Sample Rate Question.

[Psalad]

Originally Posted by Doug Young

This stuff gets difficult really fast.... There was a serious attempt just to compare different converters a few years back on gearslutz. Had some very qualified people involved - AES speakers, recording engineers, etc. Turned into a giant flame war that ended with the entire thread not only being locked, but deleted. I imagine some others here saw it at the time before it disappeared. Somehow, these discussions/tests are not only never conclusive, but they never end well. Seems easy enough, but clearly it's not.

I personally think the idea is to respect dissent and disagreement. I suspect neither of us are scientists. .

As far as testing goes it can be an ongoing thing IMO. I have done a bit but I can always learn more. The other idea is to keep an open mind.

[KevWind]

Originally Posted by Psalad

Why would you need two of those? Two of any of the same converters would work, no? Wouldn't you think it would be somewhat easy to put out a note to friends to borrow two that match?

Yes any two would work. But I have one OMNI in my system already which was why I sated two of them ? Yes I could possibly borrow the equip. if I were really concerned about it.

Even if they didn't match... might be worth a try to see if people can even hear the difference between two good quality modern converters. I doubt if I could tell.

Probably true, but as one who keeps bringing up " the science " I would think you would understand the goal of "science" is to eliminate variables that could even possibly skew results , not create or overlook them.

Again, if you think it needs a test to that degree, then you should do it. That would be great!

Yes I think in order be touted as objective science the minimum would be the same converters.

In my case, the one computer that was doing the playback and running a/b/x software was running at 88.2 (so the software was up converting for the test). I know that's not good enough for you, and that's fair enough, you are welcome to create your own test.. but it's good enough to me for a basic test.

Ok so to be clear the the 44.1 file was up converted to 88.2 before play back ? So in reality you were comparing a file recorded in 44.1 and up converted to 88.2 for play back, against a file recorded in 88.2 and played back in 88.2 (kind of the reverse of the M & M test)
And if that is good enough for you then and that's good enough for you . Although isn't it interesting that those who say they can hear a difference say their testing methods are "good enough" for them.


And as for the interfaces:

Putting aside the question of exactly how accurate 8 channels of conversion in single box multi purpose $800 and $600 interface units are actually going to be and what that might possibly mean or not ?

It should be noted for clarification that you recorded the 88.2 signal through the older $600 Steinburg unit and the 44.1 through the newer $800 dollar MOTU unit, . And while it does not seem unreasonable to assume that at the price point, and the modest difference in the price of the units, they are not going to be that different quality wise , But I think we all agree, that every year the units get better and better for the money, especially at the lower mid to upper mid price point.
But , once again the introduction of more variables that can possibly skew results serves to thwart the scientific goal of eliminating them.

"Audio d/a" is "distribution amplifier." Active splitter. line level.

ah ! makes perfect sense.

Two computers for recording, and one for playback as it means you can simply run software a/b/x testers. But the more complex two computers/two interfaces test would be even better if you have the ability to level match, have someone switching and tabulating in a way that ensures there are to tells, etc.

Agreed.

In summary and back to the beginning. I have yet to be convinced that the "testing" being done so far , really does meet the minimum standard of eliminating enough variables, to actually establish a baseline of "science". That would then arguably cast the burden of proof to those proffering an alternate view.

Thus I do not feel the need or weight of burden of proof, to justify my personal opinions or choice of sample rate or the need or to conduct a test to justify it.

[KevWind]

Originally Posted by Psalad

I personally think the idea is to respect dissent and disagreement. I suspect neither of us are scientists. .

As far as testing goes it can be an ongoing thing IMO. I have done a bit but I can always learn more. The other idea is to keep an open mind.

So far other than 1 quip I made misunderstanding your intent, I think our exchange has been pretty respectful if direct.
I am certainly not a scientist, That said with false modesty aside, I think my strong suit is analyzing design and implementation. I spent the last 15 years of construction career analyzing possible flaws and problems between the 2 dimensional architectural and engineering plans and the reality of implementing those design ideals and numbers into a 3 dimensional construct.

[Psalad]

Originally Posted by KevWind

Probably true, but as one who keeps bringing up " the science " I would think you would understand the goal of "science" is to eliminate variables that could even possibly skew results , not create them.

Sheesh dude, why the snark?? Not at all necessary.

Point is, if one cannot hear the difference between two separate converters running at two different sample rates, then chances are one cannot hear the differences between two sample rates. The differences are greater in my example. Of course, as you might suspect, I already knew it was impossible for myself to hear the difference in these converters, because.. yea, I tested them.

I just bought a new interface, a Focusrite Clarett. I will be having some fun doing a test of that soon as well.

Yes I think in order be touted as objective science the minimum would be the same converters.

...I never claimed my simple listening exercise to be ANYTHING near "objective science."

