Testing audiophile claims and myths

[gregorio]

I think I wasn't actually *judging* anything. Just making an empirical observation that this particular earphone gave me more goosebumps than others with the same music.

But you weren’t just making an empirical observation that you got more goosebumps, you also “judged” what was causing it (euphony).

Afair, I never described it as "sound quality". Maybe you misunderstood my remarks.

You quoted bigshot saying that maybe “Fidelity” should be used instead of sound quality and responded “Or euphony”. Maybe I did misunderstand that you meant Euphony should be used instead of sound quality?

SQ, to me, is measured by realism, does the kick drum sound like a kick drum and clarity, can you hear the kick drum? Sound preference is whether you like to have lots of bass, such as kick drum in your music.

That definition is highly problematic for two reasons: Firstly, a kick drum sound (deliberately) almost never sounds like a real kick drum. Therefore, there would be almost no recordings (which include a kick drum) that could have good sound quality, which obviously is not the case. Secondly, how do you know this realism you experience (or don’t experience) isn’t just due to a personal bias or other perceptual error?

“Fidelity Quality” is the only rational measure of “Sound Quality”. Although we still have to be careful about defining the fidelity of what; fidelity to the performance is almost never entirely achievable and is very rarely even desirable. Do we mean fidelity to the recording (which again is rarely desirable), fidelity to the mix or fidelity to the master?

but this is also not the discussion i meant when i said ‘the discussion will never come to an end’.

I didn’t mean my response to appear like I was disagreeing with what you posted, more like just expanding on it.

The Sony NW-WM1A and NW-WM1Z for example measure identical but sound different.

No they didn’t/wouldn’t! This is the mistake many audiophiles make, instigated/encouraged by audiophile marketing. Certain measurements maybe identical, possibly (but rarely) even ALL the typically published specs and in very rare circumstances, even the typical measurements which go beyond the published specs. But if they sound different there is ALWAYS at least one measurement that will reveal it, albeit very rarely not a common/typical one.

So either they do not actually sound different or they do sound different and will measure differently, given the correct test/measurement. Let’s try YACA; let’s say we modify a VW Polo so it achieves a 0-100kph time the same as a new Ferrari. Does this mean our polo has a measured performance identical to a Ferrari? The Ferrari will have performance differences obvious to our senses, does that mean there are performance differences we can sense but not measure? If instead of a 0-100kph time we measured say a 0-268kph time, then we’d easily see this measured difference but no one publishes a 0-268kph time and possibly no one has ever done that specific test, does that mean this performance difference is unmeasurable?

G

 

[Vamp898]

But you weren’t just making an empirical observation that you got more goosebumps, you also “judged” what was causing it (euphony).

You quoted bigshot saying that maybe “Fidelity” should be used instead of sound quality and responded “Or euphony”. Maybe I did misunderstand that you meant Euphony should be used instead of sound quality?

That definition is highly problematic for two reasons: Firstly, a kick drum sound (deliberately) almost never sounds like a real kick drum. Therefore, there would be almost no recordings (which include a kick drum) that could have good sound quality, which obviously is not the case. Secondly, how do you know this realism you experience (or don’t experience) isn’t just due to a personal bias or other perceptual error?

“Fidelity Quality” is the only rational measure of “Sound Quality”. Although we still have to be careful about defining the fidelity of what; fidelity to the performance is almost never entirely achievable and is very rarely even desirable. Do we mean fidelity to the recording (which again is rarely desirable), fidelity to the mix or fidelity to the master?

I didn’t mean my response to appear like I was disagreeing with what you posted, more like just expanding on it.

No they didn’t/wouldn’t! This is the mistake many audiophiles make, instigated/encouraged by audiophile marketing. Certain measurements maybe identical, possibly (but rarely) even ALL the typically published specs and in very rare circumstances, even the typical measurements which go beyond the published specs. But if they sound different there is ALWAYS at least one measurement that will reveal it, albeit very rarely not a common/typical one.

