[gregorio]

Feel free to enjoy what's in your mind. Cheers.

I will thanks. You on the other hand should NOT feel free to continue to promote pseudoscience, you should of course realise that is an attempted perversion of science and therefore highly insulting in a science discussion forum!

By the way, feel free not to answer the following question:

I would but what’s the point? There’s no point if either you don’t understand it or you just attempt to misrepresent it.

G

 

[sunjam]

I will thanks. You on the other hand should NOT feel free to continue to promote pseudoscience, you should of course realise that is an attempted perversion of science and therefore highly insulting in a science discussion forum!

I would but what’s the point? There’s no point if either you don’t understand it or you just attempt to misrepresent it.

G

Cool, we do agree that people should apply critical thinking and not promote pseudoscience.

 

[gregorio]

Cool, we do agree that people should apply critical thinking and not promote pseudoscience.

If you did agree then obviously you would actually apply critical thinking and you would not be promoting pseudoscience. So either you mean that only other people should apply critical thinking and not promote pseudoscience but not you, or you are not telling the truth. Either way, we do not agree.

G

 

[bigshot]

The problem we seem to always find popping up around here is that people will decide on a conclusion and then go looking for evidence that shows that they're right. They don't let the evidence lead them to a conclusion. Cherry picking and ignoring evidence that doesn't go the direction you want it to go is no way to figure out how things work.

This one is being approached backwards. Instead of looking for a way that a DAC can reproduce sound with high fidelity beyond the point of transparency, he's looking for proof that one kind of DAC and filter is theoretically more accurate than another. But the fact that the DAC and filter that he's talking about isn't audibly transparent doesn't seem to discourage him.

The truth is a typical oversampling DAC with a typical reconstruction filter will produce sound that is audibly perfect to human ears, much better than the option he is trying to champion. Sometimes the simple answer is the best one. Everything doesn't have to be the theory of relativity.

 

[sander99]

Sorry for my ignorance, feel free to think the way you like as I cannot understand what you are talking about

What do you not understand? That 44.1/16 is audibly transparent?

Maybe you thought you heard a difference with "high-res"? Maybe you did a uncontrolled, sighted or not properly level matched listening comparison?
Or maybe you inadvertently compared two different masters (or otherwise differently processed versions) of the same music?
Or maybe you compared using playback equipment that produces audible intermodulation distortion due to ultrasonic content in the "high-res" version (making "high-res" objectively worse but audibly different)?

(Or maybe you compared using a NOS DAC that messes up the 44.1/16 version more than it messes up the "high res" version?)

 

[sunjam]

The problem we seem to always find popping up around here is that people will decide on a conclusion and then go looking for evidence that shows that they're right. They don't let the evidence lead them to a conclusion. Cherry picking and ignoring evidence that doesn't go the direction you want it to go is no way to figure out how things work.

Bingo! You have very sharp mind. That's exactly the "confirmation bias" in Psychology.

Let's find out who is cherry picking and ignoring evidence.

believe 1: "Hi Res is useless"
believe 2: "Hi Res is useful"

Before we proceed, we have to define something objective. (note: "audiby transparent" is a subjective term. It is not good for any definition.)

Let's define "useful" as follow:

A_original is the original analog audio signal
D_CD is the 44/16 digital signal created from A_original
D_Hi_Res is the 768/32 digital signal created from A_original
A_CD is the reconstructed audio signal from D_CD
A_Hi_Res is the reconstructed audio signal from D_Hi_Res

Delta_CD = the difference between A_CD and A_original, (The smaller the better. It is zero for perfect system)
Delta_Hi_Res = the difference between A_HiRes and A_original

if Delta_CD > Delta_HiRes for the same A_original, then Hi Res is useful
if Delta_HiRes >= Delta_CD for the same A_oritinal, then Hi Res is useless

"decide on a conclusion" => one's believe
"Hi Res is useless" <== believe 1
"Hi Res is useful" <== believe 2

"ignoring evidence"
=> "44/16 can reconstruct perfect sine wave as shown in the Monty's videi" <== factual or not? <== question 1
=> "768/32 can recontruct an audio signal better than using 44k/16" <== factual or not? <== question 2
=> "768/32 has less quantization less than 44/16" <=== factual or not? <=== question 3

if question 1 is factual, it support the believe 1 is correct and believe 2 is incorrect <== agree?

if question 2 is factual, it support the believe 2 is correct and believe 1 is incorrect <== agree?

if question 3 is factual, it support the believe 2 is correct and believe 1 is incorrect <== agree?

