IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Audio Memory and Sound comparison
Yee
post May 26 2010, 18:07
Post #1





Group: Members
Posts: 24
Joined: 25-November 09
Member No.: 75291



There is an interesting discussion at the Harbeth User Group about audio memory and sound comparison. Alan Shaw, the designer of Harbeth speakers, provided some interesting perspective from the point of view of a speaker designer.

A few interesting points were being raised. I am very interested to hear what you guys have to add about this interesting subject.
Go to the top of the page
+Quote Post
drbarney
post May 26 2010, 19:57
Post #2





Group: Members
Posts: 7
Joined: 20-December 09
Member No.: 76169



The temporal lobes of the brain are mentioned in sound processing. There was an interesting paper published in December 2000 entitled "Pop music and frontal temporal dementia" by Giovanni Frisoni in the journal Neurology. It described people who upon losing tissue in the frontal and temporal regions of the brain spontaneously changing their musical preferences to pop music, typically teenage disco and rap.
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post May 26 2010, 20:17
Post #3





Group: Members
Posts: 3700
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Yee @ May 26 2010, 13:07) *
There is an interesting discussion at the Harbeth User Group about audio memory and sound comparison. Alan Shaw, the designer of Harbeth speakers, provided some interesting perspective from the point of view of a speaker designer.

A few interesting points were being raised. I am very interested to hear what you guys have to add about this interesting subject.


Looks to me like Alan A. Shaw, Designer and owner Harbeth Audio UK, is well on his way to re-inventing the ABX box. ;-)

There's a book named "This is your brain on Music" by Levitan that covers this topic pretty well. Zillions of footnotes to scientific papers.
Go to the top of the page
+Quote Post
Woodinville
post May 26 2010, 21:37
Post #4





Group: Members
Posts: 1402
Joined: 9-January 05
From: JJ's office.
Member No.: 18957



Yikes.

Please see www.aes.org/sections/pnw/ppt.htm for a slide deck somewhat on this subject. Look for "why do we hear what we hear".

There is also an mp3 of a recording of the talk that went aong with it. It's pointed to at thewombforums.com in the "university" forum under the "Seattle Audio Talk" thread. Don't have it on me right now.

This does dovetail in both with auditory memory and with noticing distortions (or there lack), etc.


--------------------
-----
J. D. (jj) Johnston
Go to the top of the page
+Quote Post
Yee
post May 27 2010, 04:11
Post #5





Group: Members
Posts: 24
Joined: 25-November 09
Member No.: 75291



QUOTE (Woodinville @ May 26 2010, 21:37) *
Please see www.aes.org/sections/pnw/ppt.htm for a slide deck somewhat on this subject. Look for "why do we hear what we hear".

Interesting stuff. Thanks for the link.

I think our memory of sound is related with our usual cognition processes associated to sound stimuli - interpretation as speech, melody etc.

I believe that we have a mental model of spoken language in the brain - e.g how the word 'English' usually sounds like. But everybody speak the same language and the same words slightly differently, with their own characteristic voice. It takes the brain some time and efforts to adjust to the new speech pattern of a person, especially those with thick foreign accent. I believe that the adjustment corresponds to forming of a modulation circuit, an equalizer of some sort, in the brain to translate/normalize the new accented speech to our mental model.

What I am trying to say is that we do not remember the tonal characteristic of a person. Our 'memory' of the person voice is by virtue of the modulation circuit formed in our brain.

The same would apply to how we recognize say the sound of a cello playing together with an orchestra. We have a mental model of how the instruments should sound. If the speaker is highly colored, it will sounds strange at first, and the mind will have to adjust to the new sound of the cello. But after some time of listening (some research says less than 20 minutes) we will adjust to the colored sound and listening away with total ease and enjoyment. Even very crappy sound system can sound pretty good when we get used to its sound!

Is anybody aware of any research materials regarding this subject?
Go to the top of the page
+Quote Post
googlebot
post May 27 2010, 12:25
Post #6





Group: Members
Posts: 698
Joined: 6-March 10
Member No.: 78779



I think voice recognition is a too high order concept to be a prime example. It is interwoven with a multitude of data: semantical and grammatical context, even references to own mirrored movement of tongue, lips, and mouth, when someone is hardly understandable.

But in personal experiments, I have indeed experienced a type of highly detailed auditory memory, that is very well suited for audio comparisons. Depending on content I would say that its capacity is rather 1-3 seconds, than 2-3 and I need special mental preparation to tap into this ability. It is also very fragile! I don't think that said mental preparation requires higher order capabilities of the human brain. It rather feels the opposite, like I am shutting of as much access from "interested parties" as possible, so that the perceived data stays untouched until the actual (1 pass) comparison.

