Some drives do not support reading into the lead-out. However, in conjunction
with a positive sample-offset (particularly, if greater than 588), CDParanoia
may blindly try to read into the lead-out and thus generate errors while doing
so. Rubyripper must thus handle these drives carefully so as to rip all of the
music which can be ripped without reading into the lead-out.
This will require some careful splicing of wave files (which is already needed
for Issue 220 and Issue 240) as I expect that a positive sample-offset which is
not evenly divisible by 588 will result in missing samples even if the end of
the rip is adjusted so as to not read into the lead-out. (If the end of a rip
is specified explicitly to avoid reading into the lead-out after taking the
sample-offset into account, some samples may end up getting ignored between
that adjusted time and the real lead-out). Thus, in such a scenario, the last
sector will need to be read again without any offset adjustment, and then any
missing samples spliced in.
This will also require fixing Issue 490, as any de-facto truncation of the
reading to provide for this splicing will require an explicit end sector to be
specified to CDParanoia.
Please note that the remaining samples should optionally be either filled in
with zero-samples or truncated, per another preference which will need to be
set (Equivalent to EAC's "Fill missing offset samples with silence").
A similar fix might be needed for negative offsets, but a lack a drive with
which to test this.
Original issue reported on code.google.com by comradec...@gmail.com on 1 Nov 2011 at 5:43
Original issue reported on code.google.com by
comradec...@gmail.com
on 1 Nov 2011 at 5:43