==================
What's your answers for question 1, 2, and 3?

 

[bigshot]

Your syntax is convoluted, but if I understand you correctly...

"Audibly transparent" is not a subjective term. It means that a difference is so small, it can't be discerned with human ears. That is a binary "yes/no", not a judgment call. The thresholds of perception are determined objectively with controlled listening tests. These thresholds have been documented and understood for a very long time.

1) 44.1 can perfectly reproduce all frequencies within the range of human hearing (Nyquist) / The noise floor of 16 bit is low enough to not be heard in even the most demanding home audio situations even without dithering.

2) Higher sampling rates don't reproduce those audible frequencies any better. Inaudible noise floors and even more inaudible noise floors sound the same. Bigger numbers don't mean more audible resolution in the range of human hearing. Beyond what is necessary, it's just redundant data for no reason.

3) Inaudible error is irrelevant when it comes to playing recorded music in the home, because it is inaudible.

The bottleneck isn't in the digital theory... it's in your all too human ears. Your hearing is finite. Once you surpass the threshold of audibility, nothing more matters because you can't hear it anyway. For the purposes of playing commercially recorded music in the home, CD sound is all you need.

Question: If a tree falls in the forest and no one is there to record the sound, what sampling rate and bit rate is required? Answer: None, because no one is there to record it.

 

[sunjam]

Your syntax is convoluted, but if I understand you correctly...

"Audibly transparent" is not a subjective term. It means that a difference is so small, it can't be discerned with human ears. That is a binary "yes/no", not a judgment call. The thresholds of perception are determined objectively with controlled listening tests. These thresholds have been documented and understood for a very long time.

1) 44.1 can perfectly reproduce all frequencies within the range of human hearing (Nyquist) / The noise floor of 16 bit is low enough to not be heard in even the most demanding home audio situations even without dithering.

2) Higher sampling rates don't reproduce those audible frequencies any better. Inaudible noise floors and even more inaudible noise floors sound the same. Bigger numbers don't mean more audible resolution in the range of human hearing. Beyond what is necessary, it's just redundant data for no reason.

3) Inaudible error is irrelevant when it comes to playing recorded music in the home, because it is inaudible.

The bottleneck isn't in the digital theory... it's in your all too human ears. Your hearing is finite. Once you surpass the threshold of audibility, nothing more matters because you can't hear it anyway. For the purposes of playing commercially recorded music in the home, CD sound is all you need.

Question: If a tree falls in the forest and no one is there to record the sound, what sampling rate and bit rate is required? Answer: None, because no one is there to record it.

Thanks a lot for your reply as it helps to find the real issue of the topic.

You bring a very good point: "inaudible"

If you look at your arguments, you are pretty much using "inaudible" to defend the ideal of "Hi Res is useful". You have "inaudible" for every single point you mentioned in your reply. Agree?

Let's look at a simple analogy, I think "seeing is believing", HDTV vs UHDTV.

Full HD: 1920 x 1080
8k Ultra HD: 7680  ×  4320

Fact: 8k is higher resolution than Full HD

Believe 1: 8k is useless (as Full HD can reconstruct perfect picture on regular TV you can find at home)
Believe 2: 8k is useful (as it can reconstruct better picture on a TV due to its higher resolution)

Argument for people who support the believe 1: your eyes cannot tell the difference between Full HD and 8k source.
Argument for people who support the believe 2: it is a fact that 8k has higher resolution than Full HD

=================================================================

Sounds similar to our discussion about "Hi Res is useless" vs "Hi Res is useful"?

For the TV, are you a supporter for believe 1? or believe 2?
Similarly, for audio, are you a supporter for believe 1? or believe 2?

Note: to me, the "Monty's video" is attempting to prove with "evidence" that "8k is useless" in the audio space.

 

[bigshot]

You didn't hear what I was saying I guess...

For the purposes of playing commercially recorded music in the home high res files are not at all useful. HD audio doesn't have any more resolution than 16/44.1. See the link in my sig file "Bit Rate Is Not Resolution".

According to Nyquist in order to reproduce a frequency PERFECTLY, all you need is two points. You can use a hundred points to do the same thing, but it isn't any better because two points make it PERFECT. 5 kHz is EXACTLY the same in 16/44.1 as it is in 24/96.

TVs are not a good analogy, because you can get up close to a TV and squint at it through a magnifying glass and see more detail. There is no equivalent for sound. Sound is either transparent or it isn't, and human ears can't hear differences between two sources that are transparent BY DEFINITION. You can't squint and hear more. There is a comfortable listening level and you hear what you can hear.

