halcyon
Headphoneus Supremus
- Joined
- Oct 21, 2002
- Posts
- 1,877
- Likes
- 283
I must clarify few points.
First of all, Bangraman, thank you for your contributions. I should have said this the first time, but better late than never...
Just as it is difficult to admit to having heard differences between cables in a forum where nobody believes in such differences, it is likewise difficult to admit having NOT heard differences in a forum where many people rave poetically about the differences. You have done the latter and I respect and admire your pursuit of personal truth and honesty.
I have also personally battled with this question for some years now. I was initially a very strict cable disbeliever. After having studied some basic electrical pinciples about audio cables, I was even more so.
However, after having studies psychoacoustics and having read up on listening studies that contradict what measurements would make you think about cables... I'm now a doubter rather than a disbeliever.
I can't claim universal "no differences" nor can I claim "yes, differences are abundant, just listen to them!".
I have personally taken part in both IC and speaker cable blind listening tests. Sometimes I have heard absolutely no difference between an entry level good cable (i.e. 250 USD) and a zip cord. Sometimes I've heard a difference and not only have I heard it, but I've heard it under blind conditions and my impressions have correlated almost 100% with other listeners in the blind test (I say almost 100%, because no two listeners have identical vocabulary to describe auditory experiences).
For me, the differences, when they've been audible, have been very slightly audible and more to do with imaging/soundstage/placement issues or apparent roughness/microdynamism of the sound, rather than high-end/low-end extension/sibilance.
Can I honestly say that I could always reliably differentiate between a zip cord and a higher priced cable that I have once been able to distinguish sonically?
No, I can't.
The differences are so subtle for me personally that not only do I have to absolutely believe in my ability to hear differences, I'd also have to listen like a machine.
Now the first point is important: If you don't believe in the _possibility_ of differences, it makes very little sense to conduct any tests to begin with. This would just be a null test. While a good test methodology can rule out results where people are imagining differences, there is no test methodology available that can correct for not hearing differences that are already there. If you don't even believe in the possibility, you're not going to hear them in the first place, even if they are there (basic psychological test procedure).
As for listening like a machine - unfortunately, hearing is not a machine.
For example, it is well known in the audiology (hearing measurement/diagnosis) community that the above 11 kHz hearing sensitivity can be as much as 15-30 dB off depending on the test day (same test subject, same controls, same test signals, etc).
Also, it is well known from both neurological testing and psychophysical testing that senses adapt to unchanging or slightly differing signals rapidly over time.
This means that the initial differences even if audible, quickly become inaudible, untill we get a new baseline for our echoic memory to which new incoming sounds are compared to.
Our sense are, after all, difference engines: they calculate (if you accept the computing analogy) differences to previous temporal signals.
There are no absolute baseline reference signals that we can always detect accurately under all conditions. What we hear, is based on the previous sounds we heard just previously (effect on echoic memory) AND what kind of high level schemas (categorisation imprints) and long term auditory memories we have learned.
So, listening experience is a combination of what you heard just before (short term memory, to simplify the memory model here for the sake of discussion) + what you've heard/scrutinised all throughout your life (long term auditory memory). Of course hereditary, occupational, disease and anatomical features play into this as well.
So, to make a long story shorter: to be able to hear really small differences every single time they are played back to me, I should be able to function like a machine (in a repeatable/consistent manner). Unfortunately, hearing doesn't function like this. It's (likely) impossible to make it function like this.
The echoic memory / adaptation of senses -part is the reason why quick switching is often detrimental to detecting very small impairments between two audio signals.
You can test this for yourself, if you don't believe me.
Find a large enough difference that you can detect between two signals in a normal ABX setup.
Then try doing 100 repeats of listening and see your accuracy scrore drop as a function of number of repeats.
Two factors come to play here: sensory adaptation and attention control. The sensory adaptation you can't control, unless you listen to say, white noise/silence between tests (i.e. neutralize the baseline). The attention is something that one can learn to control, but is very hard for most humans on repetitive tasks.
Also, it should be noted that quick switching is just a way to switch between test signals.
It doesn't automatically include/exclude ABX.
ABX is a way of trying to identify X as test signal A or B.
It can be done with very rapid back-and-forth switching (often detrimental in small differences, esp. as the number of switches goes up).
... OR it can be done with "slow switching": listen to A for week, listen to B for week, listen to X for week and the push the button to indicate that X was either A or B. Take three weeks off and repeat.
For small signals, the neurological and psyhochological literature implies that "slow switching" could be in many cases much more useful method to find really small differences among signals.
Now the last point: if the differences are so small that one can only hear them every now and then and even then they are not huge - does it really make any sense to be bothered about them?
I think this last point is highly subjective, this is a hobby after all.
If it was my job, of course I wouldn't be bothered about it, unless the success of the job dependent on that very last 1% of performance (it rarely does, in my experience).
