Head-Fi.org › Forums › Equipment Forums › Sound Science › 320 kbps MP3 vs. normal audio CD listening Sound quality
New Posts  All Forums:Forum Nav:

320 kbps MP3 vs. normal audio CD listening Sound quality - Page 18

post #256 of 516

Quote:

Originally Posted by jvandyk View Post


C'mon Mr H. Already posted about how horrible the 256k version of Sunken Condos sounded vs the 24 bit version. Do I need to go back and check myself on that one? Or should I goon about how badly the HD tracks 24 bit Damn The Torpedoes trashes the Mobile Fidelity CD? And I already mentioned Gaucho.
I cannot comment on those examples but I'll offer something else.  Way back in the late 1980s masters and Clark did a very interesting set of tests in an article called "Do All Amplifiers Sound the Same?" They assembled several audio enthusiasts and tested their ability to detect the differences between pairs of amplifiers ranging from $230 Pioneer receivers up to $12,000 Monoblocks. In the sighted test part several listeners opined that they could easily tell pairs of amps apart and described in great detail the differences. When the tests were repeated blind the listeners could no longer tell them apart. The certainty that many of the sighted listeners had did not mean they could hear differences without knowing what they were listening to. Yet nothing had changed between the two sessions apart from the blinding. Look on the web and you will find many many such examples where utter certainty under sighted conditions vanishes away when the awareness of what you are listening to is removed.
We need blind tests to keep us honest. Sometimes we just imagine differences that are not there. My favorite example of such self-deception is the guy comparing his beloved Bryston amp to an Onkyo. He described the merits of the Bryston and all the failings of the Onkyo in great detail (thin, grainy, distorted)  then his pal turned off the Onkyo and the music kept on playing, he was still listening to his Bryston. There are accounts of cable swaps which were not swaps yet differences were still heard. When we think a change has been made we tend to hear it regardless of how real the change is. With a short search you can find loads of similar examples. Blind tests help us to face reality and are very sensitive. In this forum we have members who can hear differences (sometimes and with some tracks) between lossy and lossless and have used DBTs to prove it. I can (blind) hear the differences between my CD players (one is 0.7db louder than the other)  or the effect of low pass filters (at some frequencies) .
 

Edited by nick_charles - 12/28/12 at 2:58pm
post #257 of 516
Quote:
Originally Posted by jvandyk View Post

Im just curious, as most of you suggest using foobar to do blind testing. Foobar cannot play 16 bit and 24 bit back to back in a bitperfect fashion period. Everybody keeps asking me to do this, but I would assume everybody is first manipulating the 16 bit tracks first with dither/ software before their tests. And this would also require upsamling to work in foobar.

 

What is wrong with converting the 16 bit sample back to 24 bit (which is a simple lossless operation of padding the sample with zero bits), and resampling as well from 44.1 kHz back to the original higher sample rate, if necessary ? Once a high resolution sample is converted to 44.1/16 format, the loss of information is irreversible, and converting it to 96/24 (or whatever else it originally was) again will not magically restore it. So, you cannot use that as an excuse when you fail to hear a difference.


Edited by stv014 - 12/28/12 at 2:40pm
post #258 of 516
Quote:
Originally Posted by Nirvana Woman View Post

http://www.head-fi.org/t/415361/24bit-vs-16bit-the-myth-exploded

 

In short: 24 bit isn't actually able to make a difference with modern hifi equipment

You might want to let the manufacturers of 24 bit DACS know that! And while you're at it, let the software manufacturers know too..those misled folks who provide upsampling and such.

post #259 of 516
Quote:
Originally Posted by jvandyk View Post

You might want to let the manufacturers of 24 bit DACS know that! And while you're at it, let the software manufacturers know too..those misled folks who provide upsampling and such.

You can read, right? The extra bits can be used e.g. for digital volume control, and upsampling = processing and processing <> playback.

 

Btw, for vinyl recordings you wouldn't need more than 13 bits.


Edited by xnor - 12/28/12 at 2:46pm
post #260 of 516
Quote:
Originally Posted by stv014 View Post

 

What is wrong with converting the 16 bit sample back to 24 bit (which is a simple lossless operation of padding the sample with zero bits), and resampling as well from 44.1 kHz back to the original higher sample rate, if necessary ? Once a high resolution sample is converted to 44.1/16 format, the loss of information is irreversible, and converting it to 96/24 (or whatever else it originally was) again will not magically restore it. So, you cannot use that as an excuse when you fail to hear a difference.

Because then you're not using the original file. Upsampling 16 bit tracks often results in a different sound ( I know, not in this forum it doesn't). But in the rest of the audiphile world, upsampling is often desired to improve the sound of 16 bit collections. It has been reviewed by every major audio mag as a way to improve quality.

 

But I don't expect the SS thread to give that any credence.

