How far have ss amps really come in the last twenty years??

[Thomas savage]

FUFU

Unfortunately this posts falls into the offensive category. Combined with the members persistent antagonism and obvious disdain for the membership I chosen to banish him.

Member @daveyf is formally ejected from the premises.

We are here to learn and share knowledge and experience, not for pissing contests and not for pointless arguments over sighted listening fallibility, unless of course those arguments contain some new evidence and may actually progress the discussion.

Cheers

 

[SIY]

See: https://www.audiosciencereview.com/...ity-and-reliability-of-abx-blind-testing.186/

And:
http://www.aes.org/e-lib/online/browse.cfm?elib=5549

View attachment 29136

And another bibliography:
http://djcarlst.provide.net/abx_peri.htm

Lipshitz's "The Great Debate" paper from 1981 was the one that got me off my butt and setting up DBTs of my own.

 

[amirm]

Remember, on this site, there are folks who only listen with their measuring gear and get confused if they happen to see anything with their eyes. One thing is certain, they do NOT like to listen with their ears, LMAO. Probably a good thing, as I seriously doubt that they have the ability to hear much anyway...at least according to what they report!!!

Every one of my headphone amplifier tests includes listening tests. I probably have done 40 to 50 of them. I do them blind and level matched so they don't generate the results you wish but we do plenty of such tests here.

We love subjective listening tests. We just want them done correctly so that what our ears hear is judged and not other factors.

As to whether we can hear differences, as you well know, I have passed double blind tests that subjectivists don't dare taking. So let's stop that claim. In my blind testing I often hear distortion in poorly measured gear (again, blind without knowing which is which) confirming the validity of measurements. And disproving the theory of "measures bad but sounds good." Indeed these tests clearly prove that all the people who claim to hear goodness in these products are clearly blind to all the distortions they produce.

See also the other reference I just provided on how bad audiophiles are in "long term" testing of really high levels of distortion: https://www.audiosciencereview.com/...ity-and-reliability-of-abx-blind-testing.186/

 

[amirm]

And another bibliography:
http://djcarlst.provide.net/abx_peri.htm

Lipshitz's "The Great Debate" paper from 1981 was the one that got me off my butt and setting up DBTs of my own.

Thanks. I kept looking for that but missed the keywords to find it.

 

[BigRez]

This is a great thread for me right now. I even appreciate the opinions that I do not agree with because they give me insight into a line of thinking. I’ve been playing around with audio equipment for 40 some years but I just recently started getting more serious about it. I’m learning all the testing procedures I can and designing speakers. I’m encouraged by one of my photography and design clients who runs a tube amplifier company. He very much believes in measuring everything, as do I. That being said, I want to explain why I think the measurements still aren’t enough. The measurements are a good place to start but there are many interactions that can’t be easily measured. The idea that at a certain level all equipment is good enough for our ears, is not something I believe. There are qualities to sound that we can perceive but can not define or put into words. My clients tag line is “Just Listen”. This discussion is about amplifiers but the same theoretical concepts are true of every component in a system: cables, DACs, speakers, etc.
I’ve studied the art and science of photography in depth and I strongly believe that people can “perceive” image sharpness and quality beyond what their brains can explain in words. I’ve seen it time and time again. A study of sharpness shows that there is a pin-point area of sharp and a “circle of confusion” around that point. To say that an image is sharp enough because of a measurement at a particular viewing distance is naive. Sharpness is perceived in a very complex way, just as the listening experience is very complex and cannot be boiled down into simple measurements. The same principles apply to motion picture viewing and HiFi listening. The measurements are very important and a great place to start. I feel we should trust measurements more than our perception, but we don’t have measurements for everything that is important to the listening experience. My point is that we can tell there is a difference but we can’t explain it or even define it. A good example is an eye exam: it’s sometimes difficult to tell which is the better test lens, they are different but is one truly better than the other? In a certain situation you will see better with that lens but change the brightness a little and just like volume levels, everything changes.
Amplifiers do sound different, speakers respond differently, we can hear things beyond explanation and it does matter to me. The hobby is based on trying to accomplish the best sound possible in any given situation. I trust testing to get me in the ball park and then I use my ears to see if it’s making me happy. My perception may change over time and a different setup may or may not make me happy. Audiophile listening is both an Art and a Science, and it should be treated as such. I see posts stating that it is just science and posts stating that it is just art, I believe it’s both.

 

[amirm]

My point is that we can tell there is a difference but we can’t explain it or even define it.

The problem is that when we setup a test where we 100% know the answer in advance, audiophiles fail to give that answer in controlled testing. For example, I can pretend to switch an amplifier to another one but keep the same amp, the listener will report hearing differences!

The reason is simple: we adapt our hearing all the time. We choose to focus on detail, or not. We choose to search for "black background" or not. We choose to listen for "micro detail" or not. We choose to listen for pitch ("PRAT") or not. Normally when we listen to music, we are not trying to capture this level of information. We are just enjoying it and only noticing high level aspects of fidelity. But when we do comparisons, we are constantly changing our telephoto zooms to see more detail. Problem is, we don't do this consistently across our evaluations, leading to conclusions that things are different when they are not.

