This page is a draft.

You may have found that some discs did not read correctly on the first attempt. You can pass different arguments to ddrescue to try and coax more sectors out of the discs. An example of a second-stage command line would be

 ddrescue -d -r1 -b2048 /dev/cdrom my_files.iso my_files.log

The key difference here is attempting to use direct mode IO, rather than buffered and telling ddrescue to do one further retry. You can also try -R, for reverse read mode. I am not sure that this helps with optical discs (but it's unlikely to hurt).

If you have access to more than one optical drive, it is definitely worth trying to read the troublesome discs from multiple drives. In my experience, older drives seem to be better at reading older discs. I choose to copy the in-progress image and log file to the other computer and run ddrescue as illustrated, but it is possible to do a completely separate rip attempt and try to combine the results later. I prefer to keep it simple.

Retries like this can take a long, long time: I've left one running for a full week and managed to recover a further 20M or thereabouts of data. Depending on how many drives you have access to, what else you are up to and how many discs you have to try and recover, you could orgnaise your "pipeline" such that one machine was performing slow retries whilst another was performing the faster initial reads of other discs.

Further extraction of data when you lack a ToC

TODO: photorec, etc.