Audio Science: Does it explain everything about how something sounds?

[jkeny]

Yes we don't know much about the ear brain interface, but pray tell me, as one who constantly belittles sine wave testing, if I pass any frequency, any amplitude, and any single or multiple sinewaves through a device, and I monitor amplitude, frequency and time on the output plus look at any distortions that arise, what is the missing test here then? Whats exactly not being picked up by such a test with sinewaves.

As I've said before, we have come a long way with simple test signals & measurments - I'm not belittittleing them just saying that it might be time to stop clinging to them like luddites & move onto a more profitable search for measurements & test signals which better represent what we perceive using our auditory perception.

You also don't seem to understand that an audio reproduction system is NOT a linear system. By using simple test signals you are limiting the potential excursions from linearity of this system

Or if you prefer, you can do a null test, which is also an accepted test, albeit a bit harder to do? The device is characterized, the listener is not.

And what is the best null that can be achieved with music as the source? Is it above or below "defined thresholds of audibility"

Whats the elasticisity in such a test? Do the test every minute, every second for ever, and you will reveal the way the device acts for audio signals passing through at that instant in time.

Huh? What's this got to do with the "elasticity of auditory perception in listening (blind or sighted)?

Your comments such as this continue to confuse the problem. Interpreting measurments vs what folks like, does not mean that measurments are somehow missing something.

I see you just ignore what I write! I pointedly stated that this is incorrect binary thinking - that we have either "sound" or "likes", nothing in between. If there was no in-between, how could Toole's statement that people converge towards the same universal liking of speakers in blind testing? This couldn;t possibly happen if hearing was so "elastic" as stated by JJ via Amir. This has been pointed out to you before by orb but I see you choose to also ignore that post too!

And yes, the audio industry, and that includes you, fail to provide anything new in their tests because if they published what full bore IMD tests look like, folks would wonder, as I have long been amazed out, how audio (and our ears ability to ignore stuff) works as well as it does.

I'm not sure what you are saying here - can you expand on this, please?

 

[amirm]

The first "datapoint - the paper - is a specific that cannot be universalised - so it is therefore not a datpoint for this discussion.

Of course it is a data point. Multiple data points then result in conclusions that can be drawn. If you are dismissive of these data points, what to do about the fact that for the opposite argument there are not even these data points? Just a lay supposition of long term testing being better.

This second "datapoint" - your experience - is no different from the "data point" that some of us have related - our experience.

No, my experience is formal and as part of my professional career. On the former, I had ways of validating whether my experience was right or wrong in formal, controlled testing. You all's experiences are not such. And for the latter, my decisions defined my livelihood and that of countless employees and company's success and reputation. So they had considerable weight behind them requiring the conclusions to be properly drawn. You all's experiences have no such weight.

 

[amirm]

As has been pointed out before - this "elasticity of our experience" is surely one reason why a single blind test is less reliable than a longer-term period of listening ?

The problem with elasticity is assigning a cause to it. "My music sounds better now so it must have been due to warm up, lifting the cable, etc." These things are controlled in blind tests and hence do not allow you to draw improper conclusions from them as you can in sighted evaluation.

A single blind test can easily be biased by this "elasticity". Your anecdote of "bad sound" was not explained to be based on anything other than your internal state so blind or not this would have been your impression of the sound.

The problem with my observation was again, assigning a cause, i.e. someone must have changed the demo system. In a blind test such possibility is not there so we don't get to draw improper conclusion.

Given a longer listening experience, over many days or weeks, a larger variety of such "elasticity" as a result of inner emotional state happens & what's constant in the sound can be perceived through any such biases. Not to mention a wider variety of music used, etc.

And countless opportunities to think this and that is an explanation for a change in sound that in reality does not exist. But sure, show me any evidence of this being true. Anything other than an appeal to an argument.

 

[AJ Soundfield]

Given a longer listening experience, over many days or weeks, a larger variety of such "elasticity" as a result of inner emotional state happens & what's constant in the sound can be perceived through any such biases. Not to mention a wider variety of music used, etc.

John, that's your "theory" based on long term viewing/touching/knowing/moods/lord knows what else/etc.
But as has been mention ad infinitum, there is no time limit on controlled blind testing.
You just can't peek at that cherished amp/dac/DUT. Or touch. Or know, etc, etc.
You have to trust your ears.
Of course, if one is a chronic peeker with untrusted ears and must rely on social innuendo, etc., obviously one steers clear of valid scientific controlled tests. Sort of a catch 22.

cheers,

AJ

 

[jkeny]

Of course it is a data point. Multiple data points then result in conclusions that can be drawn. If you are dismissive of these data points, what to do about the fact that for the opposite argument there are not even these data points? Just a lay supposition of long term testing being better.

Without more knowledge about the distortion used, I would have a large question mark over the Clarke paper being relevant to the discussion. Do you have such information other than what you already reproduced? Not all measurements/experiments represent datapoints - we've already agreed that THD is NOT a relevant measurement of audibility - if it's not a trusted measure then how can it be useful as a suitable measure of a test signal or a test device? OK, I hear you say that it was audible in an A/B test so therefore it is audible. As stated we need more information about the whole experiment - was the A/B test done with test tones or music - as I presume music was used in the long-term listening?

