• 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!

Master Thread: Are measurements Everything or Nothing?

I am speechless, I am writing a serious answer to a very relevant question.
I get called a troll, kicked out and then further belittled.

I have six years of electronics education, more than 20 years experience as a hifi customer, but not least I am also a Diy practitioner and have built and constructed hifi from scratch, amplifiers, DACs , speakers and of course cables, made lots of modifications, so no, I am not affected by the fact that I have spent a lot of money on hifi and storytelling.

I only tried to answer the thread starter's question from my experience
It does not take much to experience something mysterious, connect a 2nF and 4.7uF in parallel, if the 2nF is of a different type, you will find that it can give the overall sound a different sound signature, or change a coupling or decoupling capacitor or a crossover capacitor it can also give a different sound signature, I would say that as a diy practitioner it is very difficult not to hear and detect sound differences that should not be possible according to theory.

Is ASR a cult where you get kicked out if you don't follow the manifesto that an AP Analyser and blind test tells all

Electronic knowledge does not help me in the above cases to explain what I hear, the only thing I can use my knowledge for is to be 100% sure that I should not hear the differences I hear on cables for example, but apparently I do anyway even in the blind test.

Furthermore, in my opinion, there are no measurements for much of what we hear and emphasise when we listen and evaluate hifi. Something I elaborated on in my post.

If anyone knows what an AP analyser can and can't do it's Jonathan Novick, he covers that topic in this video.
Note at 57:14 mins the first person to grab a microphone and talk is our host @amirm . :cool:
 
I am speechless, I am writing a serious answer to a very relevant question.
I get called a troll, kicked out and then further belittled.

I have six years of electronics education, more than 20 years experience as a hifi customer, but not least I am also a Diy practitioner and have built and constructed hifi from scratch, amplifiers, DACs , speakers and of course cables, made lots of modifications, so no, I am not affected by the fact that I have spent a lot of money on hifi and storytelling.

I only tried to answer the thread starter's question from my experience
It does not take much to experience something mysterious, connect a 2nF and 4.7uF in parallel, if the 2nF is of a different type, you will find that it can give the overall sound a different sound signature, or change a coupling or decoupling capacitor or a crossover capacitor it can also give a different sound signature, I would say that as a diy practitioner it is very difficult not to hear and detect sound differences that should not be possible according to theory.

Is ASR a cult where you get kicked out if you don't follow the manifesto that an AP Analyser and blind test tells all

Electronic knowledge does not help me in the above cases to explain what I hear, the only thing I can use my knowledge for is to be 100% sure that I should not hear the differences I hear on cables for example, but apparently I do anyway even in the blind test.

Furthermore, in my opinion, there are no measurements for much of what we hear and emphasise when we listen and evaluate hifi. Something I elaborated on in my post.

If anyone knows what an AP analyser can and can't do it's Jonathan Novick, he covers that topic in this video.
But we still have the same problem: if you or someone else can perceive a real difference, then it will be a frequency more amplified than other, or a ring that give that sound an effect of being enhanced (or attenuated).

I personally made some research after sustaining the same argument: spectral analysis of an electrical signal is orders of magnitude more precise than human hearing even the best -15 or -20 dB ever measured in clinical studies.

So the only explanation rest on variables that had not been considered, or in psychoacoustic phenomena, I can’t find any other path.

As a signal is a voltage function of time, is difficult to think on non-measurable variables, one falls into negation of the definition: a signal is more than a voltage function of time.

For me is a nonsense, but I’m not an expert in electronics
 
Is this your 7th shadow account?
This is about right +/-. We know this and we decided to give him some rope. Now he can use that rope to pull himself out of the hole he has dug. Or he can do the other thing with it. So we are extending an olive branch by leaving this account active. ;) For now…
 
This is about right +/-. We know this and we decided to give him some rope. Now he can use that rope to pull himself out of the hole he has dug. Or he can do the other thing with it. So we are extending an olive branch by leaving this account active. ;) For now…
You are the nicest moderator in the world :) this kind of behavior should in my book ban you from the whole of internet forever .
 
I am not a scientist and I am sure that the test does not follow all the rules, I am the member you are looking for who wants to discuss hifi and contribute with experiences.

