An argument that aural memory is enduring

mike1127 · April 18

Maybe some of you know "jj" the sound scientist. I watched the jj video on YouTube "What does accurate even mean?" It reminded me of what he used to say on the Usenet forum dedicated to audio 24 years ago. (rec.music.audio? don't remember the forum name)

He mentions that aural memory is only a few hundred milliseconds long. That's surprising to those of us who believe we can hear differences between components when auditioned minutes, hours, or days apart.

I have an informal argument that aural memory can last hours or days in musicians or musical instrument designers, and probably in other experts such as audio engineers, experienced audiophiles, etc.

First, we observe that musicians or musical instruments designers are honing their sound. They often have a unique, recognizable sound, and it's often of very high quality.

Next we observe that it takes a long time to develop that sound. Years of daily practice and experimentation. What are they doing during that time? I suggest they are "navigating"... that is, through past experimentation, they have some concepts of sound they like, and they are trying to move closer and closer to an image of what they would like to sound like in the future.

I observe that to "navigate" somewhere requires that you be able to determine where you are at this moment relative to previous locations. So, imagine that every morning, Yo-yo Ma gets up and starts practicing. Because aural memory only lasts a couple hundred milliseconds, he has no idea where he left off the night before, right? He has no image of the sound he'd like to produce today. And if he does something that sounds better today than it did yesterday, he has no idea, because of course he can't remember what he sounded like yesterday.

As this is an obvious absurdity, we conclude that he does have an enduring memory of his sound and that he can make fine distinctions in sound even in practice sessions that are days apart. By the same logic, so do expert instrument makers, audio engineers, etc.

The Computer Audiophile · April 18

I believe @Jud has researched this.

Jud · April 18

14 minutes ago, mike1127 said:

Maybe some of you know "jj" the sound scientist. I watched the jj video on YouTube "What does accurate even mean?" It reminded me of what he used to say on the Usenet forum dedicated to audio 24 years ago. (rec.music.audio? don't remember the forum name)

He mentions that aural memory is only a few hundred milliseconds long. That's surprising to those of us who believe we can hear differences between components when auditioned minutes, hours, or days apart.

I have an informal argument that aural memory can last hours or days in musicians or musical instrument designers, and probably in other experts such as audio engineers, experienced audiophiles, etc.

First, we observe that musicians or musical instruments designers are honing their sound. They often have a unique, recognizable sound, and it's often of very high quality.

Next we observe that it takes a long time to develop that sound. Years of daily practice and experimentation. What are they doing during that time? I suggest they are "navigating"... that is, through past experimentation, they have some concepts of sound they like, and they are trying to move closer and closer to an image of what they would like to sound like in the future.

I observe that to "navigate" somewhere requires that you be able to determine where you are at this moment relative to previous locations. So, imagine that every morning, Yo-yo Ma gets up and starts practicing. Because aural memory only lasts a couple hundred milliseconds, he has no idea where he left off the night before, right? He has no image of the sound he'd like to produce today. And if he does something that sounds better today than it did yesterday, he has no idea, because of course he can't remember what he sounded like yesterday.

As this is an obvious absurdity, we conclude that he does have an enduring memory of his sound and that he can make fine distinctions in sound even in practice sessions that are days apart. By the same logic, so do expert instrument makers, audio engineers, etc.

There's a vast difference between echoic memory (short-term memory for sounds) and pattern recognition (something you've heard hundreds of times and can recognize even when distorted).

Short-term memory for subtleties like timbre (essentially all characteristics of a note aside from pitch), or even pitch itself will last perhaps 4 seconds under good conditions, according to scientific experiments that have been published in peer reviewed journals and corroborated. Still skeptical? Try it yourself: https://deutsch.ucsd.edu/psychology/pages.php?i=209

On the other hand, if I heard even a poor recording of the voice of my mother, who passed away more than a decade ago, I'm confident I'd recognize it instantly, because that pattern was ingrained in me from the womb and for nearly 6 decades thereafter. Humans are so good at pattern recognition that they have been used to discern patterns in radio waves from Jupiter when computers utterly failed to separate patterns from noise.