And if that is good enough for you then and that's good enough for you . Although isn't it interesting that those who say they can hear a difference say their testing methods are "good enough" for them.

Look, it's "good enough" until I find a better test. See, I've done a bunch of tests. But, I have learned a great deal about listening.

Putting aside the question of exactly how accurate 8 channels of conversion in single box multi purpose $800 and $600 interface units are actually going to be and what that might possibly mean or not ?

A great question, and one that can easily be understood with simple listening tests. I don't find it that much of an inconvenience to test this stuff.

Sure, it gives you something else to complain about in the methodology.. and sure, maybe it's fair enough in a way, but my own testing shows me there really is no audible difference in decent modern converters, at least the ones I've tested.

So I definitely agree my testing is flawed and imperfect, but it's only through direct blind listening tests can we really learn about real difference vs. perceived differences. Bias is strong and unavoidable.

In summary and back to the beginning. I have yet to be convinced that the "testing" being done so far , really does meet the minimum standard of eliminating enough variables, to actually establish a baseline of "science". That would then arguably cast the burden of proof to those proffering an alternate view.

Thus I do not feel the need or weight of burden of proof, to justify my personal opinions or choice of sample rate or the need or to conduct a test to justify it.

Anyway, I guess you and I differ in our approaches. Carry on, and if you ever decide to do a test, please invite me, I love to learn.

[KevWind]

Originally Posted by Psalad

Sheesh dude, why the snark?? Not at all necessary.

Point is, if one cannot hear the difference between two separate converters running at two different sample rates, then chances are one cannot hear the differences between two sample rates. The differences are greater in my example. Of course, as you might suspect, I already knew it was impossible for myself to hear the difference in these converters, because.. yea, I tested them.

I just bought a new interface, a Focusrite Clarett. I will be having some fun doing a test of that soon as well.



...I never claimed my simple listening exercise to be ANYTHING near "objective science."



Look, it's "good enough" until I find a better test. See, I've done a bunch of tests. But, I have learned a great deal about listening.



A great question, and one that can easily be understood with simple listening tests. I don't find it that much of an inconvenience to test this stuff.

Sure, it gives you something else to complain about in the methodology.. and sure, maybe it's fair enough in a way, but my own testing shows me there really is no audible difference in decent modern converters, at least the ones I've tested.

So I definitely agree my testing is flawed and imperfect, but it's only through direct blind listening tests can we really learn about real difference vs. perceived differences. Bias is strong and unavoidable.



Anyway, I guess you and I differ in our approaches. Carry on, and if you ever decide to do a test, please invite me, I love to learn.

I was not intending to be snarky if you took it so, I do apologize.
I was simply trying to point out that you are the one that early an often kept stating you were going by "the science"
You and I do not differ in our approaches per se, because we do not disagree about the scientific value of objective blind testing.
We do seem to disagree about the value to science, of the testing that has been done so far, and not just yours, but all of the testing as far as I can tell. The test you described doing is to my mind, every bit as valid as the M&M test touted by members of AES. But both, to my mind have at least one fundamental flaw in design reasoning . If that's complaining I think it a valid complaint.

[Psalad]

Originally Posted by KevWind

But both, to my mind have at least one fundamental flaw in design reasoning . If that's complaining I think it a valid complaint.

Right, I said it was kind of a valid complaint.

I think everyone benefits from doing their own tests, especially if they believe in things that are unproven. I think that is apparently one of the main differences in our philosophies.

[Psalad]

Originally Posted by KevWind

I think my strong suit is analyzing design and implementation.

That seems to be true.

You seem to be continually missing where the burden of proof lies in this discussion though.

You have the known average top limit of hearing at 20k, you have a digital audio system designed to capture all the audio that humans can hear at 44.1. Those decisions were made with current understanding of audio. It's up to those who believe there are problems with the existing specs to prove those potential problems are audible.

It's not up to the pro audio community to prove the opposite, any more than its up to the pro audio community to prove that $5k IEC power cables don't make a difference. Any and all claims of the limitations of the existing system need to be proven by those making the claim.

[KevWind]

Originally Posted by Psalad

Right, I said it was kind of a valid complaint.

I think everyone benefits from doing their own tests, especially if they believe in things that are unproven. I think that is apparently one of the main differences in our philosophies.

Again really no difference in that portion of our philosophy either, Could not agree more that people should do there own testing "especially if they believe in things that are "unproven" ....the only difference is the notion of what has actually been proven and what hasn't actually been proven and if it is valid and applies .

Originally Posted by Psalad

You seem to be continually missing where the burden of proof lies in this discussion though.