The problem is that you don't understand the most basic thing about digital audio... what sampling and bit rate does and how it works. You've glommed onto a bunch of complicated technical details and you're trying to use them to prove your point without any understanding of the basics. It's like you're building a birthday cake entirely out of icing.

There may be a language problem here. I'll try to state it as clearly as I can.

There is nothing your ears can hear in HD Audio that is different than in 16/44.1. HD Audio doesn't sound better. You don't need to worry about DACs with adjustable filters, because a modern oversampling DAC with a typical reconstruction filter can produce sound that is PERFECT to your ears. There might be tiny errors if you measure them, but you will never ever be able to hear them. 16/44.1 and a modern DAC or player is all you need. There are other things that deserve your attention more.

 

[sunjam]

You didn't hear what I was saying I guess...

For the purposes of playing commercially recorded music in the home high res files are not at all useful. HD audio doesn't have any more resolution than 16/44.1. See the link in my sig file "Bit Rate Is Not Resolution".

According to Nyquist in order to reproduce a frequency PERFECTLY, all you need is two points. You can use a hundred points to do the same thing, but it isn't any better because two points make it PERFECT. 5 kHz is EXACTLY the same in 16/44.1 as it is in 24/96.

TVs are not a good analogy, because you can get up close to a TV and squint at it through a magnifying glass and see more detail. There is no equivalent for sound. Sound is either transparent or it isn't, and human ears can't hear differences between two sources that are transparent BY DEFINITION. You can't squint and hear more. There is a comfortable listening level and you hear what you can hear.

The problem is that you don't understand the most basic thing about digital audio... what sampling and bit rate does and how it works. You've glommed onto a bunch of complicated technical details and you're trying to use them to prove your point without any understanding of the basics. It's like you're building a birthday cake entirely out of icing.

There may be a language problem here. I'll try to state it as clearly as I can.

There is nothing your ears can hear in HD Audio that is different than in 16/44.1. HD Audio doesn't sound better. You don't need to worry about DACs with adjustable filters, because a modern oversampling DAC with a typical reconstruction filter can produce sound that is PERFECT to your ears. There might be tiny errors if you measure them, but you will never ever be able to hear them. 16/44.1 and a modern DAC or player is all you need. There are other things that deserve your attention more.

I believe I hear clearly what you say (and what's in your mind).

The number one supportive fact for your believe is that "it is proven in the sampling theory"

I learned the sampling theory decades ago when I was in college. I am not an expert about the theory but I know what it said.

Most people use it in a "religious way" to defend any discussion about the benefits about Hi-Res. We all know that the sampling theory is proven correct. I am not going to argue regarding its correctness. I truely believe that it is correct too.

However, most people don't aware of its limitations in reality....

Example in your reply:

Your statement "According to Nyquist in order to reproduce a frequency PERFECTLY, all you need is two points. You can use a hundred points to do the same thing, but it isn't any better because two points make it PERFECT. 5 kHz is EXACTLY the same in 16/44.1 as it is in 24/96." is mis-leading as it misses a few key words.

In my eyes, the correct statment should be "According to Nyquist in order to reproduce a frequency PERFECTLY, all you need is two points. You can use a hundred points to do the same thing, but it isn't any better because two points make it PERFECT. 5 kHz is EXACTLY the same in 16/44.1 as it is in 24/96 under the ideal situation."

Agree?

In reality, the ideal situation does not exist. i.e.

1. There is no perfect analog filter.
2. There is quantization noise.

The ideal situation only exists in your imaginary world. This is the thing the "Monty's video" avoid to talk about intentionally.

With these two limitations above, we can see easily 44.1/16 could not reconstruct a perfect sine wave ouput. It is same for Hi-Res, it cannot reconstruct a perfect sine wave too.

Which one would reconstruct better sine wave output?

One may use only two sampled points for the imperfect reconstruction
the other use 256x more sampled points for the same imperfect reconstruction

Which one would be better during the imperfect reconstruction process? My answer is the 256x one.

I agree with you 100% about what you said if the ideal situation exists in reality but it is not.

That's why in my eyes, your statement "According to Nyquist in order to reproduce a frequency PERFECTLY, all you need is two points. You can use a hundred points to do the same thing, but it isn't any better because two points make it PERFECT. 5 kHz is EXACTLY the same in 16/44.1 as it is in 24/96." is mis-leading.

Do you follow my reasoning?