The examples you're giving are of pattern recognition born of long practice. Are you familiar with the aphorism "10,000 hours to become an expert"? That's the instant grasp of some skill or information you attain when spending many, many hours laying down a pattern. But disrupt the pattern, and we're back to echoic memory. There's a famous experiment where college music majors were asked to identify instruments with the attack and release portions of the notes removed, so essentially just the pure tones and harmonics from each instrument. These students had spent thousands of hours in orchestra practice listening to those instruments, but their performance in the experiment was pretty dismal - the familiar patterns had been disrupted.

A really wonderful book about these and other aspects of the human auditory system, including classical composers' use of auditory illusions and techniques to affect your emotions, is this: https://www.amazon.com/Musical-Illusions-Phantom-Words-Mysteries/dp/0197672280/ref=sr_1_1?crid=3AO7JTMH3EIVP&dib=eyJ2IjoiMSJ9.iaFc8-F_V4cF48re2zOSSeV6FEhPDZUByr8a3q3wlSDFInS30o_49XrICo6_fwWny-pQshh8oB-baQbrwMioFb4YlCaI0bduEBqQRPy3Li3WTZn05JA3WMko6_4ueG3COoqMlLyaBiFJ_ggCgdnExggFzyi0xEKMKZ9RzXDeDflDL96rUpjFVhGkP5j6wxaD-KTGiq_69lVd91PkoNmPhvR-VhJmEYo55vfttdIrLM0.c0O-KQFaSaz1rJTxM1CS17LddSTM9wJCVryEwq1oCGg&dib_tag=se&keywords=diana+deutsch&qid=1713480622&sprefix=diana+deutsch%2Caps%2C123&sr=8-1

botrytis · April 18

Based on articles I have read, in peer-reviewd science magazines, @Jud is spot on. Between 4 to 10 seconds is max. But research on long term memory has shown, every time one accesses a specific memory,, t is biased by what is in short term memory. So, the memery is pulled out, chanded and put back in. So, long term memory is fallable also

Jud · April 18

1 minute ago, botrytis said:

Based on articles I have read, in peer-reviewd science magazines, @Jud is spot on. Between 4 to 10 seconds is max. But research on long term memory has shown, every time one accesses a specific memory,, it is biased by what is in short term memory. So, long term memory is fallable also

Long term memory is fallible under various circumstances, but although you almost certainly haven't heard "Row, Row, Row Your Boat" in a very long time, I bet if you heard even a short passage now you'd recognize it right off.

botrytis · April 18

4 minutes ago, Jud said:

Long term memory is fallible under various circumstances, but although you almost certainly haven't heard "Row, Row, Row Your Boat" in a very long time, I bet if you heard even a short passage now you'd recognize it right off.

I am talking specific important memories. Music memories, such as that, are stored in different parts of the brain. Since the advent of MRI's this field has advanced but still a way to go.

I have been interested in this area since taking care of my mother-in-law with dementia.

PeterG · April 19

The issue is not that the studies are wrong, it is that they are often misused to create nihilistic arguments that people should not spend a lot of money on hifi gear. There are plenty of good reasons not to spend a lot of money, but the idea that attentive listeners will not recognize a change in their system or cannot compare two power cables, just to pick a contentious example, is silly.

mike1127 · April 19

1 hour ago, Jud said:

There's a vast difference between echoic memory (short-term memory for sounds) and pattern recognition (something you've heard hundreds of times and can recognize even when distorted).

By that definition of "pattern recognition." I'm not talking about pattern recognition. I'm talking about Yo-yo recognizing small changes in his sound (including timbre) from day to day. He may recognize a new sound that he's getting -- definitely not something "heard hundreds of times."

mike1127 · April 19

The pitch experiment (recognizing the return of a pitch heard before) is an example of a test that would turn out completely differently if it were done on experts - by which I mean people with absolute pitch recognition. In a similar way, when I read descriptions of audio science, I see that a lot of these experiments were performed on laypeople.

botrytis · April 19

1 hour ago, PeterG said:

The issue is not that the studies are wrong, it is that they are often misused to create nihilistic arguments that people should not spend a lot of money on hifi gear. There are plenty of good reasons not to spend a lot of money, but the idea that attentive listeners will not recognize a change in their system or cannot compare two power cables, just to pick a contentious example, is silly.