 

BTW, this all started when I was comparing the 256k version of the Fagen to the 24 bit. I suggested the 24 bit material sounded very good...What I wasn't trying to say was that 16 bit lossless was easily differentiated from 24 bit. I will admit right now that comparing the uncompressed 16 bit CD to the 24 bit version would be a much smaller difference indeed. But I also don't need a blind test to tell you the 256k version is inferior in multiple ways.

post #261 of 516
Quote:
Originally Posted by xnor View Post

You can read, right? The extra bits can be used e.g. for digital volume control, and upsampling = processing and processing <> playback.

 

Btw, for vinyl recordings you wouldn't need more than 13 bits.

Patronizing again. it's so common in this forum. But I guess, based on your response, that you admit 24 bit can sound alot better than 16 bit using your example of digital volume control or EQ without losing resolution. In fact, the better software available uses 64 bit floating processors and can effectively provide digital volume control without loss of resolution. Digital EQ can also be done with larger bit depths with loss of resolution. All of this, controllable by the end user in playback.

 

BTW, Foobar cannot do the above. You'll need at least JRiver or better.

post #262 of 516
Quote:
Originally Posted by nick_charles View Post

 

 

Quote:

Originally Posted by jvandyk View Post


C'mon Mr H. Already posted about how horrible the 256k version of Sunken Condos sounded vs the 24 bit version. Do I need to go back and check myself on that one? Or should I goon about how badly the HD tracks 24 bit Damn The Torpedoes trashes the Mobile Fidelity CD? And I already mentioned Gaucho.
I cannot comment on those examples but I'll offer something else.  Way back in the late 1980s masters and Clark did a very interesting set of tests in an article called "Do All Amplifiers Sound the Same?" They assembled several audio enthusiasts and tested their ability to detect the differences between pairs of amplifiers ranging from $230 Pioneer receivers up to $12,000 Monoblocks. In the sighted test part several listeners opined that they could easily tell pairs of amps apart and described in great detail the differences. When the tests were repeated blind the listeners could no longer tell them apart. The certainty that many of the sighted listeners had did not mean they could hear differences without knowing what they were listening to. Yet nothing had changed between the two sessions apart from the blinding. Look on the web and you will find many many such examples where utter certainty under sighted conditions vanishes away when the awareness of what you are listening to is removed.
We need blind tests to keep us honest. Sometimes we just imagine differences that are not there. My favorite example of such self-deception is the guy comparing his beloved Bryston amp to an Onkyo. He described the merits of the Bryston and all the failings of the Onkyo in great detail (thin, grainy, distorted)  then his pal turned off the Onkyo and the music kept on playing, he was still listening to his Bryston. There are accounts of cable swaps which were not swaps yet differences were still heard. When we think a change has been made we tend to hear it regardless of how real the change is. With a short search you can find loads of similar examples. Blind tests help us to face reality and are very sensitive. In this forum we have members who can hear differences (sometimes and with some tracks) between lossy and lossless and have used DBTs to prove it. I can (blind) hear the differences between my CD players (one is 0.7db louder than the other)  or the effect of low pass filters (at some frequencies) .
I am not new to the concept of blind testing. But in our world where the new itunes 256k version of Sunken Condos sounds like crap, you are wasting your breath.
 
 
 
Btw, the analog vinyl sounds better than any of them. Ready to go on that one?
I suggest glancing at my link from earlier and maybe considering some subjective opinions. A/b blind testing exposes the obvious but is truly subjective because the listener is focused on only the obvious. Listening over time reveals the finer things in life. Just my opinion. If I hear a click or a pop on a record, it biases any possible blind test....but the vinyl still sounds better in the long run. Same with 24 bit vs 16 bit. And vs lossy mp3.... The obvious.
post #263 of 516

As long as we're having so much fun, here is a good link from one of the most respected DAC Enginneers in the industry. It backs up the debate against me and also my own points.

http://www.thewelltemperedcomputer.com/KB/Bit1624.htm

 

Interestingly, all this business about decay and noise floor is pretty relevant. If one listener thinks the 24 bit audio sounds more natural, but they cannot put their finger on it, it is because they are hearing the finest details of decay and the audible sound spectrum. And todays audio equipment does reproduce all of this.

 

24 bit audio is superior technically during playback..at least in the very fine details.


Edited by jvandyk - 12/28/12 at 3:14pm
post #264 of 516

Are you kidding?

post #265 of 516
Quote:
Originally Posted by jvandyk View Post

As long as we're having so much fun, here is a good link from one of the most respected DAC Enginneers in the industry. It backs up the debate against me and also my own points.

http://www.thewelltemperedcomputer.com/KB/Bit1624.htm

 

 

 

That is one of the most dishonest audio-related web pages I have ever seen !

The guy deliberately chooses one of the worst DACs ever tested by Stereophile and presents it as a typical example of 16 bit linearity !

post #266 of 516
Quote:
Originally Posted by nick_charles View Post

 

That is one of the most dishonest audio-related web pages I have ever seen !

The guy deliberately chooses one of the worst DACs ever tested by Stereophile and presents it as a typical example of 16 bit linearity !

