[nick_charles]

Quote:

Too bad the thread got moved to "sound science". 
 

Yes, heaven forfend that audiophile orthodoxy about expensive tweaks or culty low tech or expensive add-ons be challenged in one of the popular sub-forums

 

[chesebert]

no ABX.
 
You can't exactly hear 16khz sine wave, not that young; but I am classically trained so my ears are ok I guess. Instrument with higher registers sounds absolutely more organic on 24/96. There is that distinctive lack of hardness, edginess, digitalness, in its place is a more harmonically rich instrument more like the real thing.  I guess you would have to be familiar with what the real thing sounds like to make that kind of judgment.
 
Quote:

Did you test this blind?
 
Also, what's your age and listening volume, to be able to hear harmonics over 16kHz with ease?

 

[Head Injury]

Quote:

no ABX.
 
You can't exactly hear 16khz sine wave, not that young; but I am classically trained so my ears are ok I guess. Instrument with higher registers sounds absolutely more organic on 24/96. There is that distinctive lack of hardness, edginess, digitalness, in its place is a more harmonically rich instrument more like real the real thing.  I guess you would have to familiar with what the real thing sounded like to make that kind of judgment.

Then you should be able to blind test it no problem. With two files it's simple, too. Put them both in a playlist and set it on Random ordering and repeat. Close your eyes and pick which you think is which. If you can use Foobar, there's also the ABX plugin which is a great tool for this.
 
You're in the Science domain now, gotta expect this sort of thing 

 

[leeperry]

download some 24/96 Linn tracks, downsample yourself and listen [..]
Symphony - very audible difference
 
Any instrument with 2nd, 3rd and 4th harmonics over 16khz sounds more real, and more organic on 24/96.
 
I am well aware for most music 24/96 makes no difference.

 
OK great, now we have real world experiments to rely on

 
I'll certainly agree that some 24/96 tracks can sound noticeably worse when downsampled/downscaled on your own computer...but it certainly boils down to:
-the algorithm being used to downsample, and any resampling will increase THD/THD+N anyway(easy to measure w/ WaveSpectra)
-the dithering/noise shaping algorithms being used(and there's no "one size fits all")...most consumer-grade audio software will use a very basic downscaling algorithm.
 
A simple experiment is to take a 24bit lossless soundtrack, downconvert it to 23>16bit in eac3to...and notice how each conversion gets audibly worse than the former.
 
The only point to play 24/96 is to keep those 24/96 soundtracks unmolested...whether they'd sound better than 16/44.1 or 48 files encoded from the same 32/64bit floating point master is still open for debate, though.
 
Last time a 24/96 audio seller came here on head-fi to tell us how 24/96 audio is sheer awesomeness, we kindly asked for 16/44.1 and 24/96 samples from the very same floating point master...but he started telling us that CD and HD audio couldn't possibly be mastered identically(you can easily end up w/ 8bit audio or worse on a CDDA if the volume's too low...a real problem for classical music for instance) 
 
24bit in a DAC isn't even real, as no DAC chip can reach 144dB dynamics...the only point is to transport an unmolested stream, but the other drawback of 24bit over S/PDIF is that it requires uber-low jitter: http://www.agoraquest.com/viewtopic.php?topic=34057&forum=51

The difference is not loss of Data, but the corruption of the timing relationship between the Data and the 4 timing clocks in the S/PDIF Bitstream. This timing discrepancy is called jitter and for 16 bit data there can be no more than 100 picoseconds (1 picosecond =1/1,000,000,000,000th of a second) of jitter for full resolution whereas 24 bit Data can have no more than 0.5 picoseconds of jitter to maintain full resolution.

 
so yes, 24/96 files will/could/would/should sound better on a 24/96 DAC....but I'd bet that +90% of your sound library is 16/44.1 anyway, and the improvement due to 24/96(which is way overkill for headphones use) might be destroyed by the S/PDIF jitter at those high rates IMHO.
 