 

[bigshot]

I misspoke. You’re correct. There is theoretical perfection and practical perfection. I was intending to always keep it in context of listening to music on a home audio system in a living room, but I strayed into theory, which I was trying not to do. Theory requires caveats that cloud the actual practical results that can be expected. That’s exactly what’s gotten this fellow into the weeds.

A typical oversampling consumer DAC would produce audible perfection, which ultimately is all that is important to consumers. And HD audio would sound the same as 16/44.1 to human ears. As you say, you have to go out of your way to end up with audible imperfection.

 

[sunjam]

I misspoke. You’re correct. There is theoretical perfection and practical perfection. I was intending to always keep it in context of listening to music on a home audio system in a living room, but I strayed into theory, which I was trying not to do. Theory requires caveats that cloud the actual practical results that can be expected. That’s exactly what’s gotten this fellow into the weeds.

A typical oversampling consumer DAC would produce audible perfection, which ultimately is all that is important to consumers. And HD audio would sound the same as 16/44.1 to human ears. As you say, you have to go out of your way to end up with audible imperfection.

Cool, thanks a lot for your reply. It's good that we are on the same page finally.

I do agree that CD with a good modern DAC is doing a very good job for most people. I've been listening to CD for decades without any complaints.

I don't think people HAVE TO "upgrade" to Hi-Res as most people cannot tell the difference between CD and Hi-Res. Everyone's need is different.
However, there is still a bunch of crazy people (like me?) who are looking at every single corner to improve their systems even the improvement is very small.

It just happened to me that I found that Hi-Res is more enjoyable (as I just upgraded recently). When I shared my experience with other people, many just give me mis-leading info like "Hi-Res is useless" based on questionable claims (e.g. "the Monty's video").

Someone even considered I am mentally blocked as I say "Hi-Res is not useless". LOL...

Anyway, have a good weekend and enjoy your music. For me, while I am enjoying my music, I will keep on updating my blog and share my experiences and view points in helping people to identify misleading claims. Cheers!

 

[bigshot]

It’s fine to go chase down rabbit holes as long as you realize it’s not going to make any audible difference. There is a point beyond which no human can hear, and even inexpensive digital audio gear and CD quality sound far exceeds that. The audible improvement beyond that isn’t small, it’s non existent. No one needs to upgrade beyond what they can hear, and spending a lot of money to do that is a waste of money that would be better spent on music.

I’m not keen on people who use audio equipment as status symbols. And folks who claim to hear the unhearable are often unbearable. I respect people who keep the focus on music, and the ones who solve the problem of audio fidelity efficiently and simply. Convoluted complexity for the sake of it is too common in the audiophile community. If you want to offer advice to other people, efficiency and simplicity is more useful to them than convoluted complexity.

 

[sunjam]

It’s fine to go chase down rabbit holes as long as you realize it’s not going to make any audible difference. There is a point beyond which no human can hear, and even inexpensive digital audio gear and CD quality sound far exceeds that. The audible improvement beyond that isn’t small, it’s non existent. No one needs to upgrade beyond what they can hear, and spending a lot of money to do that is a waste of money that would be better spent on music.

I’m not keen on people who use audio equipment as status symbols. And folks who claim to hear the unhearable are often unbearable. I respect people who keep the focus on music, and the ones who solve the problem of audio fidelity efficiently and simply. Convoluted complexity for the sake of it is too common in the audiophile community. If you want to offer advice to other people, efficiency and simplicity is more useful to them than convoluted complexity.

"I’m not keen on people who use audio equipment as status symbols. And folks who claim to hear the unhearable are often unbearable." <== I understand what you mean. For me, I am using SMSL and iFi Zen DACs (less than $200 each), and Fosi V3 Amp (less than $100). I don't believe in the magic cable.

Regarding audible difference, transparency, and measurements, it is another large topic. I will leave it for next time.

To me, bottom line is that "Hi-Res is useful as it help to better reconstruct the final audio output in reality when compared with CD" (statement 1)

Regardless if someone can hear or cannot hear the difference, statement 1 is still valid.

 

[gregorio]

That's exactly the "confirmation bias" in Psychology.

Oh brilliant, that seems to be your modus operandi. Yes, indeed your quote from Wikipedia is correct, you know what confirmation bias is, so why are you employing it?

Let's find out who is cherry picking and ignoring evidence.
believe 1: "Hi Res is useless"
believe 2: "Hi Res is useful"

Yes, perfect. To demonstrate “who is cherry picking and ignoring evidence” your very first premise is a strawman argument! This does indeed indicate that YOU are cherry picking and/or ignoring evidence, so now you and us have “found out who”, well done!

