Re: Dither source code?

On Wed, 03 Dec 2003 05:16:12 GMT, Bob Olhsson <[EMAIL PROTECTED]>
wrote:

>In article <[EMAIL PROTECTED]>, Don Pearce
><[EMAIL PROTECTED]> wrote:
>
>>Dither will be performed by the sum of
>>all noise-like signals reaching the ADC.
>
>Not at all, the noise could easily not be random enough or cover the
>proper spectrum. "Self dithering" is a common misconception that is
>just plain wrong.

Noise is always more random than the PRBS used for generating a dither
signal. And I will repeat, what does the dithering is not a matter of
choice - it is whatever is there. If the spectrum is not what you
would have liked, then tough, I'm afraid. When digitising vinyl, the
dominant noise, by a long, long way comes from the vinyl itself.

And as for your claim that the concept is wrong, perhaps you could
explain exactly how an ADC knows the difference between noise that
accompanied the signal already and noise that accompanies the signal
because it has just been added. There is of course no difference
whatever.

d

_____________________________

http://www.pearce.uk.com