So - a newbie asks about 96kHz

[Tom Page]

I am so sorry - please do point me to a nesgroup thread or web site if
this has been successfully answered before.

I know that the well-heald axiom is that you only need have a sampling
rate that is twice the highest sound you can hear - so 44.1kHz was
chose for CDs - most people have trouble hearing 20kHz, let along
22kHz.

Now of course we see lots about 96kHz sound - does this mean that the
above axiom is wrong, or is it more complicated than this - I wondered
if resonances are the answer, but currently cannot understand the way
it works

Thanks, Tom

[Eberhard Sengpiel]

Tom Page > wrote:
> I am so sorry - please do point me to a nesgroup thread or web
> site if this has been successfully answered before.
> I know that the well-heald axiom is that you only need have
> a sampling rate that is twice the highest sound you can hear -
> so 44.1kHz was chosen for CDs - most people have trouble
> hearing 20kHz, let along 22kHz.
> Now of course we see lots about 96kHz sound - does this
> mean that the above axiom is wrong, or is it more complicated
> than this - I wondered if resonances are the answer, but
> currently cannot understand the way it works.


Hi Tom,

so the number battle goes on. The higher, the better.
Look here, because you have to start with microphones:
http://www.schoeps.de/E/cmc6xt.html
The end of the chain are loudspeakers.
The industry is needing this -
and the bats.

Cheers

Eberhard Sengpiel
German forum of microphone recordings
and soundstudio techniques
http://www.sengpielaudio.com

[Lionel]

Tom Page wrote:
> I am so sorry - please do point me to a nesgroup thread or web site if
> this has been successfully answered before.
>
> I know that the well-heald axiom is that you only need have a sampling
> rate that is twice the highest sound you can hear - so 44.1kHz was
> chose for CDs - most people have trouble hearing 20kHz, let along
> 22kHz.
>
> Now of course we see lots about 96kHz sound - does this mean that the
> above axiom is wrong, or is it more complicated than this - I wondered
> if resonances are the answer, but currently cannot understand the way
> it works
>
> Thanks, Tom

96kHz isn't a sound frequency but a *sample* frequency !

[Arny Krueger]

"Tom Page" > wrote in message
om
> I am so sorry - please do point me to a nesgroup thread or web site if
> this has been successfully answered before.
>
> I know that the well-heald axiom is that you only need have a sampling
> rate that is twice the highest sound you can hear - so 44.1kHz was
> chose for CDs - most people have trouble hearing 20kHz, let along
> 22kHz.

It turns out that the situation is even more extreme than you say. The key
question is not what is the highest frequency you can hear, but what is the
lowest frequency you notice the removal of. The answer to that is more like
15-16 KHz and it's true for people who can hear 20 KHz tones quite clearly.

> Now of course we see lots about 96kHz sound - does this mean that the
> above axiom is wrong, or is it more complicated than this - I wondered
> if resonances are the answer, but currently cannot understand the way
> it works

96 KHz is a great example of an tried and proven audio marketing
principle - numbers for the sake of numbers.

Listen for yourself at

http://www.pcabx.com/technical/low_pass/index.htm

and

http://www.pcabx.com/technical/sample_rates/index.htm

[Todd H.]

(Tom Page) writes:

> I am so sorry - please do point me to a nesgroup thread or web site if
> this has been successfully answered before.
>
> I know that the well-heald axiom is that you only need have a sampling
> rate that is twice the highest sound you can hear - so 44.1kHz was
> chose for CDs - most people have trouble hearing 20kHz, let along
> 22kHz.

Caveat: the nyquist rate (of 2x the max frequency) assumes a perfect
low pass anti-aliasing filter with infinite dB/octave slope... which
of course doesn't exist, and those that comes close introduce
non-linear phase issues in the passband.

The traditional benefit of oversampling is that the antialiasing
low-pass filter doesn't have to be a steep...and doesn't have to
introduce those phase artifacts.

> Now of course we see lots about 96kHz sound - does this mean that
> the above axiom is wrong, or is it more complicated than this - I
> wondered if resonances are the answer, but currently cannot
> understand the way it works

I think we've surpassed the sonic benefits of first order that's for
sure. 96kHz technology is sure as hell not a must-have. Spend it on
better tranducers, techniques, and musicians. I agree with Arny is
that this is innovation the industry is making to sustain itself, and
not borne out of a real and dying need for rather small incremental
fidelity impact.

--
/"\ ASCII Ribbon Campaign | Todd H
\ / | http://www.toddh.net/
X Promoting good netiquette | http://triplethreatband.com/
/ \ http://www.toddh.net/netiquette/ | "4 lines suffice."