Nah, not really. In fact, he is one of the engineers for DCS (regarded as one of the finest DAC ranges in the world). 

post #267 of 516

try closer reading, which points out that 24 bits is not necessary for the headroom - post processing

 

and he fails to credit psychoacoustic noise shaped dither of 16 bits with delivering perceived noise floor and linearity below 16-bit's lsb - linearity basically to your DAC's limit

 

and today PC sound software can recognize, use a 24 bit DAC to "preserve the bits" of a 16 bit source stream with applied local EQ, digital volume calculations giving "more bits", sending the top 24 to the DAC

 

redithering (while pointless at 24 bits, Johnson noise in your electronics is higher) can be done too in most of today's digital audio hardware

 

even low power 16 bit uC in DAP can redither after digital volume, EQ - see RockBox code - but many now have 24bit DAC 


Edited by jcx - 12/28/12 at 6:22pm
post #268 of 516
Quote:
Originally Posted by jvandyk View Post

Interestingly, all this business about decay and noise floor is pretty relevant.

 

Do you have any idea how loud your stereo would have to be to be to be able to hear sound below -96dB? We're talking threshold of pain and hearing damage. No one listens to music like that.

 

I'll spare you the pie chart, but keep in mind that when you paraphrase from websites, it helps if you know what those numbers mean. It also helps to paraphrase from sites of people who know what they're talking about. That guy isn't a DAC engineer. He is an internet equipment reviewer on a web forum. He might as well be a chimp with those credentials.

 

Edit: I just found why you think he works for DCS. There is a footnote at the bottom of one of his pages that links to a sales pamphlet on digital audio by DCS. That is a footnote. He isn't an engineer for them.

 

Honestly, if you ever hope to know anything about digital audio, you are going to have to open your mind and ears a bit and work on understanding the principles involved here. It isn't that hard. It's actually kind of fun, and armed with that information, you can sort out the chimps with a website from the folks like Nick who actually know a great deal about the subject.


Edited by bigshot - 12/28/12 at 3:48pm
post #269 of 516
Quote:
Originally Posted by jcx View Post

try closer reading, which points out that 24 bits is not necessary for the headroom - post processing

 

and he fails to credit psychoacoustic noise shaped dither of 16 bits with delivering perceived noise floor and linearity below -96 dB - basically to your DAC's limit

 

and today PC sound software can recognize, use a 24 bit DAC to "preserve the bits" of a 16 bit stream with local EQ, digital volume applied

 

redithering (while pointless at 24 bits, Johnson noise in your electronics is higher) can be done too in most of today's digital audio hardware

 

even low power 16 bit uC in DAP can redither after digital volume, EQ - see RockBox code - but many now have 24bit DAC 

Like I said, it points out both areas of the debate. I also read this closely-

 

"Can we expect the same when moving from 16 bit (CD audio) to 24 bit?
If we look at the numbers, the answer is yes.
16 bit integer allows for 2^16= 65536 different values
24 bit integer has 2^24= 16777216 different values, a 256 times better resolution!
No doubt, this must be a substantial audible difference."

 

And this-

"Playback

The headroom argument is no longer valid, one knows the maximum level when producing the final master.

 

Noise floor: your gear must have an S/N better than 96 dB otherwise the extra bits 24 offers will be drowned in the noise. But most gear does.

 

DSP: if you use digital volume control or any other kind of DSP e.g. re-sampling, you can profit by using 24 bits.
In case of 16 bits the result must be dithered otherwise the artifacts of the DSP become audible.
Again 24 bits has the advantage because of the quantization error.
If you have hardware supporting 24 bit words, padding 16 bits audio with 8 bits does the job too."

 

Ok guys, don't mean to "try to exit the conversation" as another poster mentioned...but it's Millertime in So Cal.

post #270 of 516
Quote:
Originally Posted by bigshot View Post

 

Do you have any idea how loud your stereo would have to be to be to be able to hear sound below -96dB? We're talking threshold of pain and hearing damage. No one listens to music like that.

 

I'll spare you the pie chart, but keep in mind that when you paraphrase from websites, it helps if you know what those numbers mean. It also helps to paraphrase from sites of people who know what they're talking about. That guy isn't a DAC engineer. He is an internet equipment reviewer on a web forum. He might as well be a chimp with those credentials.

 

Edit: I just found why you think he works for DCS. There is a footnote at the bottom of one of his pages that links to a sales pamphlet on digital audio by DCS. That is a footnote. He isn't an engineer for them.

 

Honestly, if you ever hope to know anything about digital audio, you are going to have to open your mind and ears a bit and work on understanding the principles involved here. It isn't that hard. It's actually kind of fun, and armed with that information, you can sort out the chimps with a website from the folks like Nick who actually know a great deal about the subject.

Uhm. Wrong. He works for DCS. But that's OK...cheers.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › 320 kbps MP3 vs. normal audio CD listening Sound quality