No not at all missing that you think there some kind of "imperative" in this discussion of sampling music to establish a burden of proof.
You seem to be continually missing .. that is simply your opinion, and not an inherent fact of this discussion

You have the known average top limit of hearing at 20k, you have a digital audio system designed to capture all the audio that humans can hear at 44.1. Those decisions were made with current understanding of audio.

Yes generally human detection of single sine wave frequency tops at 20k.
Yes the 44.1k was the default decision made in 1980 based on a theorem published in 1928, and human hearing tests of single sign wave frequencies. Trust me I got it.

It's up to those who believe there are problems with the existing specs to prove those potential problems are audible.

Thats just the point, only ones who feel that there is some kind of imperative to so, believe that.

It's not up to the pro audio community to prove the opposite, any more than its up to the pro audio community to prove that $5k IEC owner cables don't make a difference. Any and all claims of the limitations of the existing system need to be proven by those making the claim.

That's just exactly the point again. Nobody is asking "the pro audio community to prove the opposite"
That seems to be the parlance of those defending their belief in a 35 year old standard based on single sign wave hearing tests . If that's the anchor of criteria for you, that's great, nobody is asking you change, or think, or prove, any differently. Those shouts of burden are only flowing one direction.

And this is getting past circular. see ya

[Psalad]

Originally Posted by KevWind

That's just exactly the point again. Nobody is asking "the pro audio community to prove the opposite"

Actually you did earlier in the thread I believe.

That seems to be the parlance of those defending their belief in a 35 year old standard based on single sign wave hearing tests

That's exactly it. That old standard stands until proven otherwise. Not belief, but logic in creating an adequate system. If the system is inadequate or problematic it can easily be shown by a well run test. 35 years and still no, right?

[KevWind]

Originally Posted by Psalad

Actually you did earlier in the thread I believe.

Wrong .

That's exactly it. That old standard stands until proven otherwise. Not belief, but logic in creating an adequate system. If the system is inadequate or problematic it can easily be shown by a well run test. 35 years and still no, right?

Circular

[Psalad]

Originally Posted by KevWind

So while it is probsbly accurate to say (as Psalad) pointed out
"There is no scientific evidence anyone can reliably pick out a higher sample rate and bit depth audio file from 44.1/16 bit."

It is also probably accurate to say:
There is no scientific evidence that recording at higher sample rates does not make any kind of audible difference.

This is what I was referencing.

[KevWind]

Originally Posted by Psalad

This is what I was referencing.

"So while it is probably accurate to say (as Psalad) pointed out
"There is no scientific evidence anyone can reliably pick out a higher sample rate and bit depth audio file from 44.1/16 bit."
It is also probably accurate to say:
There is no scientific evidence that recording at higher sample rates does not make any kind of audible difference."

Is an observation of lack of unflawed evidence that has been already been presented either way. Not a demand for proof
Still wrong.....

[Psalad]

Originally Posted by KevWind

Is an observation of lack of unflawed evidence that has been already been presented either way. Not a demand for proof
Still wrong.....

LOL.. OK Kev.

Let me put it this way. There is a ton of magical thinking in the audio world, and there has been as long as I can remember. Bridge pins, analog summing, IEC power cables, speaker cables, 16 vs 20 vs 24 bit, "tonewoods" on solid body guitars, converters, sample rates above 44.1, etc.

In all these cases I've heard the differences called out as "night and day." "Night and day" signifies a HUGE difference. In NONE of these cases is there a huge difference, most likely there is NO audible difference.

Unlike a lot of other things, in audio, it's impossible to directly and simultaneously compare. The "memory" of the ear is very short, plus sound changes significantly through the air to the point that moving your head just an inch or two makes a big difference in how sound appears in a room. Combine this with the existing problem of massive expectation bias (and our lack of awareness of it) and this is why a lot of these things get traction. This is where testing comes in, and the only test we have is a/b/x. It's not perfect, but it's all we have.

Makes me think about video... makes me wonder why nobody is clamoring for all those colors that humans can't see to be shown on my big screen. There is a reason we use infrared for our remote controls. Food for thought.

Anyway, the bottom line is feel free to use whatever sample rate you choose. There are real audio things to worry about that make huge audible differences, and sample rate is not one of them.

There will always be people who want to make audio more complex than it really is, by making it seem like magic. While there is no substitute for experience, it's not as magical as some in the audio world would make it appear. The most important thing to getting good at recording is experience, and lots of people are out there these days willing to share theirs for free. Make music and don't sweat the small stuff.

[rick-slo]

This is a good article for you guys to read through:
http://www.trustmeimascientist.com/2...-when-it-isnt/

[Psalad]

Originally Posted by rick-slo

This is a good article for you guys to read through:
http://www.trustmeimascientist.com/2...-when-it-isnt/

Good piece, thanks for posting, read that a while back.

This one too:

http://people.xiph.org/~xiphmont/demo/neil-young.html