• 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!

Human bottleneck

Kal Rubinson

Master Contributor
Industry Insider
Forum Donor
Joined
Mar 23, 2016
Messages
5,273
Likes
9,787
Location
NYC
Switching fast between two sources to see differences, is assuming that ALL the hearing is in the ears... Reality is that the brain will be overwhelmed and blend those two together.
First, fast switching, as in A/B, does not mean that the time listening to either A or B is brief. It can be as long as the subject wants.
Second, the brain (and its perception mechanisms) is better at recognizing contrast than subtle detail. Thus, fast switching lets it detect differences.
Heck, that's why we see movies as continuous when they are a bunch of static pictures. Brain leveling the experience.
Different mechanism in operation. Google "persistence of vision."
 

Everett T

Addicted to Fun and Learning
Forum Donor
Joined
Aug 2, 2020
Messages
648
Likes
486
well that's fine, and it's still useful to know where the lines are as far as (trained) awareness.
The audibility thresholds are pretty straight forward and with minor exception, always apply. When it comes to anything below that, it's usually people not wanting to leave anything on the table because they can.
 

NTK

Major Contributor
Forum Donor
Joined
Aug 11, 2019
Messages
2,656
Likes
5,819
Location
US East
As to why we need fast switching ... Here is article on how we our memory of auditory information works. The TL; DR version is that after half a minute, we will not remember the details of the sound we heard, only its abstract representations.

Some excerpts from the article ...

auditory_memory.png

Preperceptual Auditory Storage
Preperceptual auditory storage retains the uncategorized representations of auditory inputs that have not yet been fully processed (Massaro 1975) and is also referred to as short auditory storage (Cowan 1984).​
...​
The duration of preperceptual auditory storage is very short. Most researchers agree that it lasts less than 300 ms.​

Synthesized Auditory Memory
The auditory features stored in preperceptual auditory storage can be further analyzed to form integrated representations of sound.​
...​
The term “synthesized” refers to the process in which auditory features such as pitch, loudness, and aspects of timbre are analyzed and combined into integrated auditory representations. The duration of the synthesized auditory memory appears to vary from less than 1 s up to 30 s, depending on how it is measured, but it is most often found to be several seconds (Cowan 1984).​

Generated Abstract Memory
The integrated representations in synthesized auditory memory can be further processed to form abstract representations in generated abstract memory (Massaro 1975). The abstract representations are considered to be domain general, meaning that they do not carry information about specific sensory details. Thus, abstract representations generated from each sensory domain (hearing, vision, touch, and so on) are all stored together in the generated abstract memory.​
In more recent literature, generated abstract memory is often referred to as “the focus of attention” and is reported to have a core capacity of three to five items when various memory strategies are controlled (Cowan 2001). It is thought that information must be saved in generated abstract memory before high-level thinking about it can occur.​
 

SoNic

Member
Joined
Mar 14, 2022
Messages
93
Likes
41
Is there any serious research which confirms your claim?
You can't prove a negative (that A/B doesn't do what people think it does).
Is there any serious research that confirms that A/B tests can distinguish small differences?
People usually ask those questions when their personal belief doesn't agree with that statement. So they ask for a "higher authority".
Example:
"There are no supranatural beings"
"Are there any studies to prove that?"
 

SoNic

Member
Joined
Mar 14, 2022
Messages
93
Likes
41
First, fast switching, as in A/B, does not mean that the time listening to either A or B is brief. It can be as long as the subject wants.
To be "correct", per A/B mantra (belief), it has to be very fast. Like about a second to 10 seconds fast.
Otherwise, is just normal listening, only combined with a mental stressor...
Second, the brain (and its perception mechanisms) is better at recognizing contrast than subtle detail. Thus, fast switching lets it detect differences.

Different mechanism in operation. Google "persistence of vision."
Brain persistence is about the same mechanism. It doesn't happen in eye or in ear, but in the brain itself. It's an adaptation that we cannot go around...
A/B belief system is like imagining that where are robots and hearing is only due to ears, like a pair of microphones. Ignoring the measurement "tool" itself that those microphones are feeding.
You can hear BIG differences and that validates your belief. But you can't really catch minute differences by A/B fast switching. And that results in beliefs like "every device sounds the same" and "measurements don't matter".

I listened several musical programs over and over, to the point that I know how they should sound. And I can compare that "should" expectation with what I actually hear.
Rarely, it happens to play them through a new device that sounds "better" and that becomes my new reference, after listening a few more times on that device, until my brain memory gets adjusted.
I avoid listening to those programs over "bad" devices (like in the car, or streamed, or mp3 via phone headphones), so I don't get used to that bad sound.
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,054
To be "correct", per A/B mantra (belief), it has to be very fast. Like about a second to 10 seconds fast.
Otherwise, is just normal listening, only combined with a mental stressor...

Brain persistence is about the same mechanism. It doesn't happen in eye or in ear, but in the brain itself. It's an adaptation that we cannot go around...
A/B belief system is like imagining that where are robots and hearing is only due to ears, like a pair of microphones. Ignoring the measurement "tool" itself that those microphones are feeding.
You can hear BIG differences and that validates your belief. But you can't really catch minute differences by A/B fast switching. And that results in beliefs like "every device sounds the same" and "measurements don't matter".