I have been involved in so many tests that I am 100% convinced that there are sound differences that cannot be confirmed with theory and measurement. I have given examples, try it.
I personally don’t care if you can hear the difference between one capacitor and another of the same value, if you’re happy with your convictions go on…

But don’t try to convince the others: we are here to learn from science, “non measurable and non explained by theory” are more things to tell in another forum which has no “science” written on his title.

Apart from the fact that acoustics and electronics are well known stuff, we’re not talking about the first milliseconds after Big Bang
 
more than 20 years experience
I only tried to answer the thread starter's question from my experience
It does not take much to experience
apparently I do
Furthermore, in my opinion,

Experience - by itself - means absolutely nothing if there is no understanding to explain what the experience actually is. A witch doctor can have 20 years experience casting out evil spirits. That doesn't mean that he understands or can explain disease.

Saying "apparently" also doesn't have any meaning. "Apparently" means, "... "to all appearances" (but not necessarily "really") ...". Notice that last qualifier; (... but not necessarily "really")

When you say, "... in my opinion...", that is not a statement of fact nor authority. Opinion is simply, " ... "a judgment formed or a conclusion reached, especially one based on evidence that does not produce knowledge or certainty, ...".

People desperately need to understand the difference between objectivity and subjectivity.

Objectivity (and its partner, "science") is a method to understand the universe around us in a logical, dispassionate manner. For the majority of the past, voodoo, superstition and esoteric cults sought to explain and understand the world around us. They were full of passion (emotion) and desire, but for the most part, other than astronomy, they did not have any ability to predict the outcome of ongoing processes .... much less the future.
OTOH, the Scientific Method was developed slowly, over millennia, from the Smith and Ebers papyri on up to the present day, as a method of inquiring and observing. It used rational thinking to understand and categorize the processes of the world around us. It has the advantage of being able to predict the outcome of ongoing processes. (If you mix 1 cup of ammonia with 1 cup of chlorine bleach, you WILL get an offgassing of chlorine gas ... and you'd best get out of the room! :eek: )

Subjectivity, OTOH, is based on experience (not observation) and emotion (not rational thinking). Subjectivity is a huge part of our lives; it is the result of the biases that contribute so effectively to our survival ... biases that control our reactions faster than we can think about them.
But subjectivity is not rational. It is not dispassionate (just the opposite; it feeds on emotion) and it certainly is not logical. Through the action of many biases, subjectivity can be used to fool our brain time and time again, through illusions and through assumptions.
That makes using it for accurately predicting the outcome of any certain process almost impossible. Subjectivity is not replicable (or repeatable or reproducible; different people use different terms.) Whereas the language of objectivity is definite and measurable, the language of subjectivity is ambiguous and indefinite, and not able to be accurately measured.

So ... feelings, experiences, opinions and appearances are subjective ... and not useful for repeatability. They cannot be used to transfer information from one person to another with any accuracy.

OTOH ... logic, observations, deductions and reprodicible tests and measurements are objective ... and useful to me, to you, to anyone in the world, yesterday, today and tomorrow.

This site was founded to use objectivity to ferret out useful (and dispassionate) information in audio using science-based methods. That doesn't mean that we do not have opinions ... we do. We have personal opinions. but those opinions are not proselytized.

Jim
 
Last edited:
I am not a scientist and I am sure that the test does not follow all the rules,

In that case I'm sorry to say that it holds no value whatsoever.

The claim of cables having a sound is so preposterous from a technical point of view, that it takes some extremely solid evidence to make it less so.

It needs be obtained from a rigorous test with controls that shows statistical significance, and at the same time follows a clear protocol making it possible for other people to replicate the test in order to further cement said statistical significance.

I am the member you are looking for who wants to discuss hifi and contribute with experiences.

I know. It didn't take a huge amount of brain power to figure that out :D

I have been involved in so many tests that I am 100% convinced that there are sound differences that cannot be confirmed with theory and measurement.

Sounds more like conviction than knowledge to me. Am I wrong in assuming that no matter what argument or evidence is being presented to you by other members of this forum, it will have absolutely no chance of changing you mind?

I have given examples, try it.

No thanks.

