[Libcdio-devel] libcdio cd-paranoia: fix to rip hidden track one audio

[R. Bernstein]

Eric Shattow writes:
 > Hi,
 > 
 > libcdio/cd-paranoia is a useful example. However, it refuses to rip audio
 > from the pregap of track 1 on my 'blind melon' / 'soup' pressing.
 > 
 > To work-around, I comment out the 'exit(1);' line where a check is made... {
 > if !cdda_track_audiop() }   line 1057 in cd-paranoia.c from
 > libcdio-0.78sources. After the work-around it rips the audio okay.

 > 
 > The vanilla cdparanoia command has no complaints ripping this span of the
 > album, no complaints about it not being audio. For your reference, I run
 > using the following commandline:
 > 
 > (cdparanoia) -d $DRIVE -O $OFFSET --never-skip -Z [00:00:00.00]- $ALBUM.wav

Hmmm. At least on GNU/Linux what should be happening should be
identital. Vanilla cdparanoia makes a check in its code. And the
cdda_track_audiop test is basically checking to see if a bit in
ctde_ctrl is set which is done pretty much the same way. So what kind
of track does cd-info report the track is?
 > 
 > 
 > By the way, the libcdio cd-info indicates that something needs a clean-up
 > for amd64...  "Audio CD, CDDB disc ID is ffffffffce0ad30e".  That should be
 > outputting is a 32-bit cddb sum.

Ok. Thanks. I've made the change.

 > 
 > These are the kind of things I don't want to write a patch for, I just would
 > like to point out - hopefully you have a much better idea than I do about
 > what the proper fix will be.

 > 
 > Always happy to test out and discuss.

As for fixing cd-paranoia, what would be helpful is to understand why
libcdio is getting different results.  You may be in a better position
to determine this. For cdparanoia the line that does the test is around
151 of interface/toc.c and for libcdio on GNU/Linux it is around 525 of 
lib/driver/gnu_linux.c (get_trace_format_linux)

 > 
 > Eric
 > 
 > P.S.  The code I write is in Ruby language, my pet project is F/OSS
 > sample-accurate audio ripping. 

There are Ruby bindings for libcdio on
rubyforge.net. http://rubyforge.org/projects/rbcdio/ 

This was the Ruby code I wrote; a lot could be improved. 

So far there's a lot to do with EAC program
 > and AccurateRip checksum being the industry standard, and they're
 > closed-source and single platform. I would want to develop my applications
 > to run on many systems (cdio looks like the way to do this). If you're
 > interested I have a webpage set up with links to my work,
 > http://osclab.net/foss.html

Looks interesting. The libcdio project is always looking for
contributions. Maybe you could write a EAC CUESHEET image parser for
libcdio?