Above is why we repeat tests many times to build up statistical confidence in listening tests. We look for patterns that tease out what is really perceived, and what is a function of variability of our hearing and perception.

Until one goes into a test where the answer is known in advance, and provides the same answer, then what they think about accuracy of their hearing is just wrong.

 

[andreasmaaan]

My point is that we can tell there is a difference but we can’t explain it or even define it.

But these blind tests don't ask listeners to explain or define anything. Simply to differentiate.

There's an interesting description given in one paper of the subject who (to my knowledge) had the lowest documented nonlinear distortion audibility threshold of any person ever tested (as low as 0.003% in respect of one musical sample used to test him, although it is not clear what the spectrum of the distortion was - presumably very high-order). This subject was unable to say what sounded different, yet he could still reliably discern the difference under controlled conditions.

Anyway that's just an aside; my point is that controlled tests ask no more than for listeners to discern a difference, any difference at all...

 

[BDWoody]

But these blind tests don't ask listeners to explain or define anything. Simply to differentiate.

There's an interesting description given in one paper of the subject who (to my knowledge) had the lowest documented nonlinear distortion audibility threshold of any person ever tested (as low as 0.003% in respect of one musical sample used to test him, although it is not clear what the spectrum of the distortion was - presumably very high-order). This subject was unable to say what sounded different, yet he could still reliably discern the difference under controlled conditions.

Anyway that's just an aside; my point is that controlled tests ask no more than for listeners to discern a difference, any difference at all...

I wish it wasn't $33...looks like an interesting read.

 

[andreasmaaan]

Amplifiers do sound different, speakers respond differently, we can hear things beyond explanation and it does matter to me.

This is the issue though. If we can hear differences beyond explanation, why does this never almost never happen under controlled conditions?

EDIT: corrected myself. Occasionally surprising outcomes do occur. Although these are usually unrepeatable, and therefore best attributed to flawed experiment design, if the outcome can be repeated, it shows that something in not fully understood, ofc.

 

[andreasmaaan]

I wish it wasn't $33...looks like an interesting read.

In case you're interested, I discussed this study in a bit of depth in post #2 of this thread.

 

[amirm]

I wish it wasn't $33...looks like an interesting read.

I can quote the relevant parts:

Subjects 2, 5 and 3 on Piano could detect such low levels. But we need to be careful. The distortion was computed relative to integration of 250 milliseconds of music. They call that RMS distortion which. RMS distortion is usually the sum of distortion products (square root of sum of all of them squared). This is what we show in our measurements I have not seen it stated as RMS energy of music over that much time as they are doing. They are relying on "integration time of hearing" for that justification. When the ratio is computed relative to just peaks in music, they get:

So wildly different and higher numbers in actual percentages, not a fraction.

The test is the old concern about transient intermodulation distortion (TIM) so a different animal than what we test or care about today.

The source was Vinyl by the way.

 

[BDWoody]

In case you're interested, I discussed this study in a bit of depth in post #2 of this thread.

Wow...Thats a fair bit of work. Very interesting.

 

[Sgt. Ear Ache]

This is a great thread for me right now. I even appreciate the opinions that I do not agree with because they give me insight into a line of thinking. I’ve been playing around with audio equipment for 40 some years but I just recently started getting more serious about it. I’m learning all the testing procedures I can and designing speakers. I’m encouraged by one of my photography and design clients who runs a tube amplifier company. He very much believes in measuring everything, as do I. That being said, I want to explain why I think the measurements still aren’t enough. The measurements are a good place to start but there are many interactions that can’t be easily measured. The idea that at a certain level all equipment is good enough for our ears, is not something I believe. There are qualities to sound that we can perceive but can not define or put into words. My clients tag line is “Just Listen”. This discussion is about amplifiers but the same theoretical concepts are true of every component in a system: cables, DACs, speakers, etc.
I’ve studied the art and science of photography in depth and I strongly believe that people can “perceive” image sharpness and quality beyond what their brains can explain in words. I’ve seen it time and time again. A study of sharpness shows that there is a pin-point area of sharp and a “circle of confusion” around that point. To say that an image is sharp enough because of a measurement at a particular viewing distance is naive. Sharpness is perceived in a very complex way, just as the listening experience is very complex and cannot be boiled down into simple measurements. The same principles apply to motion picture viewing and HiFi listening. The measurements are very important and a great place to start. I feel we should trust measurements more than our perception, but we don’t have measurements for everything that is important to the listening experience. My point is that we can tell there is a difference but we can’t explain it or even define it. A good example is an eye exam: it’s sometimes difficult to tell which is the better test lens, they are different but is one truly better than the other? In a certain situation you will see better with that lens but change the brightness a little and just like volume levels, everything changes.
Amplifiers do sound different, speakers respond differently, we can hear things beyond explanation and it does matter to me. The hobby is based on trying to accomplish the best sound possible in any given situation. I trust testing to get me in the ball park and then I use my ears to see if it’s making me happy. My perception may change over time and a different setup may or may not make me happy. Audiophile listening is both an Art and a Science, and it should be treated as such. I see posts stating that it is just science and posts stating that it is just art, I believe it’s both.