Many studio work in 48bit these days, should we get ready for the 48/384 files? oh my oh my, that'll sound eye popping for sure.

 

[chesebert]

1 user with no control group is not an experiment. real world experience, perhaps.
 
Most high resolution files are wasted on headphone, that's why I have speakers.
 
I use soundforge for downsampling.
 
No ABX cuz my computer is not even close to my audio system.  Everything is streamed over ethernet.
 
Any competently made DAC these days should easily have more than 16bit in actual resolution playing back 24/96 files.
 
If you can't hear the different or don't have the gear to hear the difference, that's your problem.  Please don't impute your experience or "inexperience" on others.
 
the 48bit studio files are there for keeping all the decimals in place when manipulating  files.  I don't see how 48/384 is relevant for home use.
 
This is what happens when threads are moved to "sound science".  You get bunch of inexperienced "science" folks who still thinks 1 is 1 and 0 is 0. And if you can't calculate it, you can't hear it.
 
What a waste of time.
 
Quote:

 
OK great, now we have real world experiments to rely on

 
I'll certainly agree that some 24/96 tracks can sound noticeably worse when downsampled/downscaled on your own computer...but it certainly boils down to:
-the algorithm being used to downsample, and any resampling will increase THD/THD+N anyway(easy to measure w/ WaveSpectra)
-the dithering/noise shaping algorithms being used(and there's no "one size fits all")...most consumer-grade audio software will use a very basic downscaling algorithm.
 
A simple experiment is to take a 24bit lossless soundtrack, downconvert it to 23>16bit in eac3to...and notice how each conversion gets audibly worse than the former.
 
The only point to play 24/96 is to keep those 24/96 soundtracks unmolested...whether they'd sound better than 16/44.1 or 48 files encoded from the same 32/64bit floating point master is still open for debate, though.
 
Last time a 24/96 audio seller came here on head-fi to tell us how 24/96 audio is sheer awesomeness, we kindly asked for 16/44.1 and 24/96 samples from the very same floating point master...but he started telling us that CD and HD audio couldn't possibly be mastered identically(you can easily end up w/ 8bit audio or worse on a CDDA if the volume's too low...a real problem for classical music for instance) 
 
24bit in a DAC isn't even real, as no DAC chip can reach 144dB dynamics...the only point is to transport an unmolested stream, but the other drawback of 24bit over S/PDIF is that it requires uber-low jitter: http://www.agoraquest.com/viewtopic.php?topic=34057&forum=51
 
so yes, 24/96 files will/could/would/should sound better on a 24/96 DAC....but I'd bet that +90% of your sound library is 16/44.1 anyway, and the improvement due to 24/96(which is way overkill for headphones use) might be destroyed by the S/PDIF jitter at those high rates IMHO.

 

 

[leeperry]

Well, downsampling/downscaling/dithering/noise shaping algorithms is where the rage takes place these days...Apogee UV22HR that used to be the the cream of the crop 10 years ago has become unusable by today's standards...
 
Listen to this CD and tells me that it doesn't sound HD:

 
I believe it's using Sony's Super BitMapping algorithm(http://www.hydrogenaudio.org/forums/index.php?showtopic=37566), I've got a few remastered CD's using this thing and they always leave me out of speech. It was recorded in 1976 BTW

 
By using a state-of-the-art noise shaping/dithering combo, you can audibly improve a 16bit downscale...to the point that you'd be hard pressed to hear a difference over the original floating point source.

 

[nick_charles]

Quote:

1 user with no control group is not an experiment. real world experience, perhaps.
 
Most high resolution files are wasted on headphone, that's why I have speakers.
 
I use soundforge for downsampling.
 
No ABX cuz my computer is not even close to my audio system.  Everything is streamed over ethernet.
 
Any competently made DAC these days should easily have more than 16bit in actual resolution playing back 24/96 files.
 