So either they do not actually sound different or they do sound different and will measure differently, given the correct test/measurement. Let’s try YACA; let’s say we modify a VW Polo so it achieves a 0-100kph time the same as a new Ferrari. Does this mean our polo has a measured performance identical to a Ferrari? The Ferrari will have performance differences obvious to our senses, does that mean there are performance differences we can sense but not measure? If instead of a 0-100kph time we measured say a 0-268kph time, then we’d easily see this measured difference but no one publishes a 0-268kph time and possibly no one has ever done that specific test, does that mean this performance difference is unmeasurable?

G

As said, it can be revealed very easily by just recording a song.

But we are taking about people in the audio science hobby. People who sometimes spend sometimes more than 10'000€ for a measurement rig.

And they are the ones that people listen to and according to them, there is no difference in measurement because, of course.

You don't spend >10'000 on a measurement rig and then just record the line-out. And if the measurements rig says no audible difference, the following cult will spread the word

 

[Piotr Michalak]

Re atmospheric pressure , found this it seems our internal systems have it covered unless the change is sudden or extreme

Kitahara M, Ozawa H, Kodama A, Izukura H, Inoue S, Uchida K. Effect of atmospheric pressure on hearing in normal subjects. Acta Otolaryngol Suppl. 1994;510:87-91. doi: 10.3109/00016489409127310. PMID: 8128881.

Nice! Haven't read it but I wonder if takes into account only "normal hearing" (as in real life interactions, speaking, listening to music casually) or also hi-fidelity / ciritcal listening, where we are much more discerning and analytical.

 

[PhonoPhi]

I had people tell me that Sabre chips sounded different- a glare. Then I got a DAC/amp with a Sabre chip and it sounded the same as everything else I own. I can see how a manufacturer might hobble the fidelity of an expensive component to try to justify the higher price, I just haven’t seen that in the real world yet. I want to see the results of the listening test before I start hypothesizing about unproven, purely theoretical differences.

What was exactly the chip?
(Sabre, as most manufacturers, have several product lines)
What was the amp?

It is really great to be specific, especially given your persistent insistence for others performinh different blind tests.

In my experience "non-coloured" USB DACs at least are rather unique than not, similar to not selling just one red wine, 12.5 vol% for ultimate personal enjoyment.

 

[Vamp898]

I had people tell me that Sabre chips sounded different- a glare. Then I got a DAC/amp with a Sabre chip and it sounded the same as everything else I own. I can see how a manufacturer might hobble the fidelity of an expensive component to try to justify the higher price, I just haven’t seen that in the real world yet. I want to see the results of the listening test before I start hypothesizing about unproven, purely theoretical differences.

The amp (especially the capacitors in it) are the most important factor in the sound signature of an device. If you change the DAC but keep the amp, there will most likely be no audible difference. But the same DAC can sound very different, with different amps.

Get a DIY Amp, change one capacitor, even if the new one has the same ratings on paper but uses a different technology, the sound changes.

And to my experience, there also is a big difference between DAC/Amp Kombos and full digital amps. The latter sounding faster and more direct.

So it's the amp you want to change, not the DAC

 

[bigshot]

Than I get tired of different sound signatures or maybe rather styles of reproductions. And I start hearing shortcomings.

That sounds like it has more to do with your subjective feelings than it does any actual audible difference. Small differences are most clear in direct A/B switching. Ears adjust to small differences over time. You should be getting less discerning, not more. What is happening here is the length of time, which I’m sure is sighted- not blind, is allowing bias to creep in and make you think you hear differences that don’t exist.

Controls exist in controlled tests for a reason. If you throw out blind comparisons, and let perceptual error due to auditory memory affect your comparison, you’re going to imagine differences that don’t exist.

 

[bigshot]

The amp (especially the capacitors in it) are the most important factor in the sound signature of an device. If you change the DAC but keep the amp, there will most likely be no audible difference. But the same DAC can sound very different, with different amps.

