🚀 go-pugleaf

gaby@Stars.Reston.Unisys.COM ( UNISYS) writes:

%I think the oversampling is not time interpolation

Take it from someone who's played with the innards of more than a couple
CD players that it is.

%(which by Nyquist
%does not add any more information to the originial signal)

By "real-world" electronics, it doesn't add any more info, but allows
you to build an analog output filter that doesn't take away more info.

%, but more 
%error correction oversampling.  I.e. the same bit is sampled multiple
%times to determine its value.

No need to do this... it's usually real obvious or the Reed-Solomon
codes will allow recovery of the original info if a couple bits are
incorrect.

%I assume that this is done by sampling
%over the duration (space on the CD) of the bit.

The bits are not sampled off the disc like analog data.  They are
read off the disc much like serial data comes out of your modem
into your computer.

%I think it is a little generous to think that the industry truely 
%does "oversampling" as it is implied by signal processing connotations.
%This (as you say) requires more compute requirements to ensure proper
%interpolation of the sampled data.  If the interpolation (filtering)
%is done wrong, then the quality of the output would go down...

But it isn't done wrong and it is done on the fly.  There are a lot of
specialized chips on the market for just this purpose.  I used one of
them in my DSP board.  Remember, these chips are just mini-computers
with a built in program and they generally run at about 2 MHz.  They
only need to produce an update every 1/88kHz.  Even for a microprocessor
running at 2 MHz, that's all day long.

-- 
Todd Day  |  todd@appmag.com  |  appmag!todd@hub.ucsb.edu
		  ^^^^^^^^^^ coming soon!