-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 After some more looking into the cue file format, I discovered it's possible to provide for the lead-in with an "INDEX 00" entry, making metaflac happy without patching (and apparently it's illegal to have a cue file that *DOESN'T* start at 00:00:00 anyway), and without having to add wacky work-arounds to store the lead-in value.
Seek point markers in the flac file also get the right values when tagging the file after encoding (which I can notice now that I added -S X to FLACOPTS in my abcde.conf file :-). Several minor changes were required, both to abcde and to the mkcue program. For convenience, I've divided my changes into several small patches, all attached as a single gzipped tar file: # tar -tzvf abcde.patch.tgz - -rw-r--r-- root/root 3432 2006-07-07 14:19:57 abcde.cdrom.patch - -rw-r--r-- root/root 628 2006-07-07 14:13:42 abcde.cuefile.patch - -rw-r--r-- root/root 499 2006-07-07 14:12:05 abcde.leadin.patch - -rw-r--r-- root/root 1136 2006-07-07 14:18:15 mkcue.patch abcde.cdrom.patch is my previous patch to allow spaces in the $CDROM device name (required if you allow spaces in your flac filenames and want to rip from the flac file instead of the CD). abcde.leadin.patch is a simple patch to pass cdparanoia the range '0-' instead of '$FIRSTTRACK-$LASTTRACK' when doing a onetrack do_cdread, which causes any lead-in to be ripped along with the audio tracks. abcde.cuefile.patch is a simple change to the do_cleancue routine that allows for more than one INDEX entry per track. It also tries to be as picky as possible about recognizing the actual TRACK entries (regex rooted to the start of line, with optional whitespace) in case someone has a cuefile with PERFORMER, TITLE, or other fields that actually contain "TRACK" somewhere. mkcue.patch modifies the mkcue program to output a prelead INDEX entry for any disks with the first track starting at something other than the default 150 frame (2s) LBA address rather than blindly normalizing the first track's start-point to zero. The 150 frame offset is currently hard-coded, as I didn't see any obvious include files that would provide this value (such as CD_MSF_OFFSET from cdrom.h in linux). I can now run abcde -d <flacfile> and get *EXACTLY* the same cddb info pulled up with the physical CD. More importantly, there is now enough data to create an exact duplicate of the audio portion of the CD should it ever get lost, shredded, eaten, abducted by aliens or otherwise destroyed! :-) HAPPY RIPPING!!! - -- Charles Steinkuehler [EMAIL PROTECTED] -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (MingW32) iD8DBQFErrnKenk4xp+mH40RAjieAJ45HzVNVPhmnl5K9wP5BEfZx2lFnACgytEJ nPO6OLAy1ZGz07Ut1VWMTpg= =Y6FM -----END PGP SIGNATURE-----
abcde.patch.tgz
Description: application/compressed