[Lionel]

Todd H. wrote:
> (Tom Page) writes:
>
>
>>I am so sorry - please do point me to a nesgroup thread or web site if
>>this has been successfully answered before.
>>
>>I know that the well-heald axiom is that you only need have a sampling
>>rate that is twice the highest sound you can hear - so 44.1kHz was
>>chose for CDs - most people have trouble hearing 20kHz, let along
>>22kHz.
>
>
> Caveat: the nyquist rate (of 2x the max frequency) assumes a perfect
> low pass anti-aliasing filter with infinite dB/octave slope... which
> of course doesn't exist, and those that comes close introduce
> non-linear phase issues in the passband.
>
> The traditional benefit of oversampling is that the antialiasing
> low-pass filter doesn't have to be a steep...and doesn't have to
> introduce those phase artifacts.
>
>
>>Now of course we see lots about 96kHz sound - does this mean that
>>the above axiom is wrong, or is it more complicated than this - I
>>wondered if resonances are the answer, but currently cannot
>>understand the way it works
>
>
> I think we've surpassed the sonic benefits of first order that's for
> sure. 96kHz technology is sure as hell not a must-have. Spend it on
> better tranducers, techniques, and musicians. I agree with Arny is
> that this is innovation the industry is making to sustain itself, and
> not borne out of a real and dying need for rather small incremental
> fidelity impact.
>

You may want to consider that from the archivist point of view...
....for home user the incremental cost isn't very important.

Lionel

[Arny Krueger]

"Todd H." > wrote in message

> (Tom Page) writes:
>
>> I am so sorry - please do point me to a nesgroup thread or web site
>> if this has been successfully answered before.
>>
>> I know that the well-heald axiom is that you only need have a
>> sampling rate that is twice the highest sound you can hear - so
>> 44.1kHz was chose for CDs - most people have trouble hearing 20kHz,
>> let along 22kHz.
>
> Caveat: the nyquist rate (of 2x the max frequency) assumes a perfect
> low pass anti-aliasing filter with infinite dB/octave slope... which
> of course doesn't exist, and those that comes close introduce
> non-linear phase issues in the passband.
>
> The traditional benefit of oversampling is that the antialiasing
> low-pass filter doesn't have to be a steep...and doesn't have to
> introduce those phase artifacts.


For a good example of reconstruction and anti-aliasing filters operating at
44.1 KHz with minimal phase artifacts, please see
http://www.pcavtech.com/soundcards/LynxTWO/Ph-loop-1644-xfus10.gif
or
http://www.pcavtech.com/soundcards/LynxTWO/index.htm
for the full report.

I think that most experts would consider < 5 degrees of phase error @20 KHz
to be more than adequate performance for *any* audio component. Most power
amps don't do that well.

>> Now of course we see lots about 96kHz sound - does this mean that
>> the above axiom is wrong, or is it more complicated than this - I
>> wondered if resonances are the answer, but currently cannot
>> understand the way it works
>
> I think we've surpassed the sonic benefits of first order that's for
> sure. 96kHz technology is sure as hell not a must-have. Spend it on
> better tranducers, techniques, and musicians. I agree with Arny is
> that this is innovation the industry is making to sustain itself, and
> not borne out of a real and dying need for rather small incremental
> fidelity impact.

Yes, "rather small" to say the least.

[Tom Page]

>
> Caveat: the nyquist rate (of 2x the max frequency) assumes a perfect
> low pass anti-aliasing filter with infinite dB/octave slope... which
> of course doesn't exist, and those that comes close introduce
> non-linear phase issues in the passband.

For someone who is not that great at electronics, what does thie mean?
Thanks for any explanation - I'm basically interested in what benefits
96kHz can offer, and what sort of "magnitude of improvement" (over
44/48kHz) it can offer, if we can objectively talk about improvement
in sound quality.

One thought I had - probably stupidly, was that if you do have a 20kHz
sound, which we'll presume is audible, then if you only sample it
twice(ish) for every oscialtion it makes (i.e. what 44kHz does) then
this surely won't be a very good way to store the sound - reducing
this to the most stupid level, this surely means sampling the sine
curve at just two points per period!

If I'm being stupid please tell me - I'm fascinated because I'm
interested - so any details about anything will please me!

Tom

[Arny Krueger]

"Tom Page" > wrote in message
m
>> Caveat: the nyquist rate (of 2x the max frequency) assumes a perfect
>> low pass anti-aliasing filter with infinite dB/octave slope... which
>> of course doesn't exist, and those that comes close introduce
>> non-linear phase issues in the passband.
>
> For someone who is not that great at electronics, what does this mean?
> Thanks for any explanation - I'm basically interested in what benefits
> 96kHz can offer, and what sort of "magnitude of improvement" (over
> 44/48kHz) it can offer, if we can objectively talk about improvement
> in sound quality.
>
> One thought I had - probably stupidly, was that if you do have a 20kHz
> sound, which we'll presume is audible, then if you only sample it
> twice(ish) for every oscillation it makes (i.e. what 44kHz does) then
> this surely won't be a very good way to store the sound - reducing
> this to the most stupid level, this surely means sampling the sine
> curve at just two points per period!
>
> If I'm being stupid please tell me - I'm fascinated because I'm
> interested - so any details about anything will please me!

