[EMAIL PROTECTED] wrote:
>
> Ralph Smeets wrote:
>
> > Magic, Magic, Magic,
> >
> > a 22kHz sine-wave sampled at 44kHz stays a sine wave. If you would know
> > anything about Discrete Signal Processing you would now. (AD/DA conversion
> > isn't just connecting dots!!!!)
> >
> > Cheers,
> > Ralph -> very sleepy...
>
> That could be a real bastard for synthesized music at high frequencies then - most
> of it consists of sharp-edged waves not dis-similar to triangle waves. The degree to
> which the sine wave is reproduced also depends on the degree of oversampling used if
> I remember correctly. I actually meant a square wave not a triangle wave, but it's
> the same principal, just a different distortion pattern.
>
Hmm,
some DSP theory (applied in ALL MD equipment cappable of A/D D/A coversion (a
little bit simplified in order to keep the mathematic mungle out):
analog signal => low-pasfilter @ 20 kHz => ADC @ sample-frequency =>
low-passfilter in the digital domain => resampler => oversampler => DAC =>
low-pasfilter @ maximum output frequency!
Ie, in numbers:
Analog signal:
F(t)=cos(2*pi*2*10e3*t)-cos(2*pi*3*10e3*t)+cos(2*pi*4*10e3*t)-cos(2*pi*5*10e3*t)
sample-frequency : 176.8kHz (4 times oversampling at the ADC)
low-passfilter at 20kHz, -6dB
Ie, this gives a analog signal of:
F(t)=cos(2*pi*2*10e3*t)-0.5cos(2*pi*3*10e3*t)+0.25cos(2*pi*4*10e3*t)-0.125cos(2*pi*5*10e3*t)
After the ADC, we get a signal represented by a z-transformation. Something
that goes beyond the scope of this list..
Than we filter the digital signal with a digital filter. Ie, we trow every
frequency higher than 20kHz out of the signal:
F(nT)=cos(2*10e3*nT)
We some oversampling (4x, ie 176.8kHz):
F(nT)=cos(2*10e3*nT) (stays the same!)
We convert this digital signal into the analog domain: (A single frequency in
the digital domain translates into the same frequency in the analog domain + the
same frequency at sample-frequency + the same frequency at 2 * sample-frequency
+ the same frequency at sample-frequency + 3* the same frequency at
sample-frequency + etc....
F(t)=cos(2*pi*2*10e3*t)+cos(2*pi*197*10e3*t)+cos(2*pi*372*10e3*t)....
Then we pas this signal by a low-pasfilter @ 20kHz =>
F(t)=cos(2*pi*2*10e3*t)
Ie,
All frequencies higher than 20kHz have been removed! Our strange (wheter that is
triangle/square/puls shaped) signal has become a sinus!
Cheers,
Ralph
--
=======================================================================
Ralph Smeets Functional Verification Centre Of Competence - CMG
Voice: (+33) (0)4 76 58 44 46 STMicroelectronics
Fax: (+33) (0)4 76 58 40 11 5, chem de la Dhuy
Mobile: (+33) (0)6 82 66 62 70 38240 MEYLAN
E-Mail: [EMAIL PROTECTED] FRANCE
=======================================================================
"For many years, mankind lived just like the animals. And then some-
thing happened that unleashed the powers of our imagination: We learned
to talk."
-- Stephen Hawking, later used by Pink Floyd --
=======================================================================
-----------------------------------------------------------------
To stop getting this list send a message containing just the word
"unsubscribe" to [EMAIL PROTECTED]