You’re describing user error. The amp is interacting with the impedance of the transducers, so it’s really the transducers that are making the difference, not the amp. An amp should be audibly transparent, and most of the time it is. When it isn’t, it’s because the amp isn’t properly suited to the transducer. Used with the correct transducer, that amp would be audibly transparent too.

It is drop dead easy to make audibly transparent DACs and amps. Used properly, most amps and DACs on the market are audibly transparent. I say “most” just to satisfy people who say a colored amp or DAC might theoretically exist. Other than NOS DACs, I have never run across anything that isn’t transparent. And in the years I’ve been at Head Fi, I’ve been asking for evidence of a colored amp or DAC, but I have yet to get anything but subjective impressions and audiophoolery to indicate that such a thing exists. Please prove me wrong with a controlled listening test that shows there is a clear difference.

 

[gregorio]

This is strange, because at first I would agree with little differences between DACs, untill I listen to them for an hour or more.

At a glance, the iPhone dongle is as good as my $2k DAP, metaphorically speaking.

Than I get tired of different sound signatures or maybe rather styles of reproductions. And I start hearing shortcomings.

How do you know you’re hearing the shortcomings between two different DACs or different sound signatures and that you are not just experiencing the shortcomings of your hearing/perception becoming tired?

That's also how I measure transparency and neutrality of the source - the better it is, the longer can I listen to it without being tired.

But you are not measuring transparency or neutrality of the source, you are measuring how long it takes you to get tired and that is dependent on a whole bunch of variables that have nothing to do with audible transparency or neutrality. Obviously, for example, how tired you were when you started listening or when you last ate, how much mental or physical exertion you’ve done before starting to listen, your stress and emotional state, etc, etc, etc.

Maybe the bass is faster and highs are clearer, and more precise.

All that is easily measurable but you’re not measuring them, what you’re experiencing is almost certainly changes in your perception due to fatigue/tiredness.

But I I perceive it subconsciously, mostly. It only comes through after a longer listen.

After a longer listen you obviously must be more fatigued than when you started. So again, how can you claim to be hearing something that is inaudible or unmeasurable and not just a change in your perception due to fatigue (or some other factor)?

Re atmospheric pressure

There’s actually a lot of research been done on this, as it was and can be a life and death situation. Consider an aircraft pilot, subject to changes in air pressure and therefore a partial loss of hearing which will affect their ability to hear/communicate on the radio and with their co-pilot or navigator. This caused deaths and warranted considerable research, much of which was done prior to and during the 2nd WW, for obvious reasons.

Incidentally, quite a lot of the aforementioned Harvey Fletcher’s research was used in this regard.

G

 

[jagwap]

No they didn’t/wouldn’t! This is the mistake many audiophiles make, instigated/encouraged by audiophile marketing. Certain measurements maybe identical, possibly (but rarely) even ALL the typically published specs and in very rare circumstances, even the typical measurements which go beyond the published specs. But if they sound different there is ALWAYS at least one measurement that will reveal it, albeit very rarely not a common/typical one.

G

No. This is where we started pages ago. Let's use your terminology. There is NOT ALWAYS a measurement that allows an analysable difference in these sort of circumstances yet. You agreed that a few pages ago, and this, despite what you say able measurement definition, is what we mean: analysis of the cause.

 

[PhonoPhi]

You’re describing user error. The amp is interacting with the impedance of the transducers, so it’s really the transducers that are making the difference, not the amp. An amp should be audibly transparent, and most of the time it is. When it isn’t, it’s because the amp isn’t properly suited to the transducer. Used with the correct transducer, that amp would be audibly transparent too.

Let me continue with the wine analogy - the different taste of wines then equally can be called "a user error", since the taste "interacts" with the taste buds, interfering with the ultimate fidelity of the main purpose of alcohol consumption!

P. S. Finally, I am starting to get your definition of "fidelity"

 

[bigshot]

Nice! Haven't read it but I wonder if takes into account only "normal hearing" (as in real life interactions, speaking, listening to music casually) or also hi-fidelity / ciritcal listening, where we are much more discerning and analytical.

