digital theory versus reality

[johncarm]

Yeah, and this pattern even varies depending on your state of mind, arguably more so than any variance in the audio signal itself. We've all heard how a few glasses of wine will do wonders for the quality of your audio system... I call it "euphoric distortion"

This reply is going off into "we know nothing" land. I'm interested in this question:
 
Given two signals S1 and S2 representing sound pressure levels at the ear, do they trigger distinct patterns in the cochlear nerve? Do you agree that is an important area of knowledge?

 

[Joe Bloggs]

If someone makes a claim here, isn't it reasonable to expect a least the name of the theory they are basing it on?

A claim of what?

 

[RRod]

 
A lot of things are important, maybe more important than anti-aliasing filters. Does that mean there's no reason to investigate the effect of anti-aliasing filters?

 
Just trying to understand your MO.

 

[Joe Bloggs]

Given two signals S1 and S2 representing sound pressure levels at the ear, do they trigger distinct patterns in the cochlear nerve? Do you agree that is an important area of knowledge?

Look, you could cite the Butterfly Effect and say that a stray atomic particle of acoustic variation from the original recorded performance would result in a different reaction in the universe and hence ultimately different human reactions and hence that the stimuli were behaviorally differentiable.
https://en.wikipedia.org/wiki/Butterfly_effect

But what we're talking about here is conscious recognition of a difference in the sound. A LOT of difference in the sound (relatively speaking) has to happen before that happens.

My example of wine drinking (and indeed of the very very many things we could do to improve audio recordings before we started to worry about sample rates and antialiasing filters) go to show that many, many things in audio have lots lots more effect on the outcome that whatever it is you're worrying about.

 

[johncarm]

A claim of what?

 
 
 
For instance, Spruce claimed that you can tell if A and B are different from a single number, the relative level of (A-B) and A.

 

[johncarm]

Look, you could cite the Butterfly Effect and say that a stray atomic particle of acoustic variation from the original recorded performance would result in a different reaction in the universe and hence ultimately different human reactions and hence that the stimuli were behaviorally differentiable.
https://en.wikipedia.org/wiki/Butterfly_effect

But what we're talking about here is conscious recognition of a difference in the sound. A LOT of difference in the sound (relatively speaking) has to happen before that happens.

I didn't say there's a small difference between S1 and S2.
 
Also, I understand that the nervous system has noise in it. We can include the effect of that in our theory.
 
And of course there is much more to the brain than the cochlear nerve. 
 
But does that make the behavior of the cochlear nerve irrelevant? Are you saying no scientist should have any interest in the question, simply because it doesn't give the whole picture?

 

[Joe Bloggs]

But does that make the behavior of the cochlear nerve irrelevant? Are you saying no scientist should have any interest in the question, simply because it doesn't give the whole picture?

No...?

But how this relates to your original question, or indeed what is your question at this point, I have no idea. I mean, if you're not interested in the audibility of small differences (which are surely all that format differences, encoding differences and anti-aliasing filters produce), what are you interested in?

My example of wine drinking (and indeed of the very very many things we could do to improve audio recordings before we started to worry about sample rates and antialiasing filters) go to show that many, many things in audio have lots lots more effect on the outcome that whatever it is you're worrying about.

 

[johncarm]

No...?

But how this relates to your original question, or indeed what is your question at this point, I have no idea. I mean, if you're not interested in the audibility of small differences (which are surely all that format differences, encoding differences and anti-aliasing filters produce), what are you interested in?

As I said in the other thread, I'm taking a big picture view of knowledge about audio design and testing.
 
Basically, the central issue is simply this: we want to enjoy audio in our homes.
 
So what do we need to know? What math do we need? What physics do we need? What psychoacoustics do we need? What tests do we need?
 
You may note that audiophiles and objectivists go about this topic in very different ways. Different paradigms. So I'm going back to basics and asking "How do we know what we claim to know?"
 
Maybe I can get back to something specific. I was wondering earlier in this thread if the ear/brain is very sensitive to patterns in the relative timing of sonic events. You gave an example of two signals in which the relative timing was preserved even when passing them through an antialiasing filter. 
 
What I'm aiming at is the question of how band-limiting a signal affects the pattern in the cochlear nerve. 
 
But let's establish a frame of reference. Let's say we have two signals, S1 and S2. They both have two sonic events. In S1 the events are spaced by 100 ms. In S2 they are spaced by 110 ms. Would these signals produce distinct firing patterns in the cochlear nerve? That seems to be an important question, would you agree?

 

[Soundsgoodtome]

http://www.head-fi.org/t/813409/digital-theory-versus-reality/0_30#post_12705850

Bookmarked/subbed

Go on..

