On 05/10/06, James Leifer <<a href="mailto:James.Leifer@inria.fr">James.Leifer@inria.fr</a>> wrote:<div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> "Steve Flynn" <<a href="mailto:anothermindbomb@gmail.com">anothermindbomb@gmail.com</a>> writes: > How will the lookup program know which CD these are WAV tracks of? > When you are ripping a CD and you perform a CDDB lookup, the ISRN is > read from the CD TOC and that is used to identify the CD in the CDDB > database. Now that you've thrown that information away and have > nothing but a big pile of WAV files, you've got no way of telling what > that CD was... in essence, you've thrown away all of your metadata. Hmm... A quick googling suggests that a cddb query consists of the hash of the lengths of all the tracks on the CD. Maybe I've misunderstood, but if this is true then a pile of *all* the wav files from a single CD should yield this information nicely. </blockquote></div> I don't see how that can be accurate (although if that's the way CDDB works, then it's obviously good enough). I've always looked up the ISR number from the CD TOC myself. When I've ripped a wav file twice and then compared the two copies, there have always been slight diffrerences, due to jitter, silence cut-off points, error correction routines and so forth. The lengths of the two files may or may not be precisely the same.... this is why I didn't think using a hash of the track lengths would be particularly accurate. I'm happy to be corrected however - if it works with hashed lengths, then topper! :) -- Steve Despair - It's always darkest just before it goes pitch black...