Testing audiophile claims and myths

[gregorio]

You said everything is measurable. … What you have proved is only that everything can be be recorded. A form of measurement,

Exactly, so therefore everything is measurable which is the only thing I was trying to prove, hallelujah!!

What might not be measurable is everything in isolation but then I’ve not claimed that.

but not necessarily the same as analysis.

I have not claimed it is the same as analysis and when you stated I was assuming that it was, I specifically refuted that and gave the limited conditions of analysis of a difference file (in post #15,398).

If the artefacts to prove something is audible are hidden but other artefacts, it doesn't mean they are not there.

True, they are there and they have/could have been measured as part of the difference file from a null test, though not in isolation and not necessarily still audible, as I’ve said more than once.

You said anyone who could not analyse them mean they are not there or is incompetent.

No I didn’t, what I said was: “but an audible difference that doesn’t show up in ANY measurement, including a Null Test, how’s that even possible?
I’ve heard of rare situations where it was difficult to find the measurement type for a particular audible difference but not being able to measure an audible difference always ultimately comes down to the incompetence of the person attempting the measurement, albeit with mitigating circumstances on occasion.”

So in the case of TIM, was some new, previously non-existent, measuring equipment required to detect it, or could it be measured by the standard spectrum analysers of the day, but simply wasn’t? If it’s the former, then we would have an example of something audible that couldn’t be measured in isolation but could have been measurable within a null difference file. If it’s the latter, isn’t that ultimately; Incompetence of those who failed to do the measurement, albeit with mitigating circumstances?

I think we have now established that is not the case.

Where/How have we established what I stated is not the case? What you seem to think you’ve established is that we cannot individually analyse the components of a difference file but I don’t get the point, as I’ve already stated that all along!

G

 

[jagwap]

What you seem to think you’ve established is that we cannot individually analyse the components of a difference file but I don’t get the point, as I’ve already stated that all along!

G

But that IS the point. You choose to ignore it, so that you do not lose the argument. Being able to record in PCM doesn't mean we have finished the job.

If we cannot analyse an audible difference, it doesn't mean it is not there.

 

[Dogmatrix]

But that IS the point. You choose to ignore it, so that you do not lose the argument. Being able to record in PCM doesn't mean we have finished the job.

If we cannot analyse an audible difference, it doesn't mean it is not there.

Is there a logic trap here ?
Would not defining something as audible require analysis ?

 

[jagwap]

Is there a logic trap here ?
Would not defining something as audible require analysis ?

If you require more than objective listening tests to prove it, yes.

But the arguement here is are there possible audible differences that cannot (yet) be quantified in measurement.

Gregorio is arguing that a recording of the audio is a measurement and so that is proof all audible differences are measurable.

I am arguing he is being pedantic to prove a pointless point in order to win the argument. As this does not get us closer to working out what measurable difference the audible difference makes.

 

[Dogmatrix]

If you require more than objective listening tests to prove it, yes.

But the arguement here is are there possible audible differences that cannot (yet) be quantified in measurement.

Gregorio is arguing that a recording of the audio is a measurement and so that is proof all audible differences are measurable.

I am arguing he is being pedantic to prove a pointless point in order to win the argument. As this does not get us closer to working out what measurable difference the audible difference makes.

I see

Carrying the unknown is difficult in any debate , more so in this context

Aren't the measurable parameters well defined in terms of audio ?

 

[jagwap]

I see

Carrying the unknown is difficult in any debate , more so in this context

Aren't the measurable parameters well defined in terms of audio ?

Isn't this where we came in?...

Back in the beginning there was the big bang, and the universe started expanding....

 

[Dogmatrix]

Isn't this where we came in?...

Back in the beginning there was the big bang, and the universe started expanding....

Perhaps it has come full circle

There has been quite a volume of interesting , if a little obscure and technical , information in this line of debate

That's a win !

 

[gregorio]

But that IS the point. You choose to ignore it, so that you do not lose the argument.

How can I “choose to ignore it” when I mention it almost every post?

Being able to record in PCM doesn't mean we have finished the job.

Maybe not for an audio equipment designer who may need to precisely identify a difference, what’s causing it and what design changes will reduce/eliminate it, but we have finished the job if the question is simply can it be measured.

If you require more than objective listening tests to prove it, yes.

But we don’t need more than an objective listening test to prove it. Again, an audio equipment designer would probably need to know what the audible difference is but we’re just asking if there is an audible difference.

But the arguement here is are there possible audible differences that cannot (yet) be quantified in measurement.

Yes that is the argument here but that’s not the argument you seem to be having! The argument you are making is different, namely; We can have an audible difference that is not analysable in isolation.

That argument is taken as a given by music/sound engineers because we encounter it almost constantly. Take the example of a typical orchestra recording, can we analyse in isolation say the 3rd desk of the 2nd violins? Obviously a desk of violins is audible but we cannot analyse it in isolation. Does this mean there’s some unidentified audible phenomena we can’t measure?

Gregorio is arguing that a recording of the audio is a measurement and so that is proof all audible differences are measurable.

I’m being more precise than that because a null test first allows us to separate the differences from the content that isn’t different (audible or not, down to the resolution of our ADC and FFT), measure the amplitude of that difference content, by virtue of digitally recording it and measure it’s frequency content, using a FFT of that recording.

I am arguing he is being pedantic to prove a pointless point in order to win the argument.

It is not a pointless point, “Is it measurable?” is the whole point, though apparently not for you because you want to measure and analyse it in isolation.

As this does not get us closer to working out what measurable difference the audible difference makes.