If you can't hear the different or don't have the gear to hear the difference, that's your problem.  Please don't impute your experience or "inexperience" on others.
 
the 48bit studio files are there for keeping all the decimals in place when manipulating  files.  I don't see how 48/384 is relevant for home use.
 
This is what happens when threads are moved to "sound science".  You get bunch of inexperienced "science" folks who still thinks 1 is 1 and 0 is 0. And if you can't calculate it, you can't hear it.
 
What a waste of time.
 

 

The biggest benefit of the science subform is that it is a place where we don't just "take your word for it" . Here when people make claims about hearing differences that may be dubious they are expected to provide some actual evidence that stands up to scrutiny. Merely falling back on personal anecdotes, classical training, experience and golden ears is not substantive evidence, I speak as one with 37+ years of interest in high fidelity, but my experience means nothing whatsoever, and neither does yours.
 
What is meaningful here is what can be shown to be verifiable, that is why blind tests are considered as better evidence. If you can show a verifiable ability to detect the differences between high res and 16/44.1 as a consumer format then you are doing better than peer reviewed journal research using 60 trained listeners and over 500 trials and then folks here would be very interested, till then it is just the same old "I do not need to prove it to you" mantra.

 

[chesebert]

Hearing is inherently subjective.  Your hearing is a culmination of your experience, your association of sounds based on that experience and the reevaluation of those associated experience with those of other humans you come into contact with.  To try to quantify, measure, or objective determine a subjective variable is just futile.
 
I don't know why Jude even opened this sub-forum.  this is purely time wasting.  Go get drunk or stoned and I promise everything will sound better than you have ever heard before.
 
 
Quote:

The biggest benefit of the science subform is that it is a place where we don't just "take your word for it" . Here when people make claims about hearing differences that may be dubious they are expected to provide some actual evidence that stands up to scrutiny. Merely falling back on personal anecdotes, classical training, experience and golden ears is not substantive evidence, I speak as one with 37+ years of interest in high fidelity, but my experience means nothing whatsoever, and neither does yours.
 
What is meaningful here is what can be shown to be verifiable, that is why blind tests are considered as better evidence. If you can show a verifiable ability to detect the differences between high res and 16/44.1 as a consumer format then you are doing better than peer reviewed journal research using 60 trained listeners and over 500 trials and then folks here would be very interested, till then it is just the same old "I do not need to prove it to you" mantra.

 

[Head Injury]

Quote:

Hearing is inherently subjective.  Your hearing is a culmination of your experience, your association of sounds based on that experience and the reevaluation of those associated experience with those of other humans you come into contact with.  To try to quantify, measure, or objective determine a subjective variable is just futile.
 
I don't know why Jude even opened this sub-forum.  this is purely time wasting.  Go get drunk or stoned and I promise everything will sound better than you have ever heard before.

That's not what we're trying to do here! We just want you to prove that your subjective experiences are real. ABX and DBT tests have no interest in measuring them, or describing what they are, or what causes them. These tests are only interested in seeing if they still exist when bias is removed.

 

[nick_charles]

Quote:

Hearing is inherently subjective.  Your hearing is a culmination of your experience, your association of sounds based on that experience and the reevaluation of those associated experience with those of other humans you come into contact with.  To try to quantify, measure, or objective determine a subjective variable is just futile.
 
I don't know why Jude even opened this sub-forum.  this is purely time wasting.  Go get drunk or stoned and I promise everything will sound better than you have ever heard before.
 
 

When somebody says "I hear a difference between X and Y" and presents that assertion as evidence that there is a difference between X and Y it is rational to ask "'well can the subject really hear a difference betweehn X and Y". If you think that this utterly minimal level of critical inquiry is pointless there are plenty of other subforums where you are allowed to voice unverified opinions as gospel.