 

[spruce music]

  As I said in the other thread, I'm taking a big picture view of knowledge about audio design and testing.
 
Basically, the central issue is simply this: we want to enjoy audio in our homes.
 
So what do we need to know? What math do we need? What physics do we need? What psychoacoustics do we need? What tests do we need?
 
You may note that audiophiles and objectivists go about this topic in very different ways. Different paradigms. So I'm going back to basics and asking "How do we know what we claim to know?"
 
Maybe I can get back to something specific. I was wondering earlier in this thread if the ear/brain is very sensitive to patterns in the relative timing of sonic events. You gave an example of two signals in which the relative timing was preserved even when passing them through an antialiasing filter. 
 
What I'm aiming at is the question of how band-limiting a signal affects the pattern in the cochlear nerve. 
 
But let's establish a frame of reference. Let's say we have two signals, S1 and S2. They both have two sonic events. In S1 the events are spaced by 100 ms. In S2 they are spaced by 110 ms. Would these signals produce distinct firing patterns in the cochlear nerve? That seems to be an important question, would you agree?

Actually that exact thing has been studied extensively.  I have read some of it.  From memory for many aural events somewhere near 100 ms from the clear blue is a threshold.  Short of it and the hearing ignores it (though the cochlea responded) and just a bit over it you hear what happened. 
 
I too am at a loss what you really are aiming to learn.  There is abundant research going back more than 100 years to some of these basic questions.  What is important has been worked out in large measure.  Audio products haven't always kept up, but such things are so good in most ways now they simply far outstrip the ear and it capabilities. Period.  End of the story.  Don't pass go, don't collect $200.  Don't worry about it.  Transducers are what remains that aren't up to that standard.
 
You should also realize, even the ear drum is not pure.  It responds only after tuning and filtering and reflection of the outer ear and ear canal shape the sound pressure on the drum far from what the raw sound is.  There is mechanical leverage and filtering, there is filtering in the way the nerves function to feed the auditory nerve system, and filtering as the nerves fire into the brain that processes hearing. With further filtering for memory and other purposes.
 
If you really are interested in these hearing particulars I suggest:
 
Fundamentals of Hearing by William Yost
https://www.amazon.com/Fundamentals-Hearing-Introduction-William-Yost/dp/9004236384/ref=sr_1_fkmr0_1?ie=UTF8&qid=1468040827&sr=8-1-fkmr0&keywords=moore+fundamentals+of+hearing
 
Written as a college text for maybe a sophomore level class.  It really is an excellent introduction that almost any serious audiophile would learn tremendously from.
 
Another is:
Brian Moore's The Psychology of Hearing.
 
https://www.amazon.com/Introduction-Psychology-Hearing-Sixth/dp/9004252428/ref=sr_1_fkmr1_2?ie=UTF8&qid=1468040827&sr=8-2-fkmr1&keywords=moore+fundamentals+of+hearing
 
Yes, I own and have carefully gone through both books though in earlier additions.  It will answer so many questions for you.  Much better than us spoon-feeding each answer to you and then you waste time with a question you wouldn't ask if you had the full basic background in the subject.  Or you misinterpret our answer because you lack background.

 

[johncarm]

 
Actually that exact thing has been studied extensively.  I have read some of it.  From memory for many aural events somewhere near 100 ms from the clear blue is a threshold.  Short of it and the hearing ignores it (though the cochlea responded) and just a bit over it you hear what happened. 
 
I too am at a loss what you really are aiming to learn.  There is abundant research going back more than 100 years to some of these basic questions.  What is important has been worked out in large measure.  Audio products haven't always kept up, but such things are so good in most ways now they simply far outstrip the ear and it capabilities. Period.  End of the story.  Don't pass go, don't collect $200.  Don't worry about it.  Transducers are what remains that aren't up to that standard.
 
You should also realize, even the ear drum is not pure.  It responds only after tuning and filtering and reflection of the outer ear and ear canal shape the sound pressure on the drum far from what the raw sound is.  There is mechanical leverage and filtering, there is filtering in the way the nerves function to feed the auditory nerve system, and filtering as the nerves fire into the brain that processes hearing. With further filtering for memory and other purposes.
 
If you really are interested in these hearing particulars I suggest:
 
Fundamentals of Hearing by William Yost
https://www.amazon.com/Fundamentals-Hearing-Introduction-William-Yost/dp/9004236384/ref=sr_1_fkmr0_1?ie=UTF8&qid=1468040827&sr=8-1-fkmr0&keywords=moore+fundamentals+of+hearing
 
Written as a college text for maybe a sophomore level class.  It really is an excellent introduction that almost any serious audiophile would learn tremendously from.
 
