[clug] Recovering old CD-Rs

[Simon Oxwell]

Hi Scott,

Thanks for you lengthly reply.

On 18 February 2015 at 22:58, Scott Ferguson <scott.ferguson.clug at gmail.com>
wrote:

> On 18/02/15 12:38, Simon Oxwell wrote:
> > I'm going through a bunch of old CD-R's (some of which would be about 15
> > years old) and ripping them to ISO's while I still can, using GNU
> ddrescue.
> >
> > However, I'm having troubles with a couple of disks that obviously have
> > something burnt to them, but the CD-ROM (and DVD-ROM) drive isn't
> > recognizing them as valid media, as presumably the disk's header has
> become
> > corrupt. Any ideas as to how I might be able to read raw data from these
> > CD-Rs?
>
> You don't say what the discs are formatted as.
>

Ah, I think they're just plain old data discs - so iso9660 based, probably
with either Rockridge or Jolliet extensions. Other than that not sure -
cdrom drive isn't even recognising the discs as being valid media.

There are three types of problems you're maybe experiencing:-
> 1. scratched/dirty discs
> 2. sunlight damage
> 3. laser read/write is out of alignment
>

I think they're more likely heat damaged - they've sat in cases in a cd
rack in a box in a storage unit for way too long. They don't appear
scratched or dirty.


> 1. Clean your discs with a soft cloth in warm, soapy water. Rinse well
> and let them dry.
>

WIll try this.

2. You can 'try' setting the drive*s* to the lowest speed and blacking
> out the top of the disc with a permanent marker - then try data removal
> as per 3.
>

Hmm. How do you override the cdrom's native speed in Linux?


> 3. Find out what's on the disc and try the appropriate recovery technique:-
>

I've been using GNU ddrescue/gddresuce and for the majority of the discs
I've been working with it's been great.

The issue is the discs that don't seem to get recognised at all.


> 3f. unfortunately I can't remember the name of an application I used in
> the past which generates a list of unreadable sectors for a disc (though
> I've likely got a copy in my archives). It then uses that list to try
> and read those sectors on another drive. Rinse and repeat until it gets
> all the sectors.
> Anyone know the application??
>

Sounds a lot like GNU ddrescue to me ;)

Seriously nifty - does a quick read, generates the list of bad sectors,
then can be configured to come back a retry the bad blocks, which it uses a
bisection method on to get as much data on as possible. Or you can use a
different drive, or an identical copy (who actually does that?) to try and
read the missing data from.

Cheers,


Simon