Fragility refers to both, a loss of concentration and collapse due to disrupting audio input. For example, a too long*, faded transition between too sources seems to introduce a third element into my auditory memory (the transition itself) which impedes a simple A/B comparison. I need the help of higher order brain functions to "remove" or "hold away" the third element from the equation. And any "touching" seems to imply a loss of detail, as I have experienced with very hard to detect differences (high bitrate lossy compression etc.). Surprisingly, it sometimes seems to be more work for my brain to sort out a too long faded transition, than to sort out a transitional "CLICK" from direct switching. I think this might be due to the fact that fades are highly correlated to the to be compared data and thus harder to separate from them. Transitional clicks have another damaging effect, but I think it kicks in earlier, in the ear. The click doesn't damage the memory of the A-track as much but the state of receptivity for the immediately following B-track is disturbed, since the ear transitions to a state of lower receptivity for a short time frame.

This is all the result of personal experience and introspection, which I'm quite trained in - nevertheless highly subjective, so your results may vary.



*In this context 100 ms can already be very long.

This post has been edited by googlebot: May 27 2010, 12:32
Go to the top of the page
+Quote Post
Yee
post May 27 2010, 14:44
Post #7





Group: Members
Posts: 24
Joined: 25-November 09
Member No.: 75291



QUOTE (googlebot @ May 27 2010, 12:25) *
This is all the result of personal experience and introspection, which I'm quite trained in - nevertheless highly subjective, so your results may vary.

*In this context 100 ms can already be very long.

I spent some frustrating hours doing the audiophile way of trying to compare various things - the unplug, plug, can you hear that the voice is better harnessed - type of bullshit. It got me very suspicious after a while, that is what got me started on this thread.

I find your accounts very intriguing. I may just setup an ABX environment with my PC, DAC and earphone and try to get some direct experience on it.

My earlier post was actually referring to the more permanent memory of voices and sounds, for example the notion that we 'remember' the voice of our wife. Another example is that on some occasions when I get to hear a sound system I would make mental notes like the bass is not well integrated, the mid is too accentuated (as compared to my sound system).

So there is a permanent imprint in our mind of the sounds that we are familiar with, an I am trying to find out what is it constituted of. I believe that this type of 'memory' is what gives an unsuspecting person the illusion that he can hop to different shops to sample the sounds of various systems and make a meaningful choice.
Go to the top of the page
+Quote Post
knutinh
post May 27 2010, 14:58
Post #8





Group: Members
Posts: 569
Joined: 1-November 06
Member No.: 37047



I think that there is a fundamental (and interesting) difference between:
1)What is the auditory memory, how large does a perceptual difference between A and B have to be before we can remember how A sounded and that B sounds different after X seconds?
2)Given that system A and B satisfy 1), will the consumer be more happy in the long term if choosing on than the other as a hifi system?

Academically, it is very interesting to see what can be reliably identified using best-case conditions. As a consumer myself, I am a bit more relaxed. Can I remember the sound of my old loudspeakers from 1996 compared to the current ones anymore? If I can not, then what interest do I have in them sounding perceptually different?

If you are given only hamburgers to eat, surely you would object f you were used to better food. But if, after a while you forgot how other food tasted, and forgot about its existence, would you be any less happy than if you had been given good food all along? I subscribe to the idea that humans are tailormade as "differential" devices, not absolute. We can hear one sound as louder than another, not as 57.4dB SPL. When that other reference is missing, we seem to adapt to pretty much anything?

-k

This post has been edited by knutinh: May 27 2010, 15:02
Go to the top of the page
+Quote Post
Yee
post May 28 2010, 11:27
Post #9





Group: Members
Posts: 24
Joined: 25-November 09
Member No.: 75291



What Constitutes our Long Term Audio Memory?

We do have certain memory of sound - for example we can very quickly recognize the sound of say a trumpet. I am interested to find out what that memory consists of. I would like to investigate this from the perspective the cognitive processes associated with sound stimuli. Consider the following:

Music and the Human Brain
Consider the perception of timbre, the attributes that distinguish a saxophone from a trumpet. When both are playing the same note A4 each instrument creates a fundamental tone at 440 Hz, and the same spectrum of harmonics at frequencies of 880, 1320, 1760 Hz and so on. The relative amplitudes of the harmonics, and their variations over time, are what give the instrument its characteristic sound. Nerves from the regions of the cochlea excited by these frequencies send signals to the brain, and presumably a brain module recognizes the different pitches, similar to the way we recognize different colors - although at this level the recognition is probably not conscious. Now what? How on earth does the brain disentangle the overlapping saxophone and trumpet harmonic series, and re-assemble them, so we hear two distinct instruments, each playing a single note, rather than a mishmash of pitches? Levitin suggests that a difference of a few milliseconds in time between the arrival of the two harmonic series is the basis for this amazing feat. Directional clues might be used as well. I also suspect that our memory of what each of these instruments sounds like when played solo helps in this process, . The brain module that does this will even fix the bass response of a inferior stereo system. If an instrument plays a note that produces tones at 39, 78, 117, 156 Hz and so on, but your stereo system can't produce a 39 Hz tone, your brain will fill in the gap and you will hear a 39 Hz pitch! This phenomenon is called "restoration of the missing fundamental." All of this processing occurs automatically, largely in parallel, and without any conscious effort.

