything. It doesn't tell you how you experience that distortion (or whether you don't notice it at all). You can't even say that a larger number is bad, first because it might have no relevance to perception, and second because of a "third rail" possibility. See my post earlier in this thread about that. (About "euphonic distortion")

 
Distortion is described in audio as variance from the signal.  Whether it sounds like live music or not isn't really relevant beyond what you want to hear.

When we're talking about accuracy in sound science, it is accuracy TO THE SIGNAL.  This doesn't really need more explanation, or discussion in relation to perception.  If the goal is accuracy of reproduction, in relation to the signal, then distortion is considered bad. 

 

[mike1127]

Quote:

Distortion is described in audio as variance from the signal.  Whether it sounds like live music or not isn't really relevant beyond what you want to hear.

When we're talking about accuracy in sound science, it is accuracy TO THE SIGNAL.  This doesn't really need more explanation, or discussion in relation to perception.  If the goal is accuracy of reproduction, in relation to the signal, then distortion is considered bad. 

In my view, audio equipment does not exist to reproduce a signal accurately in a mathematical sense. In exists to reproduce music in a perceptually accurate sense.
 
Your point of view here is bizarre to me. It conjures up someone sitting in front of an oscilloscope, totally grooving on the waveforms.

 

[wavoman]

Posted by
 
Quote:

So why not prove it can be heard in a blind test? With your own equipment, at home, with your friend/significant other swapping the cables with the help of a coin to achieve randomness, and with that person giving you absolutely no cue of the cable being tested.
You can even test the same same several days before swapping. It's a sound approach to the problem since audio exists to be heard and experienced, not to be measured.
 
Reply by eucariote
 
X2.  If you take 15-20 trials, it will be enough for a robust statistical analysis that avoids type 2 errors (false negatives).  If the differences can indeed be discerned on the basis of hearing alone, you should be able to achieve a statistically significant result with ease.  Post your results and I'll be happy to provide a T and P value to your test results.

 
I'll add another X2 to khaos974's post.  These simple, in-your-own-environment blind tests, with the help of a friend who does not / cannot tip you, is the way to go.  Merchants often let you return, so try lots of things (or sell the losers).
 
I do A/B "which do you prefer, if either", not A/B/X.  (You can also do "same or different?", but I prefer "which do you prefer" for lots of reasons I won't go in to here).  And I tell my friend/wife to mix it up, adding in false comparisons.  Flip a penny and a dime, Heads=A, Tails = B on both coins. Penny is first sample up, dime is second sample.  So I listen to AA, AB. BA. BB with equal probability of one-fourth.  The actual statistical analysis is complex, and is muddied by response bias and lack-of-independence between trials.
 
(The simple Yes/No binomial p values suggested by eucariote will not really be right either, for the same reasons, but it hardly matters).
 
The formal statistics are not worth doing in these one-subject tests on yourself.  One of my professors, head of the department where I got my PhD in statistics, LJ (Jimmy) Savage, once said famously "the best statistical test is interoccular traumatic -- the results hit you between the eyes".  And trust me, you won't need 15 or 20 tests.  Your friend will know very very quickly whether you can hear any difference reliably between A and B, whether there is any real preference here.  And when you look at the results, so will you.  I typically tell my friend to stop after six trials (you can always do more if it is inconclusive).  And we laugh -- I get it all wrong.  Can't tell the $1000 SPDIF cable from the $1.99 Rat Shack yellow video one.

 

[Shike]

Quote:

In my view, audio equipment does not exist to reproduce a signal accurately in a mathematical sense. In exists to reproduce music in a perceptually accurate sense.  
Your point of view here is bizarre to me. It conjures up someone sitting in front of an oscilloscope, totally grooving on the waveforms.

 
Audio equipment != instrument.  One is for reproduction, the other itself is creation.

Your point of view conjures up someone that hates the music they're listening to and thus are seeking outside forms of interference to "fix" it.  If you dislike how your music sounds so much why not just listen to different music rather than introduce distortion to "fix" it?
 
Some of us don't want to have distortion in our music, and judging by ear is a poor way of doing it since auditory memory is extremely bad (and equally hearing varies from person to person as well).  Precision measuring instruments are a much better bet with the science we have.

 

[Uncle Erik]

Mike, I enjoy music the same way you do. The testing is just to make sure that I'm not missing anything, it's more of a confirmation that there's nothing more to be done. If I replace the master cylinder on a car, yeah, I test it to make sure everything is right. Then I stop worrying about it. When I put an amp together, I test it to make sure the voltages are correct. If they are, I forget about the voltages and just enjoy it.

