Tim S Kemp wrote:

Stewart Pinkerton wrote:

Whaaaat? That's rubbish, any modern 44.1k ADC is ruler flat to at
least 20kHz.

Never measured it, but the waveforms of perfectly audible stuff8khz will
be somewhat distorted at 44k sampling compared to higher rates.


Well I have measured it, and the distortion and I can tell you for a fact
that what you just said is definitely incorrect.

Ian
--
Ian Bell

"Tim S Kemp" wrote in message

Stewart Pinkerton wrote:

Whaaaat? That's rubbish, any modern 44.1k ADC is ruler flat to at
least 20kHz.

Never measured it,

So I was right, the cause of these weird claims is ignorance.

but the waveforms of perfectly audible stuf 8khz
will be somewhat distorted at 44k sampling compared to higher rates.

So what?

The ear can't hear the difference between a 8 KHz sine wave and a 8 KHz
square wave if you match levels properly.

"Ian Bell" wrote
Tim S Kemp wrote:

Stewart Pinkerton wrote:

Whaaaat? That's rubbish, any modern 44.1k ADC is ruler flat to at
least 20kHz.

Never measured it, but the waveforms of perfectly audible stuff8khz will
be somewhat distorted at 44k sampling compared to higher rates.


Well I have measured it, and the distortion and I can tell you for a fact
that what you just said is definitely incorrect.

Well, he's more talking about Fourier theory, and from that standpoint,
he's right.


Martin
--
M.A.Poyser Tel.: 07967 110890
Manchester, U.K. http://www.fleetie.demon.co.uk

Fleetie wrote:

"Ian Bell" wrote
Tim S Kemp wrote:

Stewart Pinkerton wrote:

Whaaaat? That's rubbish, any modern 44.1k ADC is ruler flat to at
least 20kHz.

Never measured it, but the waveforms of perfectly audible stuff8khz
will be somewhat distorted at 44k sampling compared to higher rates.


Well I have measured it, and the distortion and I can tell you for a fact
that what you just said is definitely incorrect.

Well, he's more talking about Fourier theory, and from that standpoint,
he's right.


Martin

No he isn't. a typical 44.1K system can reproduce up to 20KHz with no more
distortion than at any other frequency. There is nothing special about the
way it reproduces signals about 8KHz.

Ian

--
Ian Bell

Hi,

In message , Don Pearce
writes
On Wed, 3 Nov 2004 18:47:55 +0000, Glenn Booth
wrote:

Hi,

In message , Tim S Kemp
writes
Stewart Pinkerton wrote:

Whaaaat? That's rubbish, any modern 44.1k ADC is ruler flat to at
least 20kHz.

Never measured it, but the waveforms of perfectly audible stuff8khz will be
somewhat distorted at 44k sampling compared to higher rates.

Depends on the waveform. It's not true for sine waves. It would be true
for waveforms that include higher harmonics that are out of band (like a
square wave), but that's another story - those components wouldn't be
within the nyquist limit.

Just to be a bit pedantic, whatever the shape of the waveform,
everything below 20kHz is still reproduced perfectly.

Indeed. Pedantry accepted :-)

--
Regards,
Glenn Booth

Get back to us when you actually know what Fourier is all about.

You only need two samples per cycle to reproduce a given pure
frequency (sine wave) perfectly.

To reproduce a 4kHz sine wave, you only need to sample at 8kHz.


Martin
--
M.A.Poyser Tel.: 07967 110890
Manchester, U.K. http://www.fleetie.demon.co.uk

Tim S Kemp wrote:
Stewart Pinkerton wrote:


Whaaaat? That's rubbish, any modern 44.1k ADC is ruler flat to at
least 20kHz.


Never measured it, but the waveforms of perfectly audible stuff8khz will be
somewhat distorted at 44k sampling compared to higher rates.


excuse me? If the "waveforms of perfectly audible stuff 8khz" do not
have frequency content above 22 kHz (and assuming a good anti-aliasing
filter and reconstruction filter is used in the output stage) then the
output of the sampled waveform will not be distorted at 44kHz sampling
rates.

On Wed, 03 Nov 2004 20:35:30 GMT, "Fleetie"
wrote:

Get back to us when you actually know what Fourier is all about.

You only need two samples per cycle to reproduce a given pure
frequency (sine wave) perfectly.

To reproduce a 4kHz sine wave, you only need to sample at 8kHz.

Before someone jumps all over this, he meant 8.01 k........
--

Stewart Pinkerton | Music is Art - Audio is Engineering

In article , Tim S Kemp
wrote:
Stewart Pinkerton wrote:

Whaaaat? That's rubbish, any modern 44.1k ADC is ruler flat to at
least 20kHz.

Never measured it, but the waveforms of perfectly audible stuff8khz
will be somewhat distorted at 44k sampling compared to higher rates.

Only if the sample values are incorrectly and incompetently used by the DAC
and filtering arrangements used for signal reconstruction.

From information theory you can use formal mathematical methods to prove
that the entire waveform shape is represented by the sample values provided
the Sampling Theorem was satisfied when the samples were recorded. This
means the DAC and reconstruction filter can recreat the waveform *without*
the 'distortions' you assume exist.

FWIW as well as having looked at, and used, the theory I have also:

1) tested players and looked at the output. The results show your
assumption is incorrect. Hence I *have* measured it, and can say that the
results show you mistaken.

2) Produce a number of examples that I've used in my undergrad lectures for
years. These also show that your assumption is incorrect.

3) published a book that deals with these issues in detail. Copies now
available in paperback. ;-

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

In article , Fleetie
wrote:
"Ian Bell" wrote
Tim S Kemp wrote:

Stewart Pinkerton wrote:

Whaaaat? That's rubbish, any modern 44.1k ADC is ruler flat to at
least 20kHz.

Never measured it, but the waveforms of perfectly audible stuff8khz
will be somewhat distorted at 44k sampling compared to higher rates.


Well I have measured it, and the distortion and I can tell you for a
fact that what you just said is definitely incorrect.

Well, he's more talking about Fourier theory, and from that standpoint,
he's right.


I disagree. For example, a 12 kHz sinewave is "perfectly audible stuff
sic above 8 kHz". Yet should be reproduced accurately.

As should any waveform composed of an array of components between 8 and
20 kHz that does not clip.

The point is that the recording (series of samples) should form a 'complete
record' if made according to the Sampling Theorem. This means that any
signals above 22 kHz will not be present.

So you can say, "not signal components above 22 kHz should be present". But
this does not mean your comment about "Fourier theory" explains what he
said w.r.t. "above 8kHz".

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html