Different pitches heard at the same time are processed by a module that extracts harmony. Different pitches heard at different times are processed by another module that extracts melody. Again we know this because Sacks describes patients who have lost one capability without affecting the other. Sacks describes a gifted musician who had a stroke. Suddenly he was unable to recognize a tune as simple as "happy birthday." Yet his perception of pitch and rhythm was intact, and he could read music and hum a melody. So the problem was specifically an inability of auditory processing of a sequence of pitches.


My conjecture is that the human brain do not store the actual sound of the various instruments - but only the interpretive modules that recognize the various sound.
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post May 28 2010, 13:23
Post #10





Group: Members
Posts: 3700
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Yee @ May 28 2010, 06:27) *
What Constitutes our Long Term Audio Memory?

We do have certain memory of sound - for example we can very quickly recognize the sound of say a trumpet. I am interested to find out what that memory consists of. I would like to investigate this from the perspective the cognitive processes associated with sound stimuli. Consider the following:

Music and the Human Brain
Consider the perception of timbre, the attributes that distinguish a saxophone from a trumpet. When both are playing the same note A4 each instrument creates a fundamental tone at 440 Hz, and the same spectrum of harmonics at frequencies of 880, 1320, 1760 Hz and so on. The relative amplitudes of the harmonics, and their variations over time, are what give the instrument its characteristic sound. Nerves from the regions of the cochlea excited by these frequencies send signals to the brain, and presumably a brain module recognizes the different pitches, similar to the way we recognize different colors - although at this level the recognition is probably not conscious. Now what? How on earth does the brain disentangle the overlapping saxophone and trumpet harmonic series, and re-assemble them, so we hear two distinct instruments, each playing a single note, rather than a mishmash of pitches? Levitin suggests that a difference of a few milliseconds in time between the arrival of the two harmonic series is the basis for this amazing feat. Directional clues might be used as well. I also suspect that our memory of what each of these instruments sounds like when played solo helps in this process, . The brain module that does this will even fix the bass response of a inferior stereo system. If an instrument plays a note that produces tones at 39, 78, 117, 156 Hz and so on, but your stereo system can't produce a 39 Hz tone, your brain will fill in the gap and you will hear a 39 Hz pitch! This phenomenon is called "restoration of the missing fundamental." All of this processing occurs automatically, largely in parallel, and without any conscious effort.

Different pitches heard at the same time are processed by a module that extracts harmony. Different pitches heard at different times are processed by another module that extracts melody. Again we know this because Sacks describes patients who have lost one capability without affecting the other. Sacks describes a gifted musician who had a stroke. Suddenly he was unable to recognize a tune as simple as "happy birthday." Yet his perception of pitch and rhythm was intact, and he could read music and hum a melody. So the problem was specifically an inability of auditory processing of a sequence of pitches.


My conjecture is that the human brain do not store the actual sound of the various instruments - but only the interpretive modules that recognize the various sound.


I stand here watching people struggle with what is known as if it was not known. I guess we need something like "This Is Your Brain On Music" for dummies. All of the above conjectures and far more are answered in that book with tons of references to technical papers.

BTW, Art Ludwig is a someone that I know as being a good guy, but the quote above is probably taken from something that he wrote a goodly amount of time ago.
Go to the top of the page
+Quote Post
Yee
post May 28 2010, 15:03
Post #11





Group: Members
Posts: 24
Joined: 25-November 09
Member No.: 75291



QUOTE (Arnold B. Krueger @ May 28 2010, 13:23) *
I stand here watching people struggle with what is known as if it was not known. I guess we need something like "This Is Your Brain On Music" for dummies. All of the above conjectures and far more are answered in that book with tons of references to technical papers.

You mean to book Musicopilia? I ordered it in Amazon, it will take days (or weeks) the cross the great wilderness to me.

Google churns up ten of thousands of references. Still shifting through trying to find the answer to the question: "What is the nature of our audio memory". A dummy book would definitely help.

p.s. any other book you would recommend?

This post has been edited by Yee: May 28 2010, 15:08
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post May 28 2010, 23:24
Post #12





Group: Members
Posts: 3700
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Yee @ May 28 2010, 10:03) *
QUOTE (Arnold B. Krueger @ May 28 2010, 13:23) *
I stand here watching people struggle with what is known as if it was not known. I guess we need something like "This Is Your Brain On Music" for dummies. All of the above conjectures and far more are answered in that book with tons of references to technical papers.

You mean to book Musicopilia?


No, I mean "This Is Your Brain On Music".
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 29th August 2014 - 20:47