All hearing is normal hearing. You can’t will yourself to hear the unhearable. Audiophiles invest their egos in their golden ears, but those appendages become tin in controlled listening tests.

There is normal hearing and degraded hearing. Hearing can be degraded by damage and extreme old age. But you can’t hear beyond normal hearing with normal human ears. That’s just your ego fooling you.

From your previous post, I get the feeling that you listen more with your feelings than your ears. Whether or not that ham sandwich you ate for lunch agrees with you has more to do with the perceived sound signature than the fidelity of the equipment.

 

[gregorio]

The amp (especially the capacitors in it) are the most important factor in the sound signature of an device. If you change the DAC but keep the amp, there will most likely be no audible difference. But the same DAC can sound very different, with different amps.

Obviously that’s false. How does a DAC know when it’s attached to a different amp and change it’s sound accordingly? The DAC is always outputting the same analogue audio signal, regardless of what amp it’s connected to. Please provide some reliable evidence otherwise.

Let's use your terminology. There is NOT ALWAYS a measurement that allows an analysable difference in these sort of circumstances yet.

Oh god, not again. Your statement is FALSE, that is NOT my terminology, I did NOT say there is always a measurement that is analysable. How many times do I have to repeat this and how many times are you going to argue with an assertion you’ve made up and falsely attributed to me?

G

 

[jagwap]

Obviously that’s false. How does a DAC know when it’s attached to a different amp and change it’s sound accordingly? The DAC is always outputting the same analogue audio signal, regardless of what amp it’s connected to. Please provide some reliable evidence otherwise.

I do not think you know enough about these interactions to be so certain. The untrasonic outptu of DAC, while filtered, is still present. The amplifier will react differently to that even if the front end is bipolar or JFET. You say this stuff as definative, yet you do not have the experience to know the subtitles of this.

Oh god, not again. Your statement is FALSE, that is NOT my terminology, I did NOT say there is always a measurement that is analysable. How many times do I have to repeat this and how many times are you going to argue with an assertion you’ve made up and falsely attributed to me?

G

You are miss reading my intention. I am changing the terminology to fit what is actually being discussed, so you cannot keep asserting something that people take will as we know all measurements. We don't. You know we don't. But if they don't read the previous pedantic debate they will not know that. Please be more clear and use analysis instead of measurement when necessary to be clear.

 

[gregorio]

I am changing the terminology …

Right, so you admit you changed/made up the terminology and therefore your statement “Let’s use your terminology” was a lie.

I am changing the terminology to fit what is actually being discussed

No, you’re not! You’re changing the terminology to fit your agenda of “analysable” but the discussion is not about what is analysable. The discussion is that there was an audible difference while the measurements were all identical. Therefore, any measurement that isn’t identical disproves that observation/assertion, regardless of whether or not that non-identical measurement is analysable.

G

 

[jagwap]

Right, so you admit you changed/made up the terminology and therefore your statement “Let’s use your terminology” was a lie.

My terminology was "measurement", for what you called analysis. Because measurement for you includes recording it. Something you are likely well qualified to do. Yet is useless on its own in this case.

No, you’re not! You’re changing the terminology to fit your agenda of “analysable” but the discussion is not about what is analysable. The discussion is that there was an audible difference while the measurements were all identical.

No it isn't. The common measurements as you put it, which do not include your get-out clause of recording it, and saying you're done. Measurements that allow analysis did not show anything. Also uncommon measurements do not allow any analysis to find a cause.

Therefore, any measurement that isn’t identical disproves that observation/assertion, regardless of whether or not that non-identical measurement is analysable.

G

An MP3 at a high bit rate (by your own words, transparent), proves nothing on its own. Neither does PCM on its own. We need analysis to understand the difference. We do not always have that analysis available to us, so your "ALWAYS" is not necessarily correct.