Before we proceed, we have to define something objective. (note: "audiby transparent" is a subjective term. It is not good for any definition.)

OK, so now you introduce a false fact you have made-up. “Audibly transparent” is definitely NOT a subjective term when determined objectively, IE. When using controlled testing. How is that not self-evident?

Delta_CD = the difference between A_CD and A_original, (The smaller the better. It is zero for perfect system)

This is also false, another untruth you have made-up to support your false assertion! “The smaller the better” is NOT true, in fact we reach a point where the exact opposite is true, IE. “The smaller the worse”!

"ignoring evidence"
=> "44/16 can reconstruct perfect sine wave as shown in the Monty's videi" <== factual or not? <== question 1
=> "768/32 can recontruct an audio signal better than using 44k/16" <== factual or not? <== question 2
=> "768/32 has less quantization less than 44/16" <=== factual or not? <=== question 3
if question 1 is factual, it support the believe 1 is correct and believe 2 is incorrect <== agree?
if question 2 is factual, it support the believe 2 is correct and believe 1 is incorrect <== agree?
if question 3 is factual, it support the believe 2 is correct and believe 1 is incorrect <== agree?

Again yes indeed, a good example of ignoring evidence. All the above assertions and the answers depend on the definition of “perfect, which you are ignoring! Eg. “Perfect” to human ears (all imperfections are beyond the threshold of audibility) or “perfect” in an absolute sense (at least to the limits of our ability to measure it). For all consumers listening to the reproduction of commercial audio recordings (music or sound) at reasonable levels, then the first definition is the only one which is relevant and therefore the answers to your questions are:
Q1 = True.
Q2 = False. They are both the same, the reconstruction imperfections of both formats are well below the threshold of audibility.
Q3 = False. They both have quantisation and the resulting noise is inaudible in both cases.
The answers above obviously assume a DAC that is NOT broken!
So according to your criteria, Hi-res is in fact useless!

Let's look at a simple analogy, I think "seeing is believing", HDTV vs UHDTV.

You state “let’s look at a simple analogy” but then you present an example which is NOT a simple analogy and in fact it is not even analogous at all!

Note: to me, the "Monty's video" is attempting to prove with "evidence" that "8k is useless" in the audio space.

Maybe that’s true but only the “to me” part, because you are NOT applying critical thinking, contrary to your claims! The actual fact is that Monty’s Video does not even mention 8k video resolution and is not analogous to video resolution.

One may use only two sampled points for the imperfect reconstruction
the other use 256x more sampled points for the same imperfect reconstruction
Which one would be better during the imperfect reconstruction process? My answer is the 256x one.

Great, a complete absence of logic. Your point is again a strawman but even taking it as written, your answer is nonsense! If it’s the “same imperfect reconstruction” for both, then obviously the logical answer would be “two sampled points” because why would anyone want 256x more data for the same result?!

I do agree that CD with a good modern DAC is doing a very good job for most people.

You keep doing that, you make an assertion about critical thinking, learning technique, not promoting pseudoscience, not being subject to confirmation bias or ignoring evidence and then you do the exact opposite! In this case, you say you agree but then you don’t! Only you are claiming “a very good job for most people”, no one else is making that claim or agreeing with it, so how are you agreeing?

I don't think people HAVE TO "upgrade" to Hi-Res as most people cannot tell the difference between CD and Hi-Res. Everyone's need is different.

Firstly, that is again a falsehood, it is not “most people”. Secondly, Hi-res is NOT an upgrade, in fact as far as creating audio recordings are concerned the sample rates you have mentioned would in most cases actually be a serious downgrade!

However, there is still a bunch of crazy people (like me?) who are looking at every single corner to improve their systems even the improvement is very small.

Yep, there are certainly a bunch of crazy audiophiles out there who appear completely suckered by the false marketing, who’ve been fooled into believing that a cross-grade or even downgrade is actually an upgrade and/or that even if there is actually a very small improvement in one corner which is inaudible or cannot be even be resolved into sound, it’s actually an audible improvement.

Is that “like me?” (you)? As all you’ve provided is cherry picked evidence, falsehoods, fallacies and therefore pseudoscience and, if that’s not already more than bad enough, you’ve actually attempted that in a science discussion forum of all places, then what do you think?

To me, bottom line is that "Hi-Res is useful as it help to better reconstruct the final audio output in reality when compared with CD"

Again, Exactly! “To you” that is the bottom line because you are ignoring evidence, making-up falsehoods and NOT applying critical thinking. What do you think is going to happen if you keep doing that in a science discussion forum?

G