You're not being stupid, just poorly informed.

There's only one sine wave that passes through more than 2 points (e.g.,
2.0000001 points per cycle). That means that those points uniquely define a
sine wave with given amplitude and phase. Thus both the amplitude and phase
of *any* sine wave is precisely defined by just over 2 data points.

This relates to what you may have learned in your high school math classes.
The equation of any sine wave is defined by exactly two parameters -
frequency and phase. As you know, you can solve an equation with two
unknowns if you are given two knowns. Thus you can determine the frequency
and phase of any sine wave with just over two data points.

BTW, I've done this experimentally on my test bench, and it works exactly in
accordance with theory.

[cyrus the virus]

In article >, (Todd H.) wrote:


> I think we've surpassed the sonic benefits of first order that's for
> sure. 96kHz technology is sure as hell not a must-have. Spend it on
> better tranducers, techniques, and musicians. I agree with Arny is
> that this is innovation the industry is making to sustain itself, and
> not borne out of a real and dying need for rather small incremental
> fidelity impact.

well put.

--
cyrus

[Lionel]

cyrus the virus wrote:
> In article >, (Todd H.) wrote:
>
>
>
>>I think we've surpassed the sonic benefits of first order that's for
>>sure. 96kHz technology is sure as hell not a must-have. Spend it on
>>better tranducers, techniques, and musicians. I agree with Arny is
>>that this is innovation the industry is making to sustain itself, and
>>not borne out of a real and dying need for rather small incremental
>>fidelity impact.
>
>
> well put.
>

I'm interested in this subject so I insist :
"what about the archivist point of view ?"
Thanks.

[Arny Krueger]

"Lionel" > wrote in message

> cyrus the virus wrote:
>> In article >, (Todd H.)
>> wrote:
>>
>>
>>
>>> I think we've surpassed the sonic benefits of first order that's for
>>> sure. 96kHz technology is sure as hell not a must-have. Spend it
>>> on better tranducers, techniques, and musicians. I agree with
>>> Arny is that this is innovation the industry is making to sustain
>>> itself, and not borne out of a real and dying need for rather small
>>> incremental fidelity impact.
>>
>>
>> well put.
>>
>
> I'm interested in this subject so I insist :
> "what about the archivist point of view ?"

What about it?

[Lionel]

Arny Krueger wrote:

> "Lionel" > wrote in message
>
>
>>cyrus the virus wrote:
>>
>>>In article >, (Todd H.)
>>>wrote:
>>>
>>>
>>>
>>>
>>>>I think we've surpassed the sonic benefits of first order that's for
>>>>sure. 96kHz technology is sure as hell not a must-have. Spend it
>>>>on better tranducers, techniques, and musicians. I agree with
>>>>Arny is that this is innovation the industry is making to sustain
>>>>itself, and not borne out of a real and dying need for rather small
>>>>incremental fidelity impact.
>>>
>>>
>>>well put.
>>>
>>
>>I'm interested in this subject so I insist :
>>"what about the archivist point of view ?"
>
>
> What about it?
>
>
Don't you thing that from the strict archivist point of view it could be
an interest to get the *best* possible "picture" of the signal the
technic allow you to archive ?

[Arny Krueger]

"Lionel" > wrote in message

> Arny Krueger wrote:
>
>> "Lionel" > wrote in message
>>
>>
>>> cyrus the virus wrote:
>>>
>>>> In article >, (Todd H.)
>>>> wrote:
>>>>
>>>>
>>>>
>>>>
>>>>> I think we've surpassed the sonic benefits of first order that's
>>>>> for sure. 96kHz technology is sure as hell not a must-have.
>>>>> Spend it on better tranducers, techniques, and musicians. I
>>>>> agree with Arny is that this is innovation the industry is making
>>>>> to sustain itself, and not borne out of a real and dying need for
>>>>> rather small incremental fidelity impact.
>>>>
>>>>
>>>> well put.
>>>>
>>>
>>> I'm interested in this subject so I insist :
>>> "what about the archivist point of view ?"
>>
>>
>> What about it?
>>
>>
> Don't you thing that from the strict archivist point of view it could
> be an interest to get the *best* possible "picture" of the signal the
> technic allow you to archive ?

Not at all. Practicality has something to do with it. If your thinking were
followed there would be nothing but pristine 70 mm films with multi-track
sound in the archives.