It is not that, it is the fact that when you buy to improve your system, your brain fills in the rest of the details, even if they aren't there.

Classic sales technique in audio. Tell you what you should hear, play a piece. Did you hear it? They tell you again what you will hear and play the piece again. Then you obliviously heard the difference, even if it may not be there.

Auditory Neuroscience: Filling in the Gaps - ScienceDirect

The Brain Has Its Own "Autofill" Function for Speech | Scientific American

Your brain fills gaps in your hearing without you realising | New Scientist

Hearing With Your Ears, Listening With Your Brain – Association for Psychological Science – APS

There is way more out there. Just saying...

It may or may not be real.

Also think of tinnitus.

Understanding The Role Of The Brain In Tinnitus Perception (treblehealth.com)

Realize that the most common hallucinations are auditory.

Jud · April 19

1 hour ago, mike1127 said:

The pitch experiment (recognizing the return of a pitch heard before) is an example of a test that would turn out completely differently if it were done on experts - by which I mean people with absolute pitch recognition. In a similar way, when I read descriptions of audio science, I see that a lot of these experiments were performed on laypeople.

Read the book I recommended - it has quite a bit on perfect pitch.

Only a small minority of Westerners have perfect pitch. People who speak tonal languages like Chinese or Vietnamese display perfect pitch somewhat more frequently. But the vast majority of people don’t have that ability. And I don’t know anyone who displays “perfect jitter recognition,” do you?

I used the pitch experiment to illustrate how difficult it is to recall for any period of time past a few seconds even something as trivial as what note is being played, let alone the subtle differences we listen for between items of (hopefully) very good or excellent audio equipment.

Jud · April 19

2 hours ago, mike1127 said:

By that definition of "pattern recognition." I'm not talking about pattern recognition. I'm talking about Yo-yo recognizing small changes in his sound (including timbre) from day to day. He may recognize a new sound that he's getting -- definitely not something "heard hundreds of times."

And how does he understand the sound to be new? Because it is different from the old, i.e., the pattern.

Jud · April 19

3 hours ago, PeterG said:

The issue is not that the studies are wrong, it is that they are often misused to create nihilistic arguments that people should not spend a lot of money on hifi gear. There are plenty of good reasons not to spend a lot of money, but the idea that attentive listeners will not recognize a change in their system or cannot compare two power cables, just to pick a contentious example, is silly.

I agree that it’s tiresome when people continually insist that others who hear differences must be wrong. Yet I also think that people who continually insist they do hear differences are tiresome as well. In other words, I think it’s a mistake to have so much of one’s self-worth bound up in winning an argument, when our hobby is all about enjoyment.

I’ve tried to adhere mostly to reasonable logic when putting together my system, working toward objectively lower noise and distortion (past the point where strong objectivists would say it was audible).

But I also enjoy the subjective aspect of our hobby. For example, the player software I primarily use just came out with a beta version on Linux. Subjectively I feel I’m getting better sound from a minimal Arch Linux install with no desktop than from another distribution that I run with a GUI. Since the connection from player software to DAC is via optical Ethernet, there’s no reasonable explanation I can think of for any difference in sound. I’m very likely imagining things. And that’s just fine with me. There’s a joke about a father whose son thinks he’s a chicken. A psychiatrist describes to the father how he’ll cure the son. The father hesitates, and says “But we like the eggs.” Well, I like listening to the setup I think sounds better, even though it’s probably just my imagination.

About power cords, since you mentioned them: I caused a company selling expensive power cords to change its website when I pointed out in a Stereophile comment that their description of part of their manufacturing process was complete scientific BS. And there was another power cord manufacturer who claimed in these forums that his power cord was a room temperature superconductor (he also claimed on his own website that the same material was a super-lubricant for truck engines). For myself, about 35 years ago I bought some MIT power cords that are heavy gauge with built-in ferrites, and that’s been my only foray into that area.

mike1127 · April 19

1 hour ago, Jud said:

And how does he understand the sound to be new? Because it is different from the old, i.e., the pattern.

Okay... so how does that support your argument that we can only remember timbre for 4 seconds?

mike1127 · April 19

1 hour ago, Jud said:

Only a small minority of Westerners have perfect pitch.

