[paranoia] sample offset?

[Leigh Orf]

Monty wrote:

|   > Hmmm.  Not much traffic here.  Is anyone listening?  I only
|   > got one response to my first message (Thanks, Keith), and
|   > none to my second.
|
|   Folks are here, there just hasn't been much happening on
|   paranoia development.
|
|   I'm actually putting some time into a cdparanoia update right
|   now, and I'm evaluating and applying patches that have come
|   in over the past year.  Sample-offset ripping seems like
|   something best done by an editing program, but I'll still
|   look at the sample-grained patches that have come in.
|
|   Monty

Monty,

Concerning the editing program approach, I've done this and it sucks!
Having to rip a CD, paste the wav files together, and then separate them
out again is a real pain and as far as I can tell only doable under
Windows software using expensive software. I know Exact Audio Copy, the
Windows cdparanoia equivalent, allows for this offset issue and can get
'clone' extracts.

Honestly, it doesn't bother me that my cd burner is consistenly off
by 1153 samples when extracting. However, in the world of audio music
trading (see http://etree.org for instance) a lot of people are really,
really anal about getting exact clones of shortened wav files (shorten
is a lossless audio compression program that gets about 2:1 compression
on wav files). I do a lot of ftp trading via the etree and have to deal
with this well-meaning but misplaced concern about getting exact clones
of their original files.  I can see their point since a lot of "newbies"
don't use EAC or cdparanoia and when trading produce CD-Rs that have the
clicks and pops etc. that probably motivated you to write this software
in the first place. By using md5sum files generated with the original
shn files, they can download and run md5sum on the files and if they
match, rest assured that they at least got clones of the original (of
course this says nothing of problems with the original, but I'm not
going to beat that dead horse in this forum!).

So, many people who trade this music will archive .shn files (losslessly
compressed wav files) on iso9660 cdroms. Then they don't have to rip
anything, just copy the files over when trading and everything matches.
The md5 files can be used as fingerprints which verify the 'pedigree' of
the audio.

Since I use cdparanoia, I always get good extracts, but because of
the offset issue, shn files created from my cdparanoia extracts all
fail md5sum checks from the md5sum files generated from the original
shn files. However, I simply will not archive my shn files when I get
perfect audio ripping with cdparanoia just to make some people happy...
plus I don't have the time and it's just a waste of media! But even when
people get my flawless, offset copies they get nervous and they think
they are getting inferior music. It also ends up introducing a whole new
set of md5files and since some other well meaning people feel the need
to archive the original md5sum files, this confuses matters more.

Anyway, that's my reason for wanting to be able to have cdparanoia deal
with the offset issue. I only recently discovered that this offset is a
constant integer for a given drive (or am I wrong here? It must be or
wouldn't there always be clicks between DAO tracks?) so it seems to be
a logical thing to do... it would be super to make it possible via a
command line option.

Plus, having this option would make it easy to prove that cdparanoia is
indeed doing bit-perfect copying just by running 'diff' or md5sum.

Thanks for listening. It would make me very happy to be able to rip a
CD, shorten the tracks, and have the original md5sum files match... it
would save me a lot of grief. So that is my selfish motive! And it does
seem to be a relatively easy patch.

Leigh Orf
http://orf.cx

--- >8 ----
List archives:  http://www.xiph.org/archives/
Paranoia homepage: http://www.xiph.org/paranoia/
To unsubscribe from this list, send a message to 'paranoia-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.