High resolution music vs mp3.
Jan 28, 2015 at 6:19 PM Post #16 of 42
goodvibes said:
  Time to take it to sound science.
smile.gif
Click to expand...

 
Yeah, well, we can't conduct a proper investigation otherwise, can we?
biggrin.gif

 
Jan 28, 2015 at 6:19 PM Post #17 of 42
I believe the more headroom in the high resolution sound allow for more instruments, notes,tones to be played and recorded. I don't think there is a point to test listening high resolution from a track of 3-4 instruments with wasted space...it will result in more noises when recorded at 24/96, and most of the time these noises can not be heard unless you pump the music loud enough to blow out your ears. Hence I will commit finding a better recording sound track to test it myself.
 
Jan 28, 2015 at 6:26 PM Post #18 of 42
whitigir said:
I believe the more headroom in the high resolution sound allow for more instruments, notes,tones to be played and recorded. I don't think there is a point to test listening high resolution from a track of 3-4 instruments with wasted space...it will result in more noises when recorded at 24/96, and most of the time these noises can not be heard unless you pump the music loud enough to blow out your ears. Hence I will commit finding a better recording sound track to test it myself.
Click to expand...

 
24-bit is relevant to recording due to higher demands of computer processing. I think that is what you're thinking of.
 
Jan 28, 2015 at 7:46 PM Post #19 of 42
music alchemist said:
   
Yeah, well, we can't conduct a proper investigation otherwise, can we?
biggrin.gif
Click to expand...


 The investigation is just as biased ( in which particular direction depending on your personal perspective) regardless of where it's posted. It just gets observed and participated in by the correct parties in sound science. It's common courtesy as outlined in the rules so portable source gear threads stay about portable source gear. Doubting what another hears is also frowned upon other than in that forum.
 "If what you want to post includes words/phrases like "placebo," "expectation bias," "ABX," "blind testing," etc., please post it in the Sound Science forum." 
 "Discussion of blind testing is only allowed in the Sound Science forum." http://www.head-fi.org/a/terms-of-service
 Just a heads up. I'm not reporting or further participating. I wouldn't have posted this if you didn't respond to my polite reminder.
 
Jan 28, 2015 at 8:04 PM Post #20 of 42
goodvibes said:
   The investigation is just as biased ( in which particular direction depending on your personal perspective) regardless of where it's posted. It just gets observed and participated in by the correct parties in sound science. It's common courtesy as outlined in the rules so portable source gear threads stay about portable source gear. Doubting what another hears is also frowned upon other than in that forum.
 "If what you want to post includes words/phrases like "placebo," "expectation bias," "ABX," "blind testing," etc., please post it in the Sound Science forum." 
 "Discussion of blind testing is only allowed in the Sound Science forum." http://www.head-fi.org/a/terms-of-service
 Just a heads up. I'm not reporting or further participating. I wouldn't have posted this if you didn't respond to my polite reminder.
Click to expand...

 
I never used any of those words or phrases.
tongue.gif

 
And no, the investigation is objective. If you want to learn more about it, visit the relevant section.
 
In case you didn't realize, the thread I linked to was in the Sound Science section. I thought that's what you were referring to. I was agreeing with you that the investigation should take place in that section.
 
Jan 28, 2015 at 8:32 PM Post #21 of 42
OK, just so you're aware, I know enough. Unless you have the best kit in the best environment, it's always relative when actual listening is involved. Problem is you could have a decade long discussion on what constitutes best kit or environment in the same old sound vs specs format. It's an endless circle driven by personal opinion and experience. Scientific process remains a constant but experiments can be setup to more easily fail or succeed depending on parameters. It's a lot less linear when it comes to sound. For instance, if we're lucky, we can hear to 20Khz but we can differentiate as little as a cent or 2 in time of a midrange frequency when it's a differential or harmonic. Can 16/44 reproduce that sort of time differential? Maybe there's more to it, maybe not but being the only objective guy in the room is probably not the answer.
 
Jan 28, 2015 at 8:38 PM Post #22 of 42
goodvibes said:
  OK, just so you're aware, I know enough. Unless you have the best kit in the best environment, it's always relative when actual listening is involved. Problem is you could have a decade long discussion on what constitutes best kit or environment in the same old sound vs specs format. It's an endless circle driven by personal opinion and experience. Scientific process remains a constant but experiments can be setup to more easily fail or succeed depending on parameters. It's a lot less linear when it comes to sound. For instance, if we're lucky, we can hear to 20Khz but we can differentiate as little as a cent or 2 in time of a midrange frequency when it's a differential or harmonic. Can 16/44 reproduce that sort of time differential? Maybe there's more to it, maybe not but being the only objective guy in the room is probably not the answer.
Click to expand...

 
I honestly think it would be cool if high-res sounded better than Red Book when the files are from the same master. It's just that no one has ever been able to demonstrate such a thing.
 
Jan 28, 2015 at 9:39 PM Post #23 of 42
goodvibes said:
OK, just so you're aware, I know enough. Unless you have the best kit in the best environment, it's always relative when actual listening is involved. Problem is you could have a decade long discussion on what constitutes best kit or environment in the same old sound vs specs format. It's an endless circle driven by personal opinion and experience. Scientific process remains a constant but experiments can be setup to more easily fail or succeed depending on parameters. It's a lot less linear when it comes to sound. For instance, if we're lucky, we can hear to 20Khz but we can differentiate as little as a cent or 2 in time of a midrange frequency when it's a differential or harmonic. Can 16/44 reproduce that sort of time differential? Maybe there's more to it, maybe not but being the only objective guy in the room is probably not the answer.
Click to expand...