However, as it's a hobby and an exercise to learn more about sounds, equipment and my own hearing, I personally remain intrigued about the potential differences, even when I think I hear them, but am not convinced if they really are there or not.
best regards,
halcyon
First of all, Bangraman, thank you for your contributions. I should have said this the first time, but better late than never...
Just as it is difficult to admit to having heard differences between cables in a forum where nobody believes in such differences, it is likewise difficult to admit having NOT heard differences in a forum where many people rave poetically about the differences. You have done the latter and I respect and admire your pursuit of personal truth and honesty.
I have also personally battled with this question for some years now. I was initially a very strict cable disbeliever. After having studied some basic electrical pinciples about audio cables, I was even more so.
However, after having studies psychoacoustics and having read up on listening studies that contradict what measurements would make you think about cables... I'm now a doubter rather than a disbeliever.
I can't claim universal "no differences" nor can I claim "yes, differences are abundant, just listen to them!".
I have personally taken part in both IC and speaker cable blind listening tests. Sometimes I have heard absolutely no difference between an entry level good cable (i.e. 250 USD) and a zip cord. Sometimes I've heard a difference and not only have I heard it, but I've heard it under blind conditions and my impressions have correlated almost 100% with other listeners in the blind test (I say almost 100%, because no two listeners have identical vocabulary to describe auditory experiences).
For me, the differences, when they've been audible, have been very slightly audible and more to do with imaging/soundstage/placement issues or apparent roughness/microdynamism of the sound, rather than high-end/low-end extension/sibilance.
Can I honestly say that I could always reliably differentiate between a zip cord and a higher priced cable that I have once been able to distinguish sonically?
No, I can't.
The differences are so subtle for me personally that not only do I have to absolutely believe in my ability to hear differences, I'd also have to listen like a machine.
Now the first point is important: If you don't believe in the _possibility_ of differences, it makes very little sense to conduct any tests to begin with. This would just be a null test. While a good test methodology can rule out results where people are imagining differences, there is no test methodology available that can correct for not hearing differences that are already there. If you don't even believe in the possibility, you're not going to hear them in the first place, even if they are there (basic psychological test procedure).
As for listening like a machine - unfortunately, hearing is not a machine.
For example, it is well known in the audiology (hearing measurement/diagnosis) community that the above 11 kHz hearing sensitivity can be as much as 15-30 dB off depending on the test day (same test subject, same controls, same test signals, etc).
Also, it is well known from both neurological testing and psychophysical testing that senses adapt to unchanging or slightly differing signals rapidly over time.
This means that the initial differences even if audible, quickly become inaudible, untill we get a new baseline for our echoic memory to which new incoming sounds are compared to.
Our sense are, after all, difference engines: they calculate (if you accept the computing analogy) differences to previous temporal signals.
There are no absolute baseline reference signals that we can always detect accurately under all conditions. What we hear, is based on the previous sounds we heard just previously (effect on echoic memory) AND what kind of high level schemas (categorisation imprints) and long term auditory memories we have learned.
So, listening experience is a combination of what you heard just before (short term memory, to simplify the memory model here for the sake of discussion) + what you've heard/scrutinised all throughout your life (long term auditory memory). Of course hereditary, occupational, disease and anatomical features play into this as well.
So, to make a long story shorter: to be able to hear really small differences every single time they are played back to me, I should be able to function like a machine (in a repeatable/consistent manner). Unfortunately, hearing doesn't function like this. It's (likely) impossible to make it function like this.
The echoic memory / adaptation of senses -part is the reason why quick switching is often detrimental to detecting very small impairments between two audio signals.
You can test this for yourself, if you don't believe me.
Find a large enough difference that you can detect between two signals in a normal ABX setup.
Then try doing 100 repeats of listening and see your accuracy scrore drop as a function of number of repeats.
Two factors come to play here: sensory adaptation and attention control. The sensory adaptation you can't control, unless you listen to say, white noise/silence between tests (i.e. neutralize the baseline). The attention is something that one can learn to control, but is very hard for most humans on repetitive tasks.
Also, it should be noted that quick switching is just a way to switch between test signals.
It doesn't automatically include/exclude ABX.
ABX is a way of trying to identify X as test signal A or B.
It can be done with very rapid back-and-forth switching (often detrimental in small differences, esp. as the number of switches goes up).
... OR it can be done with "slow switching": listen to A for week, listen to B for week, listen to X for week and the push the button to indicate that X was either A or B. Take three weeks off and repeat.
For small signals, the neurological and psyhochological literature implies that "slow switching" could be in many cases much more useful method to find really small differences among signals.
Now the last point: if the differences are so small that one can only hear them every now and then and even then they are not huge - does it really make any sense to be bothered about them?
I think this last point is highly subjective, this is a hobby after all.
If it was my job, of course I wouldn't be bothered about it, unless the success of the job dependent on that very last 1% of performance (it rarely does, in my experience).
However, as it's a hobby and an exercise to learn more about sounds, equipment and my own hearing, I personally remain intrigued about the potential differences, even when I think I hear them, but am not convinced if they really are there or not.
best regards,
halcyon