 

[haloxt]

Questions! Questions! To sum it up
 
1. Is there an audible difference between $1000 and $5000 dac? And if there's a difference, how much?
 
2. Is there an audible difference between 24/96 and 16/44.1? And if there's a difference, how much?
 
3. Assuming either 1 or 2 is true or both are true, what is the "right" way for people to buy dac's and/or cd's?
 
The UN is working on a mathematical model to describe all of the above along with what you should wear and eat, with an ETA of 5 years. Keep tuned in.

 

[Ham Sandwich]

 
Quote:

  Is it my imagination or was this thread moved out of the Source subforum ? 

The move happened during my reply.  I typed in my reply while in the Dedicated Sources forum.  When I hit submit I got teleported here to Sound Science.
 
I never get to post in the Dedicated Sources forum.  Maybe if I just typed faster I could have succeeded for just once even if briefly. 

 

[Pio2001]

Quote:

1. Is there an audible difference between $1000 and $5000 dac? And if there's a difference, how much?
 
2. Is there an audible difference between 24/96 and 16/44.1? And if there's a difference, how much?
 
3. Assuming either 1 or 2 is true or both are true, what is the "right" way for people to buy dac's and/or cd's?

There are two categories of audible differences.
 
Differences in the delivered sound, that cause differences in what we hear. They are measurable, and quantifiable.
Differences in what we hear that are not caused by differences in the sound. They might be called "psychologic", though their belonging to the field of psychology, physiology or neurology is another debate.
 
Both differences look exactly the same to our ears. Only some advanced practice of double blind testing allows, in some case, to distinguish between the two.
 
So about question one,
-There are often differences of the first kind. High end audio is a domain where many, if not most, electronics are "coloured". Their measured performances are usually very bad. Amplifiers with THD between 1 and 10 % (more than 10 % for the most expensive amplifier in the world). I cited some sources of such differences above : mostly non-oversampling, and tampered digital filters.
How much ? Just plot the frequency response of said DACs.
 
-Differences of the first kind are nearly non-existent between 24/96 and 16/44.1. How much is "nearly" ? If you crank up the volume to insane level in the absolute silence, you can hear the 16 bits quantization noise.
 
-Differences of the second kind are extremely versatile. They vary from listener to listener. Some listeners hear no differences between nearly anything, while other hear differences between everything. They can vary according to the time of the day, the weather, anything... One of their most important characteristic is that they are not quantifiable in the same way as differences of the first kind. The most common measurment unit is the "night and day".
For convenience, we might call them differences in "musicality".
 
About question one and two, I'd say that differences of the second kind do not depend on what is tested. The fact that DACs between 1k and 5k are considered, or 24/96 vs 16/44.1 has absolutely no importance for these differences. It's an alchemy taking place between a listener and a concept. The concept of $ 5k DAC does not mean the same thing for all listeners, therefore they do not experience the same amount of 2nd kind differences between them.
 
-What is "right" ? Depends on what "right" means to the listener. Some, called "objectivist", think that 1st kind differences are right, while 2nd kind differences are wrong, and thus blind listening tests are absolutely necessary in order to know what is right and what is wrong.
Subjectivists think that both kind are right, thus blind listening test are completely unuseful.
 
Most of the audiophile don't fall in these extreme category, and while all of them agree that differences of the first kind are right, differences of the second kind are not really dealt with.

 

[haloxt]

 
Quote:

There are two categories of audible differences.  
Differences in the delivered sound, that cause differences in what we hear. They are measurable, and quantifiable.
Differences in what we hear that are not caused by differences in the sound. They might be called "psychologic", though their belonging to the field of psychology, physiology or neurology is another debate.

 
Quote:

Subjectivists think that both kind are right, thus blind listening test are completely unuseful.

 
Oh I see. Pardon me but isn't that like a murderer who tries to convince people he's not a murderer by saying he's a murderer? Or maybe you can show me some examples of "subjectivists" who say placebo is an important part of their musical enjoyment.

 

[leeperry]

hehe, Pio2k1 can be such an entertainer sometimes