"Jan Philips" <judmccrNOSPAMM@bellsouth.net> wrote in message
news3jmfvs3jgopb5fsjancrj93d45oscjhis@4ax.com
Nor does an A/D.

But you don't need sort of complexity that to just record, play back, and
transmit music.

The interpolation is generally done via a conceptually simple but
steep-sloped low-pass filter.

A conceptually-simple low pass filter suffices wonderfully. The low pass
filter gets a little complex when one tries to have good filtering at and
above Nyquist, but smooth response below. Nevertheless, its now usually
mostly done in the digital domain via oversampling, and has low analog parts
count.

I knew someone like Arny would jump in. Since I do this stuff for a living,
re-stating the obvious is torture.

"Gareth Hardy" <gareth@sentientsolutions.freeserve.co.uk> wrote in message
news:bdem2b$646$1@newsg1.svr.pol.co.uk...

"FDR" <_remove_spam_block_rzitka@hotmail.com> wrote in message
news:33GKa.117560$zm1.78420@twister.nyroc.rr.com
Yes, it's the site's home page and every page on the site is linked out of
it, directly or indirectly.

You've never heard of doing things this way before?

I guess the idea that a web site has a home page and all other pages on the
site are linked out of it is way over your head.

But, for some odd reason, that's how most of the WWW world is made. Since
your powers of observation seem to be at a low ebb, you'll have to trust me
about this...

Arny,

I think these people are reading a lot of junk about perceptually based
music encoding and
crossing it into their conception of how PCM sampling works.

"Arny Krueger" <arnyk@hotpop.com> wrote in message
news:tqydnZxatblGxGajXTWJkA@comcast.com...

On Thu, 26 Jun 2003 16:38:50 -0400, "Arny Krueger" <arnyk@hotpop.com>
wrote:

Can you briefly explain what a "conceptually-simple low pass filter"
is and how it gets from digital to analog? I don't know much about
electronics but I do know that a low-pass filter lets low frequencies
through. But I don't see how that goes from digital to analog.

If the original is a sine wave (under the Nyquist limit), it is
digitally sampled, and a D/A conversion is done with the
conceptually-simple low pass filter, you get back a sine wave (not
some step approximation to a sine wave), right?

In mathematical information theory, info, as you call it, is that that
reduces uncertainty. Therefore, no information is lost
in digital sampling. You have the audiophiles erroneous belief that the
analog signal is somehow 'true' and 'complete' and the
digital signal is not. From the standpoint of mathematical information
theory, the digital signal is complete.

"FDR" <_remove_spam_block_rzitka@hotmail.com> wrote in message
news:fysKa.114946$zm1.88992@twister.nyroc.rr.com.. .

"Arny Krueger" <arnyk@hotpop.com> wrote in message
news:-xadnaUCnJ0Lx2ajXTWJjA@comcast.com...
You call it a page, then you get pissed at those that call it a page and not
a website. If you thought it was a website, then why don't you say so
yourself????

"Arny Krueger" <arnyk@hotpop.com> wrote in message
news:jDWdnRJhqK9sgmajXTWJgA@comcast.com...
Not really. I just stand behind my opinion as you do (and that simple
opinion is that 96k is better than 44.1k). And by doing so, I do not deserve
to be called a craphead...

Good ones are always expensive. And just maybe I live in Croatia where 300$
sound card is stated as luxury in gonverment laws!

I have my ears, and don't need programs to test them. And concerning
lessons, I had mine more than enough...

Now I saw FDR's post that owner of the page is actually you. No wonder you
are byting everyone who says differently. People have different opinions
about certain subjects. And in audio, sometimes you can be right or wrong.
As you believe I'm wrong, I believe you are too. So what?

Do not catch me by a single word. pcabx IS a single web page/site/whatever.
I am here to learn from people and share knowledge. Not to try to convince,
by all means, somebody in his possible denial of truth as you are firmly
doing last couple of posts.

You can try to defend your statements. Not by, as you energically did,
denying others...

Anyhow, my appologies if U felt insulted in any way. It was not intended.
And next time, please, have more patience...

"FDR" <_remove_spam_block_rzitka@hotmail.com> wrote in message
news:97KKa.118420$zm1.20462@twister.nyroc.rr.com.. .
Sigh...at once you invoke both empiricism and mysticism. Can't argue with
that (laugh).

"Jan Philips" <judmccrNOSPAMM@bellsouth.net> wrote in message
news:u6nmfvso5oaeoj8hhm5kk1qdesg7e0rjhg@4ax.com
Ever hear of google?

Like Buckaroo I get tired of doing people's homework for them.

right.