that's all well and good...but what if the differences you think you are hearing are actually an illusion? Nobody is trying to convince anyone that they don't believe they hear what they think they hear. We're suggesting only that they may be being mislead by their hearing - and that suspicion isn't based on just random conjecture. It's based on the fact that there have been many many tests done over the years that show us that we can't necessarily trust what we think our ears are telling us.

 

[BDWoody]

I can quote the relevant parts:

View attachment 29140

Subjects 2, 5 and 3 on Piano could detect such low levels. But we need to be careful. The distortion was computed relative to integration of 250 milliseconds of music. They call that RMS distortion which. RMS distortion is usually the sum of distortion products (square root of sum of all of them squared). This is what we show in our measurements I have not seen it stated as RMS energy of music over that much time as they are doing. They are relying on "integration time of hearing" for that justification. When the ratio is computed relative to just peaks in music, they get:

View attachment 29141

So wildly different and higher numbers in actual percentages, not a fraction.

The test is the old concern about transient intermodulation distortion (TIM) so a different animal than what we test or care about today.

The source was Vinyl by the way.

Thank you.

 

[andreasmaaan]

I can quote the relevant parts:

View attachment 29140

Subjects 2, 5 and 3 on Piano could detect such low levels. But we need to be careful. The distortion was computed relative to integration of 250 milliseconds of music. They call that RMS distortion which. RMS distortion is usually the sum of distortion products (square root of sum of all of them squared). This is what we show in our measurements I have not seen it stated as RMS energy of music over that much time as they are doing. They are relying on "integration time of hearing" for that justification. When the ratio is computed relative to just peaks in music, they get:

View attachment 29141

So wildly different and higher numbers in actual percentages, not a fraction.

The test is the old concern about transient intermodulation distortion (TIM) so a different animal than what we test or care about today.

The source was Vinyl by the way.

I don't disagree that there were reasons that test was problematic. The fact that the test system had higher levels of distortion than those tested (although spectra were not discussed) was a factor, although this is normal and generally very difficult (if not impossible) to avoid in distortion audibility studies due to transducer distortion.

I mulled over the question of peak vs RMS distortion in this study too. I'm in two minds as to its importance.

We know from temporal masking studies that masking continues long after a masker has ceased, and that 200-250ms is generally the auditory system's integration time. I'm not sure how this is to be interpreted in the absence of more information about the transfer function of the distorted signal. Some information is given in the study, but it's incomplete. For example:

And yes, the type of distortion being tested is not at all relevant to the products you are measuring

 

[Sir Sanders Zingmore]

One thing is certain, they do NOT like to listen with their ears,

Why do people need their eyes in order to “trust their ears”?

 

[Sgt. Ear Ache]

Sort of ironic that we would be accused of not liking to listen with our ears when what we are advocating is testing that literally forces us to listen ONLY with our ears. In fact, it's really the subjectivists that don't like to listen with their ears...

 

[Zog]

Unfortunately this posts falls into the offensive category. Combined with the members persistent antagonism and obvious disdain for the membership I chosen to banish him.

Member @daveyf is formally ejected from the premises.

We are here to learn and share knowledge and experience, not for pissing contests and not for pointless arguments over sighted listening fallibility, unless of course those arguments contain some new evidence and may actually progress the discussion.

Cheers

For the record there was nothing he said that antogonised me. Nor did I find his post offensive.

 

[BigRez]

that's all well and good...but what if the differences you think you are hearing are actually an illusion? Nobody is trying to convince anyone that they don't believe they hear what they think they hear. We're suggesting only that they may be being mislead by their hearing - and that suspicion isn't based on just random conjecture. It's based on the fact that there have been many many tests done over the years that show us that we can't necessarily trust what we think our ears are telling us.

I'm 100% in agreement but my post may have been misleading. I don't trust myself to be an impartially judge of what I'm listening to! That's not what I was trying to say at all, I was pointing out that it's more complicated than any test. The tests we have are the best way to measure sound, but we arn't able to measure everything because the interactions are complicated, so there's still some mystery involved in any system. Hopefully equipment that tests well makes you happy as well, if it doesn't, you may never scientifically know why. It could be perception or it could be a real un-measurable phenomenon.

 

[SIY]

Hopefully equipment that tests well makes you happy as well, if it doesn't, you may never scientifically know why. It could be perception or it could be a real un-measurable phenomenon.

Yes to the former.

As to the latter, well, here's a list of all demonstrated auditory phenomena which were not easily measurable:






Quite a list!