I used the pitch experiment to illustrate how difficult it is to recall for any period of time past a few seconds even something as trivial as what note is being played, let alone the subtle differences we listen for between items of (hopefully) very good or excellent audio equipment.

Only a small minority of people are expert musicians. Only a small minority of people are fanatical audiophiles. Only a small minority on any subject are experts. So that doesn't have anything to do with it.

And an expert at using relative pitch for dictation could nail that test too. (And about half the students at the music school I went to had perfect pitch.)

Also an experiment in one context doesn't generalize to another context. An experiment on remembering pitches doesn't generalize to remembering the sound characteristics of audio equipment.

So to get back to my point, we widely recognize famous musicians to have gorgeous sound. How do you think they got there? By remembering their prior attempts to make sound. If we only remember sound for four seconds, then every day they would have forgotten what they sounded like the day before. No progress would be possible.

Jud · April 19

1 hour ago, mike1127 said:

Okay... so how does that support your argument that we can only remember timbre for 4 seconds?

Because it’s not echoic memory, it’s pattern matching. You know what does or doesn’t match the familiar pattern.

Have a look sometime at the Wikipedia article on New Coke. Tl:dr - Pepsi beat Coke in taste tests (short term memory for taste) because it was sweeter. So Coke changed its formula to make it sweeter. It completely bombed with Coke drinkers. Why? Because it didn’t “taste like a Coke” (didn’t match the pattern built up over long years of drinking Coke). So two different mental abilities, one very short term (sensory memory), one long term (pattern matching). This is also why more people who speak tonal languages have perfect pitch. They’ve been matching pitches mentally all their lives to derive the correct meaning when someone speaks to them, or when they speak to others.

Jud · April 19

1 hour ago, mike1127 said:

Also an experiment in one context doesn't generalize to another context. An experiment on remembering pitches doesn't generalize to remembering the sound characteristics of audio equipment.

The fact that blind tests of audio equipment, music resolutions, etc., pretty much never give better than chance results wouldn’t seem to square with a long term memory for those sound characteristics. (This involves not only echoic memory but the ability to consciously realize and verbally express any differences. Regarding this, see the description on Wikipedia of the experiment known as the “Iowa Gambling Task.”)

mike1127 · April 19

32 minutes ago, Jud said:

Because it’s not echoic memory, it’s pattern matching. You know what does or doesn’t match the familiar pattern.

Yo-yo's sound changes every day. There's not one familiar pattern. My point is that he, like all musicians, need to have a good memory for details of their sound ... not a fixed sound, but a changing sound .. or else they wouldn't reach the levels we expect of accomplished musicians.

You are suggesting he uses pattern matching to do this. But how long does he have to play exactly the same way to establish a pattern? 100 days of no variation in his playing? What if he can establish a pattern in 5 minutes?

botrytis · April 19

6 hours ago, Jud said:

The fact that blind tests of audio equipment, music resolutions, etc., pretty much never give better than chance results wouldn’t seem to square with a long term memory for those sound characteristics. (This involves not only echoic memory but the ability to consciously realize and verbally express any differences. Regarding this, see the description on Wikipedia of the experiment known as the “Iowa Gambling Task.”)

I posted in another thread a classic experiment that was done by the Boston Audio Society, that was a DBT, named 'Do all amplifiers sound the same.' The answer was yes.

Jud · April 19

28 minutes ago, botrytis said:

I posted in another thread a classic experiment that was done by the Boston Audio Society, that was a DBT, named 'Do all amplifiers sound the same.' The answer was yes.

Yes, and there’s also the well known one published in the Audio Engineering Society Journal, regarding failure to distinguish hi-res from RedBook.

Now either all these things really do sound the same and we audiophiles are totally wasting our money, or there are scientific experiments showing blind testing as normally done isn’t effective to distinguish differences that do exist and can reach us emotionally (or perhaps some of both). Why anyone would want to insist they were wasting their money by saying we live in a world where the best musicians on the planet are utterly inconsistent in their playing from day to day, I don’t know.

botrytis · April 19

53 minutes ago, Jud said:

Yes, and there’s also the well known one published in the Audio Engineering Society Journal, regarding failure to distinguish hi-res from RedBook.

