• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Auditory memory

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
I just hope I never hear claims re echoic memory again in any audio forum!

Scientists don't understand how or how well it all works - ok?
 

MarkS

Major Contributor
Joined
Apr 3, 2021
Messages
1,062
Likes
1,502
I just hope I never hear claims re echoic memory again in any audio forum!

Scientists don't understand how or how well it all works - ok?
If someone can hear the difference between two components (amps, say) by listening for an hour, rather than for a few seconds, fine. Just demonstrate it by blind listening. Get 18 out of 20 right. Or 68 out of 100. (Odds against doing either by chance are around 5000 to 1.) Then you will have a case.
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,522
Likes
37,050
One of the odd (and mostly overlooked) conclusions of the Rees paper with a meta-analysis of hires testing was that longer times for music listening resulted in more positive results than short times. They didn't conclude the longer the better nor how long was optimum. But that multiple tests seemed to indicate you received more results of difference with 30 second snippets vs anything shorter.

I'm not sure I believe that would hold up to scrutiny. I would say even with the constraints of echoic memory, 30 seconds is a good test length for each selection. If differences are big enough you'll hear them. And it is more like how we listen to music for enjoyment.

There are a number of very minor differences I can detect blind with instant switching and 3-5 second long listening segments, but most of them I can detect no other way nor are they large enough to matter whatsoever to me in regards to my normal music listening. Any one have any thoughts on that?
 
OP
Wes

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
One of the odd (and mostly overlooked) conclusions of the Rees paper with a meta-analysis of hires testing was that longer times for music listening resulted in more positive results than short times. They didn't conclude the longer the better nor how long was optimum. But that multiple tests seemed to indicate you received more results of difference with 30 second snippets vs anything shorter.

I'm not sure I believe that would hold up to scrutiny. I would say even with the constraints of echoic memory, 30 seconds is a good test length for each selection. If differences are big enough you'll hear them. And it is more like how we listen to music for enjoyment.

There are a number of very minor differences I can detect blind with instant switching and 3-5 second long listening segments, but most of them I can detect no other way nor are they large enough to matter whatsoever to me in regards to my normal music listening. Any one have any thoughts on that?

Do you recall how long the longer times were?
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
The auditory memory issue arises when comparisons are made requiring quick switching. IIRC, the AES paper found auditory memory was about 6 seconds on average, and was one of the foundations of rapid switching during DBT and ABX tests. The shortness of auditory memory means you want to switch rapidly, not necessarily frequently with short times between.

Listening longer before switching is something I have always found helpful, since you "get used to" the sound so differences become more apparent. Switching back and forth multiple times in a few seconds only makes sense to me if there is something very specific and short-term to compare. Back when I helped arrange DBTs, we allowed customers to listen as long as they wished before running the test, to help cement their perceptions whilst encouraging them to identify key sections they could readily identify. Testing proceeded with them listening as long as they wished before switching and then identifying what changed (if anything). ABX was a major upgrade to the test methodology (R.I.P. Arnie!)
 
Last edited:

Geoffkait

Member
Joined
Jul 4, 2021
Messages
73
Likes
23
Speaking for myself I remember systems I heard in general terms, even some specifics, say, this morning, last week, last year, even fifty years ago. I suspect most people can remember such things too. You can remember what Elvis’ voice sounds like, right?
 

pkane

Master Contributor
Forum Donor
Joined
Aug 18, 2017
Messages
5,631
Likes
10,205
Location
North-East
I just hope I never hear claims re echoic memory again in any audio forum!

Scientists don't understand how or how well it all works - ok?

Understanding how something works isn't required to study it. As to how well it works, that's been studied and results published. So, I'm not sure what your point is.
 

DVDdoug

Major Contributor
Joined
May 27, 2021
Messages
2,917
Likes
3,831
Speaking for myself I remember systems I heard in general terms, even some specifics, say, this morning, last week, last year, even fifty years ago. I suspect most people can remember such things too. You can remember what Elvis’ voice sounds like, right?
Right! I don't need quick switching to know if my subwoofer is turned-off. Or, not too long ago I was at someone else's house and I noticed the left speaker wasn't working. That didn't require A/B switching or memory.

But the guys who've done lots of ABX tests will tell you that small-subtle differences are easier to detect with quick-switching so quick testing is a better, more reliable way to do it. And then you can consider if an "upgrade" or "improvement" is worth it if you can't tell the difference between A & B after waiting a day.

P.S.
When I was in high-school & college I worked at shop that repaired car stereos (and TVs, etc.). I was working on a car radio that was distorting. After making some changes and attempted fixes and listening for awhile it started sounding more normal and started to question if I was actually hearing distortion or not. I had to listen to something else to "re-calibrate" my ears to remind myself what normal-undistorted sound was like. It was a weird experience to become sort-of "immune" to distortion.
 
Last edited:
OP
Wes

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
Understanding how something works isn't required to study it. As to how well it works, that's been studied and results published. So, I'm not sure what your point is.

My point is not about mechanistic explanations vs. systems analytic frameworks.

I am tired of all the claims people make about aural memory and pointed out that it is very poorly understood. My post is a rant!! A scientific rant, but nonetheless...
 

Geoffkait

Member
Joined
Jul 4, 2021
Messages
73
Likes
23
How about visual memory or is that different? People can remember movie scenes quite vividly. Say Lawrence of Arabia. I have committed that one to memory. And Saving Pvt. Ryan. Paintings like Mona Lisa and American Gothic. We won’t talk about eye witness reports, though.
 
OP
Wes

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
that is called feature extraction (by some) - it happens in auditory memory too and is a big red herring thrown into the echoic pot
 

danadam

Addicted to Fun and Learning
Joined
Jan 20, 2017
Messages
956
Likes
1,496
My post is a rant!! A scientific rant, but nonetheless...
Maybe your point would carry better if you included a link to what triggered your rant?

You can remember what Elvis’ voice sounds like, right?
Why are people so fond of this (bad, imo) analogy? Yes, I can remember voices, but I can recognize them as easily on good as well as on bad systems. That you can recognize voices doesn't automatically mean that you can recall differences (often subtle) between systems.
 

MarkS

Major Contributor
Joined
Apr 3, 2021
Messages
1,062
Likes
1,502
How about visual memory or is that different?
The analogy to audio would be to compare two pictures, the same except that the tint has been very slightly changed on one, or a very small amount of pixel noise added (changing the brightness, say, of each pixel randomly by a small amount). It's MUCH easier to detect this if you can compare the two pictures side by side, instead of looking at one, and then, a few hours later, looking at the other.

But, as noted on this thread: https://www.audiosciencereview.com/...alogue-of-blind-tests.8675/page-6#post-837183
I'm no longer interested in this discussion.
 
Last edited:
Top Bottom