Both audio and data CDs contain some redundancy that allow correcting single
bit errors and detecting more serious errors. Correcting a single bit error
should be undetectable audibly or in a data stream. Data is also
distributed, to some extent, so that a dropout that would affect many bits
is spread out somewhat.
If an error more serious than that occurs, audio players will interpolate
between neighboring samples. The distribution of the error means that the
interpolation will happen several times to single samples in different
places rather than having one audible pop several samples long. For low
frequency signals interpolation is not too bad, of course it affects higher
frequencies more.
A data CD, obviously, cannot interpolate if the error-correction fails. On
a computer, though, you can retry a read, which often works. I'm not sure
if higher end CD players retry reads (if there's a buffer it would be
possible) but I expect consumer audio CD players don't.
...Craig
----- Original Message -----
From: "Mike Marsh" <mmarsh@...>
To: <motm@yahoogroups.com>
Sent: Tuesday, November 05, 2002 8:53 PM
Subject: [motm] Re: digital source and its medium puzzle- the answer
> > For Data CD`s it doesn`t matter as much, but
> > for audio, it matters a lot.
>
> I disagree. If ∗one∗ zero is interpreted as a one on a data CD, it
> could screw up the whole program, if it's an executable. Data CDs
> are much less forgiving; with audio, our ear integrates.
>
> I think.
>
> Mike
>
>
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>