• 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!

Steve Guttenberg - Audiophiliac

Status
Not open for further replies.

win

Senior Member
Joined
Jun 17, 2020
Messages
430
Likes
432
Location
Irvine CA
I immediately recognized it sounded like *shit* compared to when I used the audio system before. It was muffled an boomy and that was, I know, NOT how this system sounded when I used it. Sure enough, checking the EQ, my rap-loving son had pushed the bass way up. As soon as I re-set the bass controls, it sounded as I remembered.

recognizing a poorly eq'd system (my brother used to do the same thing on my parents' car in the day) is a far cry from proclaiming that one expensive hifi component is 'warmer, deeper soundstage, etc' than another expensive high component. worth noting that it always seems like the more expensive stuff is better.
 

Sgt. Ear Ache

Major Contributor
Joined
Jun 18, 2019
Messages
1,894
Likes
4,150
Location
Winnipeg Canada
I think the point is that are you able to distinguish you mom's voice 3 days ago vs a week ago?
As we know we all sound so slightly different when we talk too much or haven't drink enough water or maybe a bit nasal.
But no, they are all your mom's voice, you most likely can't remember which days your mom's voice is particular dry.

Same with the Revel speaker analogy, are you able to distinguish bass turn up by 1 point vs 5 point? Probably,
What about 0 to 1 point. Maybe you will know something is different but cannot "test it" correctly during a test setting.

Again, How far can you take to say that blind test is the absolute conclusive evidence if our sound memory is unreliable as the study shows.
Now, I'm fully advocating that a lot of arcane audiophile equipment are snake oil.
However, if you really want to dig into accuracy of the absolute testing, then I think this is the effect that needs to be taken into consideration.

Oh I have done A/B test, there's no difference.
OK, are you an avg Joe or trained? How was the test setup
Are you able to mitigate the "sound memory" effect? What did you do to mitigate it, etc.

Apparently, some people choose to use sarcasm to avoid the question when their "religion" is being tested.


Your need to bring trained ears into the mix is an obfuscation. The fact that trained ears might have a better chance of distinguishing subtle differences doesn't in any way negate the validity of blind tests. If somebody claims they can hear something, and then in a blind test they are shown not to be able to hear that thing, the question has been answered. The fact that some third person who has special training (such as our fearless leader Amir perhaps, lol) might be able to hear the thing in question is immaterial to the point wrt this specific conversation about youtube reviewers because the question isn't "can a trained person hear the difference?" It's "can the person making the claim they can hear the difference hear the difference?"
 
Last edited:

smallricey

Active Member
Joined
Jul 5, 2020
Messages
108
Likes
17
Your need to bring trained ears into the mix is an obfuscation. The fact that trained ears might have a better chance of distinguishing subtle differences doesn't in any way negate the validity of blind tests. If somebody claims they can hear something, and then in a blind test they are shown not to be able to hear that thing, the question has been answered. The fact that some third person who has special training (such as our fearless leader Amir perhaps, lol) might be able to hear the thing in question is immaterial to the point wrt this specific conversation about youtube reviewers because the question isn't "can a trained person hear the difference?" It's "can the person making the claim they can hear the difference hear the difference?"
No, the question is if A/B test a valid set up for testing hearing difference?
If one are not able to hear the difference due to "sound memory" effect, which you claimed trained listeners could, then the test is invalid unless trained listeners were used in the testing. .

It's a simple logic. So, no it's not immaterial.
You could claim youtuber reviews are garbage because they are not able to distinguish the difference in a blind test.
Same can question if the blind test was set-up and done correctly or not.
If no one can is supposed to distinguish the difference then obviously the A/B test is flawed.
 
Last edited:

win

Senior Member
Joined
Jun 17, 2020
Messages
430
Likes
432
Location
Irvine CA
No, the question is if A/B test a valid set up for testing hearing difference?

standard practice would be to claim your hypothesis, in this case, "I can hear a difference". then the onus is on you to prove it. if you can't objectively prove you hear a difference, the result would be your hypothesis that you can is rejected.
 

Sgt. Ear Ache

Major Contributor
Joined
Jun 18, 2019
Messages
1,894
Likes
4,150
Location
Winnipeg Canada
No, the question is if A/B test a valid set up for testing hearing difference?
If one are not able to hear the difference due to "sound memory" effect, which you claimed trained listeners could, then the test is invalid unless trained listeners were used in the testing. .

It's a simple logic. So, no it's not immaterial.
You could claim youtuber reviews are garbage because they are not able to distinguish the difference in a blind test.
Same can question if the blind test was set-up and done correctly or not.
If no one can is supposed to distinguish the difference then obviously the A/B test is flawed.

More likely, there is no difference to be heard. There's no problem at all with an A/B test. I never intended, nor do intend to debate that with you. I can hear two sounds (even very similar sounds) and distinguish between them in a direct A/B comparison much more easily than I can comparing one sound now to a sound I heard a week ago. So if that's the debate you want to have, go start up another thread like the one that was shut down because it went nowhere and proved nothing.
 

smallricey

Active Member
Joined
Jul 5, 2020
Messages
108
Likes
17
standard practice would be to claim your hypothesis, in this case, "I can hear a difference". then the onus is on you to prove it. if you can't objectively prove you hear a difference, the result would be your hypothesis that you can is rejected.