It's the same kind of "argument" you'll see time and time again when people try to sell you homeopathy and other silly magical concepts. "Try it for yourself and see" :facepalm:

You probably won't find a single person here, who'll accept that as proof of anything.
 
In that case I'm sorry to say that it holds no value whatsoever.

The claim of cables having a sound is so preposterous from a technical point of view, that it takes some extremely solid evidence to make it less so.

It needs be obtained from a rigorous test with controls that shows statistical significance, and at the same time follows a clear protocol making it possible for other people to replicate the test in order to further cement said statistical significance.



I know. It didn't take a huge amount of brain power to figure that out :D



Sounds more like conviction than knowledge to me. Am I wrong in assuming that no matter what argument or evidence is being presented to you by other members of this forum, it will have absolutely no chance of changing you mind?



No thanks.

It's the same kind of "argument" you'll see time and time again when people try to sell you homeopathy and other silly magical concepts. "Try it for yourself and see" :facepalm:

You probably won't find a single person here, who'll accept that as proof of anything.
Hehe, “audihomeopathy”: I think we have found a good word to magical and unmeasurable properties of electronic components…
 
Hehe, “audihomeopathy”: I think we have found a good word to magical and unmeasurable properties of electronic components…
Including alleged effects borrowed from transmission lines, etc. Although Teo's liquid cables remains my favorite cable nonsense.

Some fun reading:

 
As I understand it, the human ear can detect and decipher "useful" information (such as speech) well below the noise floor.
That's not how I understand that. AFAIK some guys that know Morse code are capable of detecting tones well below the noise floor and guys listening with hydrophones can 'hear' things that untrained folks can't.

Also.... it depends on the spectrum of the noise and the spectrum of what is in the noise floor if you can hear things in a noise floor.
 
Including alleged effects borrowed from transmission lines, etc. Although Teo's liquid cables remains my favorite cable nonsense.

Some fun reading:

Was more funny reading Teo Audio poetry on cables, I didn’t know that pearl of literature.

Hope Elon Musk installed some of those on Starlink: if one is going to Mars must do it with style :cool:
 
As someone who uses and respects well executed tests (as in ASR) and who puts the premium on fairly represented data, I have a question, if possibly a naive one. As I understand it, the human ear can detect and decipher "useful" information (such as speech) well below the noise floor. I also believe that most of us instinctively assume that all the goodness (useful program material) in a test lives solely "above the line*" and that everything below the line can be safely ignored. I know I tend to. But, is it possible that one source of the tension between the objective absolutist and the moderate subjectivist can be found lurking in the space below the line - in the quality of the noise (or distortion products?) and the means employed by the designer to lower the line. An obvious examples in the distortion domain: given two amplifiers with identical THD results, which one would likely sound better, the one with predominantly odd harmonic distortions or the one with mostly even harmonics? Which is better in a stressed output stage, hard clipping or compression? Just a thought, not a credo.

* substitute the relevant axis as necessary or typical
That is over-simplified. If you have one noise number for 0 to 20 khz, that is fine. That is the line we can hear below. So can FFTs and for similar reasons. Our ears have the ability to slice up the spectrum into maybe 30 or so bands like a spectrum analyzer. Just as a 32 bin FFT can measure within each bin below the total bandwidth noise floor, so can our hearing. You might want to read up on ERB or effective rectangular bandwidth of hearing. Of course we can have a 1 million bin FFT showing what is far below that single noise number or the level at which our ears can hear.

So it is a mistake to think the noise floor sets a line and everything audible is above that line. Also a mistake to think everything below that composite wide band noise number is something you can ignore.

Most natural noise is pink, meaning higher at low frequencies and lower levels at high frequencies. Our hearing is of different sensitivity at different frequencies, being most sensitive around 3-5 khz. So one should be careful about noise. Sure, if the total noise level is low enough we can ignore it, like say -120 dbr. Otherwise one needs to know a bit more than the one number.
 