The same goes for cables. I haven't found any reason to use anything other than stock cables and inexpensive aftermarket ones. I've tested others to see if I'm missing anything. Nothing turned up, so I use the inexpensive cables, confident that the system is performing as well as it can. If something turns up in cable testing, I will take it seriously and reevaluate my position. But until that happens, I'll relax and enjoy the music.

 

[mike1127]

Quote:

Originally Posted by Shike /img/forum/go_quote.gif
 

Your point of view conjures up someone that hates the music they're listening to and thus are seeking outside forms of interference to "fix" it.  If you dislike how your music sounds so much why not just listen to different music rather than introduce distortion to "fix" it?

I have no idea where you get this.

 

[mike1127]

Quote:

Mike, I enjoy music the same way you do. ...

The same goes for cables. I haven't found any reason to use anything other than stock cables and inexpensive aftermarket ones. I've tested others to see if I'm missing anything. Nothing turned up, so I use the inexpensive cables, confident that the system is performing as well as it can. If something turns up in cable testing, I will take it seriously and reevaluate my position. But until that happens, I'll relax and enjoy the music.

Relaxing and enjoying music is very important, of course. But because I enjoy live music the best, I put some attention to the "musical accuracy" of equipment. It's the very reason that I enjoy what musicians do that I pay attention in this way. If I judge equipment purely by whether I enjoy this particular recording in this moment, I can get lost in a maze of interdependent variables. As I have said, accuracy of the reproduction of music can only be judged by listening to signal as music.
 
I understand doing measurements to gauge whether the equipment is performing properly. Measurements are somewhat useful to designers and technicians. But I do tend to find the best audio equipment is made by designers who put a lot of attention to listening, and make that more important than measurements.
 
As I explained elsewhere, an audio device can be regarded as a universe of behaviors. In one sense, taking a few measurements is like a tiny peek into that universe.

 

[Shike]

Quote:

I have no idea where you get this.

 
 
And I'm supposed to get your idea of me grooving to an oscilloscope?  

You're saying that a signal free of distortion may not sound accurate to you, but why are you bothering relying on distortion to fix a recording that isn't done to your standards anyway (you will never restore what was never there)?  Why do you want to fight what the artist or sound engineer did to undercut the final product?
 
Maybe, just maybe some of us want a true to the signal accurate reproduction of what was made so we don't mess with it.  In regards to signal accuracy less distortion is considered good.  Some of us like to know our music is produced as it was intended to be reproduced.  Subjective "accuracy" may be fine for you (if there really is such a thing), but we're in sound science and not a listening impressions sub-forum.
 
Regardless, distortion measurements on cables aren't really showing substantial measurable or audible differences.
 
 
And, lastly I do have a slew of different measuring headphones.  AKG K601, Stax SR-5 refurbed, UE TF10, ME M6, AT ATH-A700, Grado SR-60 with flats . . . etc etc.
 
I can listen to what I like and acknowledge it still isn't reproducing the signal accurately.  Such a conundrum I know.

 

[sohels]

Quote:

But I do tend to find the best audio equipment is made by designers who put a lot of attention to listening, and make that more important than measurements.

As has been previously suggested in this thread, measurements ensure accuracy of reproduction. 'The best audio equipment' will invariably measure well, unless what you seek is a flavored, lower-fidelity sound.

 

[mike1127]

Quote:

As has been previously suggested in this thread, measurements ensure accuracy of reproduction. 'The best audio equipment' will invariably measure well, unless what you seek is a flavored, lower-fidelity sound.

I think this is just a circular definition. Accuracy is defined as "measures well"; equipment that measures well is called "accurate." It really says nothing at all.

It may be that some people happen to like the sound of equipment that "measures well" (although even that concept is nebulous---what measurements do we mean?) and they take that as confirmation that measurements are useful. But that is not a valid conclusion. It's just a coincidence.

 

[mike1127]

Quote:

You're saying that a signal free of distortion may not sound accurate to you,
 
There is no such thing as a "signal free of distortion."
 
I can listen to what I like and acknowledge it still isn't reproducing the signal accurately.  Such a conundrum I know.
 
I explained already that sometimes I like effects that must be acknowledged as distortions.

 

 

[mike1127]

Quote:

 Subjective "accuracy" may be fine for you (if there really is such a thing), but we're in sound science and not a listening impressions sub-forum.

 
Science is relation to the reproduction of music is a different thing than, say, the science of global warming.
 
Temperature is well-modeled as a measurable quantity. Climate models are quantifiable models, which is appropriate.
 
My basic point is that it is a mistake to think that the science of audio is the same thing.
 
It is entirely possible to work with "subjective accuracy" and bring the kind of hypothesis-testing that scientists do without losing sight of the subjective nature of accuracy.
 
In my opinion, those who think "sound science" is the same as, say, climate science, are on the wrong track. At the very least they have very little useful to tell someone who wants reproduced music to sound like live music.