Another is:
Brian Moore's The Psychology of Hearing.
 
https://www.amazon.com/Introduction-Psychology-Hearing-Sixth/dp/9004252428/ref=sr_1_fkmr1_2?ie=UTF8&qid=1468040827&sr=8-2-fkmr1&keywords=moore+fundamentals+of+hearing
 
Yes, I own and have carefully gone through both books though in earlier additions.  It will answer so many questions for you.  Much better than us spoon-feeding each answer to you and then you waste time with a question you wouldn't ask if you had the full basic background in the subject.  Or you misinterpret our answer because you lack background.

 
I will get one or both books.
 
I can't parse your sentence here:
 
"From memory for many aural events somewhere near 100 ms from the clear blue is a threshold.  Short of it and the hearing ignores it (though the cochlea responded) and just a bit over it you hear what happened. "
 
What does "100 ms from the clear blue" mean? What events are you talking about? What is 100 ms referring to?

 

[gregorio]

 
Regarding wow and flutter, suppose you ask me to characterize the wow and flutter of a turntable. So I send you an FFT of the output of the turntable. Does that give you enough to go on?

 
It gives me everything! Are you changing your mind or contradicting yourself and now saying that in fact an FFT is not a complete representation of a signal?
 

 
Regarding frequency and amplitude: Why are we measuring a device? What's the practical reality here?

 
Given that the recording/reproduction of audio is: The measuring/converting of (only) acoustic frequency and amplitude, the storing of that converted data and then converting it back to acoustic frequency and amplitude, what else it is you think we should be measuring and what other practical reality is there?
 

 
I send you a plot of frequency and amplitude of its output. Does that help you know its fidelity? Does that help you check if it functioning correctly? Does it help you characterize its behavior? Does it help you with anything?

 
If you send me an FFT of a DACs output, an FFT of the input and I have a way of accurately comparing them, then the answers to your are: Yes, yes, yes and yes!
 
G
EDIT: Just realised this thread has gone off on another tangent

 

[johncarm]

   
It gives me everything! Are you changing your mind or contradicting yourself and now saying that in fact an FFT is not a complete representation of a signal?
 
 
Given that the recording/reproduction of audio is: The measuring/converting of (only) acoustic frequency and amplitude, the storing of that converted data and then converting it back to acoustic frequency and amplitude, what else it is you think we should be measuring and what other practical reality is there?
 
 
If you send me an FFT of a DACs output, an FFT of the input and I have a way of accurately comparing them, then the answers to your are: Yes, yes, yes and yes!
 
G
EDIT: Just realised this thread has gone off on another tangent

 
My central question here is: what is the appropriate measurement to characterize the behavior of a device? This question might come up when testing a new DAC design. The specific measurement we use might depend on the nature of the behavior we are characterizing.
 
The answer to my turntable question is "No." You do not have enough to characterize the wow and flutter if all I give you is an FFT of its output.
 
Note that an FFT is not just frequency and amplitude. 
 
If we want to characterize a device that is mostly linear, then measuring the transfer function will be useful.
 
However, the transfer function is not very useful for characterizing nonlinearities.
 
A measurement is useful to the extent it allows us to predict the behavior of a device.
 
The transfer function allows us to approximately predict the output of a device that is approximately linear. However, it does not allow a complete prediction.
 
A DAC is not a completely linear device. Something more than the transfer function is needed to characterize its behavior.

 

[Joe Bloggs]

... and when all the bellyaching about possible DAC, amp and file format gremlins is said and done, the "reality" is I can still stick a $50 pair of headphones into a $100 DAC/amp and get superior musical performance playing redbook CDs to TOTL rigs costing tens of thousands per part playing DSD512 or whatever, because I bothered to characterize the headphone's frequency response on my ears and my own loudspeaker-to-headphone HRTF and compensated for both using (shock horror) parametric equalizers and cross-channel convolution DSPs :rolleyes:

 

[johncarm]

... and when all the bellyaching about possible DAC, amp and file format gremlins is said and done, the "reality" is I can still stick a $50 pair of headphones into a $100 DAC/amp and get superior musical performance playing redbook CDs to TOTL rigs costing tens of thousands per part playing DSD512 or whatever, because I bothered to characterize the headphone's frequency response on my ears and my own loudspeaker-to-headphone HRTF and compensated for both using (shock horror) parametric equalizers and cross-channel convolution DSPs

 
Yes, my central question is what do we need to know to reproduce music with fidelity. But you are leaving an important concept undefined here. What is the definition of "superior musical performance"?