A null test does get us closer because now we are dealing with only the difference/s rather than the difference/s + all the content that isn’t different. It may or may not get us all the way to working out exactly what it is, but that’s a different question. Again, the question is, “Is there a measurable difference?”, not “What is the measurable difference in isolation?”.

We have tests/measurements for some specific things in isolation, such as IMD, we also have tests/measurements for groups of things, such as SINAD but we don’t have tests/measurements for every isolated thing, such as a test/measurement specifically for the 3rd desk of 2nd violins. If we want to measure this in isolation, then we have to isolate it before we measure it. EG. Record the 3rd desk of 2nd violins on their own and then measure it or, create a difference file where the audible difference is isolated (not obscured by something else) and then measure it. But of course now we’re talking about test procedure rather than the actual measurements themselves.

G

 

[bigshot]

the arguement here is are there possible audible differences that cannot (yet) be quantified in measurement.

Arguments like this remind me of how many angels can dance on the head of a pin... How are you planning to prove unquantified differences without quantifying them? The fact is, a signal off a mixing board is audibly identical to a nice shiny CD made from the bounce down. You can run it back through that process 100 times and it will still sound exactly the same to human ears. If there is a difference, it doesn't matter. Why argue about things that don't matter?

The majority of the arguments in this forum can be lumped into two categories... outright falsehoods being propped up by logical fallacies, and exceptions to the rule that don't matter in the real world. You have a choice... baloney or Brobnigag.

 

[jagwap]

How can I “choose to ignore it” when I mention it almost every post?

Maybe not for an audio equipment designer who may need to precisely identify a difference, what’s causing it and what design changes will reduce/eliminate it, but we have finished the job if the question is simply can it be measured.

But we don’t need more than an objective listening test to prove it. Again, an audio equipment designer would probably need to know what the audible difference is but we’re just asking if there is an audible difference.

Yes that is the argument here but that’s not the argument you seem to be having! The argument you are making is different, namely; We can have an audible difference that is not analysable in isolation.

That argument is taken as a given by music/sound engineers because we encounter it almost constantly. Take the example of a typical orchestra recording, can we analyse in isolation say the 3rd desk of the 2nd violins? Obviously a desk of violins is audible but we cannot analyse it in isolation. Does this mean there’s some unidentified audible phenomena we can’t measure?

I’m being more precise than that because a null test first allows us to separate the differences from the content that isn’t different (audible or not, down to the resolution of our ADC and FFT), measure the amplitude of that difference content, by virtue of digitally recording it and measure it’s frequency content, using a FFT of that recording.

It is not a pointless point, “Is it measurable?” is the whole point, though apparently not for you because you want to measure and analyse it in isolation.

A null test does get us closer because now we are dealing with only the difference/s rather than the difference/s + all the content that isn’t different. It may or may not get us all the way to working out exactly what it is, but that’s a different question. Again, the question is, “Is there a measurable difference?”, not “What is the measurable difference in isolation?”.

We have tests/measurements for some specific things in isolation, such as IMD, we also have tests/measurements for groups of things, such as SINAD but we don’t have tests/measurements for every isolated thing, such as a test/measurement specifically for the 3rd desk of 2nd violins. If we want to measure this in isolation, then we have to isolate it before we measure it. EG. Record the 3rd desk of 2nd violins on their own and then measure it or, create a difference file where the audible difference is isolated (not obscured by something else) and then measure it. But of course now we’re talking about test procedure rather than the actual measurements themselves.

G

You are choosing to miss the point. To make a point.

I don't feel like repeating it all again. That does not mean you have won the arguement. It mean you didn't engage on the arguement, but steered it away from the argument:

Some repeatable audible differences cannot yet be quantified and analysed, yet.

 

[bigshot]

Some repeatable audible differences cannot yet be quantified and analysed, yet.

Name one.

 

[gregorio]

You are choosing to miss the point.

I’m not choosing to miss the point, I’ve addressed the point more than once but it’s a different point.

To make a point.

Absolutely, to make the point that all audible differences are measurable (though not necessarily analysable).

It mean you didn't engage on the arguement, but steered it away from the argument:

Some repeatable audible differences cannot yet be quantified and analysed, yet.

I did “engage on the argument” (of audible differences being measurable), but I didn’t engage on your argument (some audible differences aren’t analysable) because I’m not disagreeing with that argument. In fact, I even gave an example of a repeatable audible difference that cannot be analysed, the 3rd desk of the 2nd violins in a typical orchestra recording!

G

 

[jagwap]

I’m not choosing to miss the point, I’ve addressed the point more than once but it’s a different point.

Absolutely, to make the point that all audible differences are measurable (though not necessarily analysable).

I did “engage on the argument” (of audible differences being measurable), but I didn’t engage on your argument (some audible differences aren’t analysable) because I’m not disagreeing with that argument. In fact, I even gave an example of a repeatable audible difference that cannot be analysed, the 3rd desk of the 2nd violins in a typical orchestra recording!

G

Does anyone here think that is relevant to the improvement of the art of audio engineering? Isn't it just a step forward in the art of pedantry?

You usually do better than this...

 

[jagwap]

I’m not choosing to miss the point, I’ve addressed the point more than once but it’s a different point.

Absolutely, to make the point that all audible differences are measurable (though not necessarily analysable).

I did “engage on the argument” (of audible differences being measurable), but I didn’t engage on your argument (some audible differences aren’t analysable) because I’m not disagreeing with that argument. In fact, I even gave an example of a repeatable audible difference that cannot be analysed, the 3rd desk of the 2nd violins in a typical orchestra recording!

G

Now, the next step is: Can audio subjectivism find differences in audio that objective analysis cannot yet qualify or quantify?

 

[71 dB]

I rather enjoy the summer than engage in this pointless-looking discussion.