I'm really not in a position to "blindly" accept this as a datapoint until I have more information

No, my experience is formal and as part of my professional career. On the former, I had ways of validating whether my experience was right or wrong in formal, controlled testing. You all's experiences are not such. And for the latter, my decisions defined my livelihood and that of countless employees and company's success and reputation. So they had considerable weight behind them requiring the conclusions to be properly drawn. You all's experiences have no such weight.

OK, you are invoking the MS card & I''ll give you that - your experience SHOULD be more relevant than others but I don't think that this is a trump card that allows you to deny others experience - just as I don't think JJ's statement (& experience) trumps all others (I see his blind spot as exemplified in this binary mindset - "it's either sound or preference")!

 

[Orb]

The problem with elasticity is assigning a cause to it. "My music sounds better now so it must have been due to warm up, lifting the cable, etc." These things are controlled in blind tests and hence do not allow you to draw improper conclusions from them as you can in sighted evaluation.


The problem with my observation was again, assigning a cause, i.e. someone must have changed the demo system. In a blind test such possibility is not there so we don't get to draw improper conclusion.


And countless opportunities to think this and that is an explanation for a change in sound that in reality does not exist. But sure, show me any evidence of this being true. Anything other than an appeal to an argument.

Amir,
are you seriously suggesting some amp designs do not subtly change over the 1st hour from cold start and it must be elasticity?
You use the example of elasticity, but unfortunately your earlier example involving yourself is a different behaviour; it sounded bad, you realised nothing changed then sounded good.
Here we are talking about it sounds ok-good from cold and then very good from about 30mins to 1hour (many anecdotally mention that part of this is an awareness of the perceived volume changing - of course it does not change physically-electronically by any meaningful amount if at all but is an interesting key variable beyond it sounds bad/it sounds good), if left on and tried a different time it sounds very good.
Using different amps does not always create the same situation, some are great from cold, some take 15mins-30mins, some take an hour.
This is not elasticity because the listening behaviour is not following the model you outlined with yourself.
Furthermore measurements show amps do behave differently from cold to warm-up, yes one can say the measurements are meaningless as they are so small but it does show something is happening with the amp from cold to warm-up.
It is like discussing jitter measurements that are already meaninglessly small to some vocal posters you argue with and then become smaller with solution tweaks or some systems, or even noise such as the Melco review and measurements (that lets be honest had a very wide S/N ratio and low noise even before the notable improvement) that should not have an influence on the audibility of sound quality.
Going with Elasticity the reviewer should had reached the same conclusion on both USB and ethernet, however his comments were consisted with the measurements he knew nothing about and that one setup (ethernet) had very marginal improvements (only jitter that were small to begin with) that could be done without the product and the other (USB) had very good improvements and interesting noise measurement improvements.
But then one may say it is meaningless because the measurements we so small to begin with.
Cheers
Orb

 

[AJ Soundfield]

Amir,
are you seriously suggesting some amp designs do not subtly change over the 1st hour from cold start?

Temperature change or Audibly change? It's possible I suppose. Where is the valid, scientific audibility evidence to support it?

 

[Orb]

Temperature change or Audibly change? It's possible I suppose. Where is the valid, scientific audibility evidence to support it?

Well until someone does an extensive test we do not have it, so it also cannot be said to be elasticity/cognitive fooling/etc as some of you are doing.
However it is backed up by engineering in terms of measurement behaviour and what the top audio engineers say
So this is the next best thing and why your wrong in your dismissive nature of posting , but lets not try to skew the context by saying it applies to all amps, it does not and comes down to topology,transistors used,etc.
Science is not just about valid scientific evidence from dbt abx testing and we ignore everything else, it is also from modelling-measuring-correlation-etc.
Orb

 

[ddk]

As has been pointed out before - this "elasticity of our experience" is surely one reason why a single blind test is less reliable than a longer-term period of listening ? A single blind test can easily be biased by this "elasticity". Your anecdote of "bad sound" was not explained to be based on anything other than your internal state so blind or not this would have been your impression of the sound.

Given a longer listening experience, over many days or weeks, a larger variety of such "elasticity" as a result of inner emotional state happens & what's constant in the sound can be perceived through any such biases. Not to mention a wider variety of music used, etc.

+1 John.

Lacking Harman's resources blind tests tend to be limited in both scope and time. The premise that there's always prejudice involved in sighted testing is absurd and I don't see how a few minutes of ABX testing can be more accurate than longer term experience.

david

 

[jkeny]

+1 John.

Lacking Harman's resources blind tests tend to be limited in both scope and time. The premise that there's always prejudice involved in sighted testing is absurd and I don't see how a few minutes of ABX testing can be more accurate than longer term experience.

david

Indeed, it's one of the platforms of my argument - this "elasticity" MUST apply to all listening. But like you, I don't believe "that there's always prejudice involved in sighted testing"

As far as I remember, JJ's comments on this were related to the fact that we can only pay attention to one thing at a time so when we listen to a second playback of the track or component we may listen to a very different aspect of the sound & therefore we will hear it differently. Again, this applies to blind testing, just as much as sighted so no get out of jail cards being dished out in this statement.