I listened several musical programs over and over, to the point that I know how they should sound. And I can compare that "should" expectation with what I actually hear.
Rarely, it happens to play them through a new device that sounds "better" and that becomes my new reference, after listening a few more times on that device, until my brain memory gets adjusted.
I avoid listening to those programs over "bad" devices (like in the car, or streamed, or mp3 via phone headphones), so I don't get used to that bad sound.
You aren't the only one to make these claims using this methodology. The problem is the brain is not an instrument the way you are thinking. As an instrument it is effected by many things other than sound. Your methodology helps the brain feel better about what it thinks about something being different. Your approach does not help the brain accurately detect differences. Yes there is research to support these ideas. It is your method which ignores how the brain works.

For instance, listening as you describe the brain will usually hear a difference when in fact nothing has changed, but the brain thinks something has changed.

Can tell us what would convince you otherwise?
 

Kal Rubinson

Master Contributor
Industry Insider
Forum Donor
Joined
Mar 23, 2016
Messages
5,273
Likes
9,787
Location
NYC
To be "correct", per A/B mantra (belief), it has to be very fast. Like about a second to 10 seconds fast.
I don't know where you come up with that "belief" but it is fallacious and unnecessary. Responsible comparisons do not have that restriction
Otherwise, is just normal listening, only combined with a mental stressor...
Nope. It is normal listening to two samples (of any length) with a minimal transition time to permit assessment of contrast.
Brain persistence is about the same mechanism.
It is the same only if you insist on that unfortunate belief that the samples on either side of the transition are almost as brief as the transition. That's clearly not the case in A/B listening tests.
A/B belief system.....
Who cares about belief systems?! Keep to the science.
I listened several musical programs over and over, to the point that I know how they should sound. And I can compare that "should" expectation with what I actually hear.
You may think so but it is highly unlikely to be reliable. As Blumlein 88 says,
You aren't the only one to make these claims using this methodology. The problem is the brain is not an instrument the way you are thinking. As an instrument it is effected by many things other than sound. Your methodology helps the brain feel better about what it thinks about something being different. Your approach does not help the brain accurately detect differences. Yes there is research to support these ideas. It is your method which ignores how the brain works.
Yup.


.
 
OP
Y
Joined
Nov 11, 2022
Messages
75
Likes
48
Why can't Amir build a random a/b switcher that mutes both inputs during switching so that he can't tell which is which?
 

Kal Rubinson

Master Contributor
Industry Insider
Forum Donor
Joined
Mar 23, 2016
Messages
5,273
Likes
9,787
Location
NYC
There is no science behind A/B switching.
There is substantial science behind psychoacoustic A/B testing and that encompasses what we are discussing.
And I am done with this... because I don't argue with religious believers.
Good but, if so, avoid looking in mirrors.
 

Head_Unit

Major Contributor
Forum Donor
Joined
Aug 27, 2018
Messages
1,341
Likes
688
Is there a scientific bar for Sinad, where we can comfortably say, bro, that's enough nads :D, going further is a waste of natural resources?
According to my calculations, 108.35 dB.

More seriously, I worked at a gigantic autosound company, and we got quite a number of complaints about beeping noise in a CD changer. Which turned out to be crosstalk from a data wire next to the L channel...at literally around -100 dB FS!!! And even rerouting the cables to reduce another 6 dB was not enough for the folks who had already heard it. Seems incredible, but true. Makes you wonder how the capacitor-tweaking Japanese engineers never noticed it, if ordinary customers complained. Likely answer is they never actually listened in a car :facepalm:

We also have to keep in mind a DAC can get daisy-chained into a preamp and maybe active crossover then an amp, so you have to look at the total noise and distortion added up.
 

Head_Unit

Major Contributor
Forum Donor
Joined
Aug 27, 2018
Messages
1,341
Likes
688
A/B tests don't reveal all the differences.
Switching fast
Your analogy about swapping pictures is kinda funny, since that's actually how Clyde Tombaugh found Pluto ha ha. But I know what you mean. As for A/B, whoever said it had to be fast? To me the interesting way to do it* would be
- Hide all the equipment
- Listen as usual each day, maybe writing comments about the sound
- Have a trusted friend or minion swap the cable, amp, whatever. And eventually swap it back. But you don't know when.
- Eventually, compare your notes versus their log of when the whatever was swapped.
- See if there is any correlation.

The main thing is that for a test to be scientific you can only change ONE variable, in this case the equipment. Changing knowledge about the equipment is introducing psychological variables and then the test is not scientifically definitive by definition.

To those who say "But I know what I hear!! It is so obvious!!" I say "then why are you afraid to try it blind?!?" then they usually get mad some more, and when I ask "why do you get so angry at the mere possibility that it is all in your head?"

*my personal answer why I don't: because it's a time consuming pain in the butt, and I don't get enough time to just listen as is.
 
Top Bottom