Indeed. I used to have terrible trouble doing it on my previous machine; if I read at anything above 1x (i.e. normal audio-playing speed) I got blocks of data transposed and repeated and generally put in the wrong place, but no errors reported. These days we have cdparanoia which appears to do a pretty good job of dealing with that sort of oddity, and I wouldn't be surprised if my current CD drives were better put together than my old one as well; but even so there's the odd problem. It's silly.

LNR's CDRW seems to read much more cleanly than my CD ROM does. I wonder if the analogue output is good enough to feed that back into the sound card and use that instead?

Don't think I haven't considered it! But first I'm trying the much simpler trick of getting in touch with someone I happen to know has the same CD, and seeing if his copy is undamaged :-)

The downside of the analogue approach is that it would be horrendously slow. It's what we'll have to do when we start on digitizing the vinyl though.

Yes, that doesn't sound like fun... Though "horrendously" seems a bit strong; I'm currently ripping CDs at an average rate of 2 to 2.5 times normal audio playing speed, so a factor of two off that doesn't seem that much worse.

(Bah. And it turns out the guy I thought had the same CD in fact doesn't. Time for a more creative approach, I fear.)