 

[treitz3]

Temperature change or Audibly change? It's possible I suppose. Where is the valid, scientific audibility evidence to support it?

Oh come on now. You have never heard a tube sound better after warm up or perhaps an SS amplifier with huge heat sink fins not sound better after they warm up? One shouldn't need scientific proof for that and I find it hard to believe you haven't heard one of the two....and I say this with the utmost of respect.....unless perhaps you or anyone else hasn't had exposure to that yet.

Some amps would be hard to tell (design dependent) if any change would have occurred but some amps have a change that would clearly show that the sound changes. I have experienced a couple of amps that have audible (not subtle) changes over the course of three days.....but I guess that doesn't matter. I don't have "scientific proof" to back up my claims (which aren't really claims BTW, they are facts of life). If I did, I'd be happy to provide it for you and others.

In the meantime, I'll just accept the fact and enjoy the music.

Tom

 

[AJ Soundfield]

Oh come on now. You have never heard a tube sound better after warm up

Yes.

or perhaps an SS amplifier with huge heat sink fins not sound better after they warm up?

No. Though I allow for this possibility due to pathological audiophile design. Sound improving after several hours, days, etc....ummm, no.

One shouldn't need scientific proof for that and I find it hard to believe you haven't heard one of the two....and I say this with the utmost of respect.....unless perhaps you or anyone else hasn't had exposure to that yet.

This whole detour was triggered by a specific claim...and yes I do need scientific proof for it, thanks.
However, this proof would then completely fly in the face of the whole premise of the thread, so I'm not sure how we detoured to this dead end...

cheers,

AJ

 

[Phelonious Ponk]

The premise that there is always bias involved in sighted listening is self-evident. That doesn't mean that there can't be bias involved in unsighted listening, but it does, by definition, remove a whole group of biases, even when it is poorly controlled. It is utterly illogical to assume anything else.

 

[jkeny]

The premise that there is always bias involved in sighted listening is self-evident. That doesn't mean that there can't be bias involved in unsighted listening, but it does, by definition, remove a whole group of biases, even when it is poorly controlled. It is utterly illogical to assume anything else.

As it is similarly illogical to ignore the ADDITIONAL influences/biases introduced by blind listening So we have some biases removed & other new ones introduced & the final result is .....................?

 

[AJ Soundfield]

As it is similarly illogical to ignore the ADDITIONAL influences/biases introduced by blind listening

Nope, nothing added to soundwaves>ears.
Just your favorite stuff taken away. Peeking, price, street cred, gold/silver content, awards, etc, etc.

 

[ddk]

Indeed, it's one of the platforms of my argument - this "elasticity" MUST apply to all listening. But like you, I don't believe "that there's always prejudice involved in sighted testing"

As far as I remember, JJ's comments on this were related to the fact that we can only pay attention to one thing at a time so when we listen to a second playback of the track or component we may listen to a very different aspect of the sound & therefore we will hear it differently. Again, this applies to blind testing, just as much as sighted so no get out of jail cards being dished out in this statement.

Even if that's true, it would only make the case for longer term experience.

The premise that there is always bias involved in sighted listening is self-evident. That doesn't mean that there can't be bias involved in unsighted listening, but it does, by definition, remove a whole group of biases, even when it is poorly controlled. It is utterly illogical to assume anything else.

If by bias you mean a preference for a type of sound, sure. And it will start with a bias for the type of music that one likes but I disagree with that premise in the context of product A vs B whether or not one is or isn't familiar with one or both.

david

 

[jkeny]

Even if that's true, it would only make the case for longer term experience.

Exactly & also it highlights one of the essentials to successful blind testing - finding the difference & being able to focus on that difference during blind testing is crucial to achieving anything other than a null result (although Amir has demonstrated he can do this, I suspect most can't - it's because of his training). Going into a blind test expecting that the difference will make itself known to you just displays an ignorance of auditory perception as does the idea that we have microphones on the sides of our heads (soundwaves>ears)

 

[ddk]

Exactly & also it highlights one of the essentials to successful blind testing - finding the difference & being able to focus on that difference during blind testing is crucial to achieving anything other than a null result (although Amir has demonstrated he can do this, I suspect most can't - it's because of his training). Going into a blind test expecting that the difference will make itself known to you just displays an ignorance of auditory perception as does the idea that we have microphones on the sides of our heads (soundwaves>ears)

Another issue with blind tests is that one can easily be led to come up with a desired result very much like a magician pushing a card on to unsuspecting audience and making it look like it was all chance.

david

 

[jkeny]

Another issue with blind tests is that one can easily be led to come up with a desired result very much like a magician pushing a card on to unsuspecting audience and making it look it was all chance.

david

Indeed, the Great Randi springs to mind & his "blind tests" travelling show - the latest Ethernet cable test being a prime example

 

[ack]

Peter, Al, ack, Tony et al

IMO AJ is a perfect reason for everyone to hit their "ignore"button and let him talk to himself

In medicine we call that ilk "shifting dullness" and it's best ignored

No worries, already done that