Well do everyone a favour then, and keep quiet when you notice someone is
wrong. It's selfish twice over to just state something is "tripe" and then
wait for someone else to fill in the gaps. Some of us use UseNet to try and
learn.

"Buckaroo" <Dipper_CA@earthlink.net> wrote in message
news:bdgk0l$1lp$1@gladiola.noc.ucla.edu...
If common sense and life experience are mysticism, then I'm guilty. If my
engineering background is empiricism, then I'm also guilty.

Once again, what do you have against higher sampling rates?

Oh, and here's what Dallas Semiconductor has to say :

"Now moving into the mathematical realm, assume the wheel is a unit circle
with sine and cosine coordinates. If one samples at the positive and
negative peaks of the cosine values (which are 180 degrees out of phase),
then the Nyquist criteria is met and the original cosine values can be
reconstructed from the two sampled data points. Thus the Nyquist limit is
essential in reconstructing the original signal. As more and more points are
added, the ability to replicate the original signal improves."

http://www.dalsemi.com/appnotes.cfm/appnote_number/928

Wow, looky there, more points = improvement.

On Fri, 27 Jun 2003 06:03:04 -0400, "Arny Krueger" <arnyk@hotpop.com>
wrote:

And although the Nyquist theorem applies to Fourier transforms, and
you said that the Nyquist theorem is at work here, the Fourier
transform is _not_ involved?

On Thu, 26 Jun 2003 13:41:17 -0700, Buckaroo wrote:

Then you'd be the perfect person to write up a web site on the technical
aspects of digital audio. Then whenever someone asks a question you can
point them to your web site. How does that sound?

--
Ian.

EOM

"Ian Hastie" <ian_a_hastie@hotmail.com> wrote in message
newsan.2003.06.27.15.49.43.513000@hotmail.com...
It would seem obvious, but this type of person doesn't seem interested in
teaching.

On Fri, 27 Jun 2003 10:09:11 -0400, Jan Philips
<judmccrNOSPAMM@bellsouth.net> wrote:

Doesn't the Nyquist frequency also rely on using a sinc function for
reconstruction? It's been awhile since I've worked with this stuff
academically of professionally (hence keeping out of this thread except to
call someone on the still-false claim that DVD-Rs are cheaper or even as
cheap per unit storage than CD-Rs), but I seem to recall that perfect
reconstruction of a signal sampled at or above 1/2 its max frequency required
the use of a perfect sinc function, which is not so easy to replicate in real
life (and is usually approximated). IF my memory on this is correct, then it
seems likely that while information technically isn't lost in the ADC
process, it may be lost in the DAC process. Whether or not that loss (if
present) would be sufficient to be audible to an untrained (or even trained)
ear, I dunno.

Given that Arny and "Buckaroo" are so adamant about 44.1kHz sampling being
absolutely perfect (forgetting bit depth for the moment) for frequencies up
to about 20kHz, I'm definitely interested in their thoughts as to what I'm
recalling incorrectly.

--
Erik Harris n$wsr$ader@$harrishom$.com
AIM: KngFuJoe http://www.eharrishome.com
Chinese-Indonesian MA Club http://www.eharrishome.com/cimac/

The above email address is obfuscated to try to prevent SPAM.
Replace each dollar sign with an "e" for the correct address.

Erik Harris <n$wsr$ader@$harrishom$.com> writes:
You and me both.

Well, there's also that pesky matter of that band limiting filtering,
the required steepness of it to effectively eliminate aliasing
components, and the phase anomalies the filter presents in real life.

That (as I recall) is why practically oversampling is typically a Good
Thing--it relieved the burden of that perfect low pass filter, and
allowed it to be a) farther from the passband, b) less steep and c)
consequently able to introduce fewer phase anomalies than otherwise
possible.

Myself as well. There's a big difference bewteen the academics of the
Nyquist rate in discrete mathemtics theory, and the realities of the
componentry and actual filters one can manufacture and wrap their ears
around.


Best Regards,
--
Todd H.
http://www.toddh.net/

You need to learn about frames of reference. The quote is correct....but
you cannot hear the improvement, which is only an
improvement in the physical frame of reference. You can find
double blind tests that prove this conclusively if you will Goggle for them.

Engineers are the worst offenders and failing to understand the slightest
thing about psychoacoustics.

Quite true. The reason that the sample rate was set at 44.1 Ks/sec was to
allow for
filter roll off at the Nyquist at the upper limit of hearing, back when the
designs called
for analog output filters after the sample and hold. With improved digital
filtering,
oversampling, and other algorithms, it is a moot point now.

As an aside:

Most individuals in industrialized nations have severe attenuation of
hearing acuity under laboratory conditions above
ca. 16, 000 Hz. Another point to consider.

I only teach post docs.