[Lionel]

Arny Krueger wrote:
> "Lionel" > wrote in message
>
>
>>Arny Krueger wrote:
>>
>>
>>>"Lionel" > wrote in message

>>>
>>>
>>>>cyrus the virus wrote:
>>>>
>>>>
>>>>>In article >, (Todd H.)
>>>>>wrote:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>>I think we've surpassed the sonic benefits of first order that's
>>>>>>for sure. 96kHz technology is sure as hell not a must-have.
>>>>>>Spend it on better tranducers, techniques, and musicians. I
>>>>>>agree with Arny is that this is innovation the industry is making
>>>>>>to sustain itself, and not borne out of a real and dying need for
>>>>>>rather small incremental fidelity impact.
>>>>>
>>>>>
>>>>>well put.
>>>>>
>>>>
>>>>I'm interested in this subject so I insist :
>>>>"what about the archivist point of view ?"
>>>
>>>
>>>What about it?
>>>
>>>
>>
>>Don't you thing that from the strict archivist point of view it could
>>be an interest to get the *best* possible "picture" of the signal the
>>technic allow you to archive ?
>
>
> Not at all. Practicality has something to do with it. If your thinking were
> followed there would be nothing but pristine 70 mm films with multi-track
> sound in the archives.
>
>
Yes but in the example you give the cost increase would be perhaps too
much important. In the case of audio signal the technic is cheap and the
support for archiving exists at a low cost.

[Arny Krueger]

"Lionel" > wrote in message

> Arny Krueger wrote:
>> "Lionel" > wrote in message
>>
>>
>>> Arny Krueger wrote:
>>>
>>>
>>>> "Lionel" > wrote in message
>>>>
>>>>
>>>>
>>>>> cyrus the virus wrote:
>>>>>
>>>>>
>>>>>> In article >, (Todd H.)
>>>>>> wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> I think we've surpassed the sonic benefits of first order that's
>>>>>>> for sure. 96kHz technology is sure as hell not a must-have.
>>>>>>> Spend it on better tranducers, techniques, and musicians. I
>>>>>>> agree with Arny is that this is innovation the industry is
>>>>>>> making to sustain itself, and not borne out of a real and dying
>>>>>>> need for rather small incremental fidelity impact.
>>>>>>
>>>>>>
>>>>>> well put.
>>>>>>
>>>>>
>>>>> I'm interested in this subject so I insist :
>>>>> "what about the archivist point of view ?"
>>>>
>>>>
>>>> What about it?
>>>>
>>>>
>>>
>>> Don't you thing that from the strict archivist point of view it
>>> could be an interest to get the *best* possible "picture" of the
>>> signal the technic allow you to archive ?
>>
>>
>> Not at all. Practicality has something to do with it. If your
>> thinking were followed there would be nothing but pristine 70 mm
>> films with multi-track sound in the archives.
>>
>>
> Yes but in the example you give the cost increase would be perhaps too
> much important. In the case of audio signal the technic is cheap and
> the support for archiving exists at a low cost.

Tell it to the archivists.

[Lionel]

Arny Krueger wrote:
> "Lionel" > wrote in message
>
>
>>Arny Krueger wrote:
>>
>>>"Lionel" > wrote in message

>>>
>>>
>>>>Arny Krueger wrote:
>>>>
>>>>
>>>>
>>>>>"Lionel" > wrote in message

>>>>>
>>>>>
>>>>>
>>>>>>cyrus the virus wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>>>In article >, (Todd H.)
>>>>>>>wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>>I think we've surpassed the sonic benefits of first order that's
>>>>>>>>for sure. 96kHz technology is sure as hell not a must-have.
>>>>>>>>Spend it on better tranducers, techniques, and musicians. I
>>>>>>>>agree with Arny is that this is innovation the industry is
>>>>>>>>making to sustain itself, and not borne out of a real and dying
>>>>>>>>need for rather small incremental fidelity impact.
>>>>>>>
>>>>>>>
>>>>>>>well put.
>>>>>>>
>>>>>>
>>>>>>I'm interested in this subject so I insist :
>>>>>>"what about the archivist point of view ?"
>>>>>
>>>>>
>>>>>What about it?
>>>>>
>>>>>
>>>>
>>>>Don't you thing that from the strict archivist point of view it
>>>>could be an interest to get the *best* possible "picture" of the
>>>>signal the technic allow you to archive ?
>>>
>>>
>>>Not at all. Practicality has something to do with it. If your
>>>thinking were followed there would be nothing but pristine 70 mm
>>>films with multi-track sound in the archives.
>>>
>>>
>>
>>Yes but in the example you give the cost increase would be perhaps too
>>much important. In the case of audio signal the technic is cheap and
>>the support for archiving exists at a low cost.
>
>
> Tell it to the archivists.
>
>
I hope some of them read this. ;-)