In this case, the hypothesis should be "Trained listener can hear a difference". Since the sound memory effect would make standard A/B test a flawed test since no once can "remember" the difference between slight sound difference in signature in short amount of time, except for trained listeners.

If the train listener part is even true. If the A/B test set up is physically not possible to test the difference then your whole test is flawed.
Hence, the result is inconclusive.
 

smallricey

Active Member
Joined
Jul 5, 2020
Messages
108
Likes
17
More likely, there is no difference to be heard. There's no problem at all with an A/B test. I never intended, nor do intend to debate that with you. I can hear two sounds (even very similar sounds) and distinguish between them in a direct A/B comparison much more easily than I can comparing one sound now to a sound I heard a week ago. So if that's the debate you want to have, go start up another thread like the one that was shut down because it went nowhere and proved nothing.

Yah most likely there is no difference to be heard, but there's a problem and it's discussed in a thread in this site and other scientific experiment. \
So you are absolutely wrong about there's "no problem at all"

Obviously there's a problem, else there wouldn't be several studies on it
You can choose to ignore it and that's your call.
But you would be wrong, since you did not disapprove the effefct, you are just ignoring it. lol
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
In this case, the hypothesis should be "Trained listener can hear a difference". Since the sound memory effect would make standard A/B test a flawed test since no once can "remember" the difference between slight sound difference in signature in short amount of time, except for trained listeners.

If the train listener part is even true. If the A/B test set up is physically not possible to test the difference then your whole test is flawed.
Hence, the result is inconclusive.
Do I need to highlight all the wrong stuff here in red?
 

smallricey

Active Member
Joined
Jul 5, 2020
Messages
108
Likes
17
Do I need to highlight all the wrong stuff here in red?
Maybe you can highlight the wrong stuff in the original scientific study instead? If you are saying human can actually "remember" the difference.
Are you going to highlight human can determine sound loudness within 3db as well?
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
Maybe you can highlight the wrong stuff in the original scientific study instead? If you are saying human can actually "remember" the difference.
Are you going to highlight human can determine sound loudness within 3db as well?
I don’t need to, it’s your post that has the wrong stuff.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
In this case, the hypothesis should be "Trained listener can hear a difference". Since the sound memory effect would make standard A/B test a flawed test since no once can "remember" the difference between slight sound difference in signature in short amount of time, except for trained listeners.

If the train listener part is even true. If the A/B test set up is physically not possible to test the difference then your whole test is flawed.
Hence, the result is inconclusive.

I am missing something or not understanding your statement. Shoot, those myriad of A/B and ABX tests I did years ago using a variety of (mostly untrained) listeners must have been all wrong? Decades of research and all those stupid AES types too. In my, and any, tests pretty much everybody picked up on differences where they existed, and not where they did not. That is the fundamental principle of blind testing that has been used for decades. Long-term memory is suspect; that is why fast switching is critical, and why it makes it easy for anybody to hear differences. Even (or maybe especially) small differences. Auditory memory is only about 6 seconds IIRC; long-term listening reinforces short-term memory (in audio and anything else). You don't need trained listeners, just people who want to listen.
 

smallricey

Active Member
Joined
Jul 5, 2020
Messages
108
Likes
17
I am missing something or not understanding your statement. Shoot, those myriad of A/B and ABX tests I did years ago using a variety of (mostly untrained) listeners must have been all wrong? Decades of research and all those stupid AES types too. In my, and any, tests pretty much everybody picked up on differences where they existed, and not where they did not. That is the fundamental principle of blind testing that has been used for decades. Long-term memory is suspect; that is why fast switching is critical, and why it makes it easy for anybody to hear differences. Even (or maybe especially) small differences. Auditory memory is only about 6 seconds IIRC; long-term listening reinforces short-term memory (in audio and anything else). You don't need trained listeners, just people who want to listen.

I think what you described is exactly why many A/B tests could've failed.
That the untrained listeners who did not know the effect did not perform fast switching, or could not perform fast switching during the test for that matter. It would not have been possible to get the correct answer. Hence the results of those tests would have been inconclusive.
You might have set up your A/B tests correctly while many didn't.

To be honest, in order to perform these tests you should find unbiased random select listeners.
Now you have to add that these listeners need to know exactly what they are listening for.
imo that's difficult, not "just people who want to listen".

It's entirely possible that one person "feel" that one system is better than the other, yet don't really know which part is better.
Hence they would not be able to perform the fast switching action like you said and the test result would not be valid.
 
Last edited:

smallricey

Active Member
Joined
Jul 5, 2020
Messages
108
Likes
17
My god, man I am so blessed that can read, write and wave hands. Isn't that better than an illiterate???

lol if yall just wanna crack retarded jokes then I can do that too. So much for "science" forum eh lol
 

Wes

Major Contributor
Forum Donor
Joined
Dec 5, 2019
Messages
3,843
Likes
3,788
I just hope you are enjoying trolling, even if you don't catch any fish.
 

smallricey

Active Member
Joined
Jul 5, 2020
Messages
108
Likes
17
I just hope you are enjoying trolling, even if you don't catch any fish.
? the answer to your question is like right above it. I don't even know why would you ask that question if you did read.

It's either you did not read, you are illiterate, or you are the one who is trolling.
Stay within your comfort zone does not automatically makes things right, I think that's what science is about.
 
Status
Not open for further replies.
Top Bottom