That is over-simplified. If you have one noise number for 0 to 20 khz, that is fine. That is the line we can hear below. So can FFTs and for similar reasons. Our ears have the ability to slice up the spectrum into maybe 30 or so bands like a spectrum analyzer. Just as a 32 bin FFT can measure within each bin below the total bandwidth noise floor, so can our hearing. You might want to read up on ERB or effective rectangular bandwidth of hearing. Of course we can have a 1 million bin FFT showing what is far below that single noise number or the level at which our ears can hear.
Of course it's simplified, but I assumed the readers of this forum would already be aware that the human 'ear' (with inherent limitations) can capture correlated sounds (like music) even when uncorrelated sounds (noise) predominate. Thus, the perceived noise floor may not map one to one with what a given test regime assumes the floor to be. So, two devices under test might sound differently while measuring essentially identical, when the test is structured to mask, remove, or avoid the sorts of information our hearing is quite adept capturing - for example music and speech.

So it is a mistake to think the noise floor sets a line and everything audible is above that line. Also a mistake to think everything below that composite wide band noise number is something you can ignore.
Exactly my point. Regardless of the shape of the lines, some will see like them as if Harry Olson had brought them down from Mount Pleasant, Iowa. Truth is very fuzzy and insisting that today's SOTA science is unassailable only stiffens the backs of the religiously subjectivist.

Most natural noise is pink, meaning higher at low frequencies and lower levels at high frequencies.
One of the prime attributes of noise, regardless of it color, is lack of correlation, particularly of correlated events. That's why noise is useful for some tests. Music, on the other hand, is highly correlated in several parameters (pitch, rhythm, timbre, etc.). Consider, that the sorts of sounds we want our systems to pump out will be correlated, unless we want to listen solely to wind, steam boilers, etc. ;) If someone can come up with an objective (scientific?) test methodology based on music (or music-like signals), they would qualify for sainthood.

In the meantime, back to Mahler.
 
Of course it's simplified, but I assumed the readers of this forum would already be aware that the human 'ear' (with inherent limitations) can capture correlated sounds (like music) even when uncorrelated sounds (noise) predominate. Thus, the perceived noise floor may not map one to one with what a given test regime assumes the floor to be. So, two devices under test might sound differently while measuring essentially identical, when the test is structured to mask, remove, or avoid the sorts of information our hearing is quite adept capturing - for example music and speech.


Exactly my point. Regardless of the shape of the lines, some will see like them as if Harry Olson had brought them down from Mount Pleasant, Iowa. Truth is very fuzzy and insisting that today's SOTA science is unassailable only stiffens the backs of the religiously subjectivist.


One of the prime attributes of noise, regardless of it color, is lack of correlation, particularly of correlated events. That's why noise is useful for some tests. Music, on the other hand, is highly correlated in several parameters (pitch, rhythm, timbre, etc.). Consider, that the sorts of sounds we want our systems to pump out will be correlated, unless we want to listen solely to wind, steam boilers, etc. ;) If someone can come up with an objective (scientific?) test methodology based on music (or music-like signals), they would qualify for sainthood.

In the meantime, back to Mahler.
Try Deltawave. Tests using music for signals.

You also seem to think our ears act differently with noise. They don't really.
 
Try Deltawave. Tests using music for signals.

You also seem to think our ears act differently with noise. They don't really.
Thanks for the Deltaware pointer and no I don't think they work differently, I was using "ears" as shorthand for the whole human listening system. Sorry if that wasn't clear in context.
 
Thanks for the Deltaware pointer and no I don't think they work differently, I was using "ears" as shorthand for the whole human listening system. Sorry if that wasn't clear in context.

The thread is long so maybe just trying it after downloading and then looking near the end of the thread makes sense.
You can download it for free here:

Also something of a testing suite including using multiple tones you can construct any way you wish:
 
Also.... it depends on the spectrum of the noise and the spectrum of what is in the noise floor if you can hear things in a noise floor
Plenty of RF modulation schemes work with signals way under the noise floor. For instance LoRa can modulate upto 20 dB below the noise floor. Obviously this pales in comparison to Rob Watts skills :rolleyes:
 
Yep, but that's a CW that can easily be isolated which speech is not. Because noise is wideband and the modulated CW is very narrow the noise energy in that band is low.
That works differently than hearing.
 
That works differently than hearing
Obviously. It’s random sounds vs something specifically designed to work below noise floor levels.

But relatively speaking LoRa is actually wide bandwidth, so very different from CW.
 
Yep but very low bit rate which makes this possible.
 
Back
Top Bottom