Now either all these things really do sound the same and we audiophiles are totally wasting our money, or there are scientific experiments showing blind testing as normally done isn’t effective to distinguish differences that do exist and can reach us emotionally (or perhaps some of both). Why anyone would want to insist they were wasting their money by saying we live in a world where the best musicians on the planet are utterly inconsistent in their playing from day to day, I don’t know.

With hearing, especially, it is very much tied into our emotional centers, etc. I think that is part of the issue.

I really can't answer the question, honestly.

With my system, I look at the ambient noise, in the area, lots of wildlife and birds as well as traffic, so for me, that affects how well one can discern things.

Also, as we age, our ears tend to be less resolving and that can play into this as well.

I feel live and let live - it is all personal taste. For me it is about enjoyment of music, not necessarily the equipment.

I have friends that have quite expensive systems and one thing I can say is, they are never happy with what they have. I tend not to change equipment often. I am happy with mine, what does that say about me? No clue.

Great discussion so far!

The Computer Audiophile · April 19

1 hour ago, Jud said:

Yes, and there’s also the well known one published in the Audio Engineering Society Journal, regarding failure to distinguish hi-res from RedBook.

Now either all these things really do sound the same and we audiophiles are totally wasting our money, or there are scientific experiments showing blind testing as normally done isn’t effective to distinguish differences that do exist and can reach us emotionally (or perhaps some of both). Why anyone would want to insist they were wasting their money by saying we live in a world where the best musicians on the planet are utterly inconsistent in their playing from day to day, I don’t know.

I don't think I want to steer this conversation into the abyss, but it kind of seems that double blind testing as we've come to know it, and the fact that echoic memory is so short, would mean that something is very off with any conclusions drawn. How can a test of sound be done, on anything other than pattern recognition?

I've never thought of it before, and maybe I'm in my own world this morning, but it seems a bit strange to expect accurate and actionable results from an audio test. I could be very wrong though.

botrytis · April 19

38 minutes ago, The Computer Audiophile said:

I don't think I want to steer this conversation into the abyss, but it kind of seems that double blind testing as we've come to know it, and the fact that echoic memory is so short, would mean that something is very off with any conclusions drawn. How can a test of sound be done, on anything other than pattern recognition?

I've never thought of it before, and maybe I'm in my own world this morning, but it seems a bit strange to expect accurate and actionable results from an audio test. I could be very wrong though.

Well, this is the point. If one reads how, for example, the amplifier test was done, it was very specific. I think they did the best they could at the time.

This is why I always question my purchases of equipment. Does it really make a difference or am I putting the 'hype' in with it. It is the scientist in me, as I always question my results. I also want to make sure that when I publish results, they are easily reproduced by others (in science, not audio 😇).

I agree, Chris, that is why we are having this discussion. Maybe there is no answer. I don't know.

botrytis · April 19

One will notice, the most enduring aural memories, also are attached to the emotions at the time one heard sound.

I just wanted to point that out and not implying anything else.

The Computer Audiophile · April 19

3 minutes ago, botrytis said:

Well, this is the point. If one reads how, for example, the amplifier test was done, it was very specific. I think they did the best they could at the time.

This is why I always question my purchases of equipment. Does it really make a difference or am I putting the 'hype' in with it. It is the scientist in me, as I always question my results. I also want to make sure that when I publish results, they are easily reproduced by others (in science, not audio 😇).

I agree, Chris, that is why we are having this discussion. Maybe there is no answer. I don't know.

Totally!

I love these discussions where a couple objective topics aren't necessarily at odds, but they are butting up against each other a bit.

Can our very limited echoic memory, invalidate all but the most stringent listening tests? I have no idea, but it seems like there's more to the story than just sitting down and having some music played on system A/B/X. I kind of don't see how an A/B/X test could be valid if the audio wasn't switched between three components" all within say 5 seconds. but then, how can you listen to something extensively? I really don't know.

I like the science of both topics, but struggle to pur all the pieces together on this one.

An argument that aural memory is enduring

Recommended Posts

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Link to comment

Create an account or sign in to comment

Create an account

Sign in