This is it, I agree. There is not a night and day difference, it is the fuller feeling you may be able to tell if you listen to both back and forth continuously. Blind test as a random test can't tell the diffence. That doesn't mean it is not there, it is just that your brain is confused on what it is perceiving. Try it with different perfumes or fragrances. I can let you blind test two totally different perfumes and smell very much a like, and you won't be able to tell the difference after 2-3 times. Unless you will do this a couple times a day for a few days, and while not being blinded, your brain will slowly differentiates it. But again, after you recognized it, and let your brain forget about it for a few months, I can blind test you again, and you won't be able to tell. It is that subtle, but it is there.
 
Jan 29, 2015 at 7:06 AM Post #26 of 42
Using 320 with my A17 on hi-rez setting gives me all the sq i will need on Jazz and Rock, but then Sony are the masters at standard downloads.  DSD is better, but not better enough. I would use it for poorly recorded classical greats however. 
 
Jan 29, 2015 at 8:09 AM Post #27 of 42
milford30 said:
   
 
I always thought they need to increase the frequency so that they can increase the sampling rate due to the 
Nyquist–Shannon sampling theorem.​
Click to expand...

 
Yes if you want to capture a frequency greater than 22.05kHz you need a sampling rate greater than 44.1ksamp/sec. But there is little evidence that humans in general can hear (at least with their ears) higher than 20kHz. The slightly higher sampling rate for DVD/Blu-ray (48ksamp/sec) I believe is tied more into syncing audio with video more than it is with a need to generate a 24kHz tone. I mean it's easy to experiment. Make a 48k file of a 23kHz sine wave and see how well you can hear it / sense it through headphones and speakers.
 
Jan 29, 2015 at 8:32 AM Post #28 of 42
rrod said:
   
Yes if you want to capture a frequency greater than 22.05kHz you need a sampling rate greater than 44.1ksamp/sec. But there is little evidence that humans in general can hear (at least with their ears) higher than 20kHz. The slightly higher sampling rate for DVD/Blu-ray (48ksamp/sec) I believe is tied more into syncing audio with video more than it is with a need to generate a 24kHz tone. I mean it's easy to experiment. Make a 48k file of a 23kHz sine wave and see how well you can hear it / sense it through headphones and speakers.
Click to expand...

My point is we want a higher sampling rate therefore we need to increase the frequency range, otherwise we can't preform fourier analysis on it to turn it into 0 and 1s. 24kHz is just due to the sampling rate we want and still want it to be digitized, so the point is if humans can tell the difference between the sampling rates and NOT hearing extra sounds at 20+ kHz that has been suggested.
 
Jan 29, 2015 at 8:46 AM Post #29 of 42
milford30 said:
  My point is we want a higher sampling rate therefore we need to increase the frequency range, otherwise we can't preform fourier analysis on it to turn it into 0 and 1s. 24kHz is just due to the sampling rate we want and still want it to be digitized, so the point is if humans can tell the difference between the sampling rates and NOT hearing extra sounds at 20+ kHz that has been suggested.
Click to expand...

 
Fourier analysis doesn't turn the waveform into 0s and 1s, it takes a waveform from the time domain to the frequency domain and vice-versa. In the case of digital audio, the waveform will be discrete (having already been turned into discrete data by the ADC) and assumed to be periodic, and thus the frequency representation will be periodic and discrete.
 
You are reversing the logic a bit: we don't chose a frequency range to get a sampling rate; we chose a sampling rate to get a frequency range. All differences between PCM formats at a given bit depth are due to the sampling rate; there's nothing else that is different! So for instance the faster rise-times for transients that 16/96 PCM has over 16/48 PCM are entirely due to the extra frequency content (24-48kHz) allowed by the higher sampling rate. And you can't hear those frequencies, so ask yourself how you are possibly sensing something like these "better" transients.
 
Jan 29, 2015 at 9:28 AM Post #30 of 42
rrod said:
   
Fourier analysis doesn't turn the waveform into 0s and 1s, it takes a waveform from the time domain to the frequency domain and vice-versa. In the case of digital audio, the waveform will be discrete (having already been turned into discrete data by the ADC) and assumed to be periodic, and thus the frequency representation will be periodic and discrete.
 
You are reversing the logic a bit: we don't chose a frequency range to get a sampling rate; we chose a sampling rate to get a frequency range. All differences between PCM formats at a given bit depth are due to the sampling rate; there's nothing else that is different! So for instance the faster rise-times for transients that 16/96 PCM has over 16/48 PCM is entirely due to the extra frequency content (24-48kHz) allowed by the higher sampling rate. And you can't hear those frequencies, so ask yourself how you are possibly sensing something like these "better" transients.
Click to expand...

 
Fourier transform is how digital signal processing works:
 
http://en.wikipedia.org/wiki/Digital_signal_processing
 
Fourier analysis is at the heart of most analogue (waves you get from your mic or out of your headphones) and digital (0 and 1s) convertions. 
 
Your mic records the signal in the time domain, then you take samples of it to convert it digitally, the sampling rate is chosen and limited by 
Nyquist–Shannon sampling theorem​
(which is closely linked to fourier transforms)​
  
"My point is we want a higher sampling rate therefore we need to increase the frequency range" <- this means we choose sampling rate which results in a higher frequency spectrum.
 
"NOT hearing extra sounds at 20+ kHz that has been suggested" <- clearly said what you have just said "And you can't hear those frequencies, so ask yourself how you are possibly sensing something like these "better" transients."
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top