Recovering data from CD-R?

  • Thread starter Thread starter Mike Tomlinson
  • Start date Start date
M

Mike Tomlinson

I have a data CD-R (blue dye) which was written in July, 1998. Although
undamaged, it has deteriorated to the point where it is now almost
completely unreadable - no drive I have tried will even read the ToC,
with the exception of one: if I persevere with a Plextor 12/20 PleX
SCSI unit, /sometimes/ I can get that to read the ToC.

I have tried several makes and models of CD and DVD readers/writers
without success. Have tried Clone CD, BadCopy Pro, CD-R Diagnostic
programs and "dd if=/dev/cdrom of=filename.img" in Linux, without
success.

Any suggestions for other things to try? The data isn't vitally
important, but it would be nice to be able to recover [some of] it for
the satisfaction of doing so (I'm stubborn.)

Thanks.
 
Mike Tomlinson said:
I have a data CD-R (blue dye) which was written in July, 1998. Although
undamaged, it has deteriorated to the point where it is now almost
completely unreadable - no drive I have tried will even read the ToC,
with the exception of one: if I persevere with a Plextor 12/20 PleX
SCSI unit, /sometimes/ I can get that to read the ToC.

I have tried several makes and models of CD and DVD readers/writers
without success. Have tried Clone CD, BadCopy Pro, CD-R Diagnostic
programs and "dd if=/dev/cdrom of=filename.img" in Linux, without
success.
ISOBUSTER.


Any suggestions for other things to try? The data isn't vitally
important, but it would be nice to be able to recover [some of] it for
the satisfaction of doing so (I'm stubborn.)

Thanks.
Click to expand...
 
Did you try using Data Recovery softwares ?!!!!. or else ..try using
dos utils like diskedit with DOS cdrom drivers loaded already..you may
read raw sectors with that.
 
Previously Mike Tomlinson said:
I have a data CD-R (blue dye) which was written in July, 1998. Although
undamaged, it has deteriorated to the point where it is now almost
completely unreadable - no drive I have tried will even read the ToC,
with the exception of one: if I persevere with a Plextor 12/20 PleX
SCSI unit, /sometimes/ I can get that to read the ToC.
Click to expand...
I have tried several makes and models of CD and DVD readers/writers
without success. Have tried Clone CD, BadCopy Pro, CD-R Diagnostic
programs and "dd if=/dev/cdrom of=filename.img" in Linux, without
success.
Click to expand...

Retry that with "dd_rescue". It will not abort on errors.
Although from my experience the chances are small. CD-R is a
cheap, unreliable medium that is not suited for anything
important. It is o.k. for temporary storage of things
that are also on a reliable medium.

Arno
Any suggestions for other things to try? The data isn't vitally
important, but it would be nice to be able to recover [some of] it for
the satisfaction of doing so (I'm stubborn.)
Thanks.
Click to expand...
Click to expand...
 
I have a data CD-R (blue dye) which was written in July, 1998. Although
undamaged, it has deteriorated to the point where it is now almost
completely unreadable - no drive I have tried will even read the ToC,
with the exception of one: if I persevere with a Plextor 12/20 PleX
SCSI unit, /sometimes/ I can get that to read the ToC.
Click to expand...

I've done a lot of work with deteriorating CDRs, and was surprised and
pleased to find that the Liteon LTD-163 DVD reader was able to easily
read CDRs (including blue dye Verbatims from around '98) that several
Plextor CD readers and burners and NEC DVD burners were unable to
touch.

This drive let me go back and recover a bunch of stuff from old CDRs
that I considered unreadable. You can find them on ebay used pretty
cheap (mine came in a Dell system); search on LTD163 and LTD-163.
YMMV, as always, but it was quite a discovery for me.
 
Arno Wagner said:
Retry that with "dd_rescue". It will not abort on errors.
Although from my experience the chances are small. CD-R is a
cheap, unreliable medium that is not suited for anything
important. It is o.k. for temporary storage of things
that are also on a reliable medium.
Click to expand...
Since the TOC is unreadable, how can Linux find the session? I assume
/dev/cdrom is the first or last session.
 
Previously Eric Gisin said:
Since the TOC is unreadable, how can Linux find the session? I assume
/dev/cdrom is the first or last session.
Click to expand...

dd_rescue allows you to write specific sector only when the source sector
was read successfully. That way you can mix retries together, even when
they were done with diffferent readers. As th OP states he
could get the TOC with the plextor drive. After that the file can be
mounted via the loop-driver. The attempt by the OP shows that he
intended to do this.

Still might take a lot of tries or never work. And no, Linux needs the
TOC just as any other OS. I did not claim anything else.

Arno
 
Eric Gisin said:
Since the TOC is unreadable, how can Linux find the session? I assume
/dev/cdrom is the first or last session.
Click to expand...

/dev/cdrom is a block device, which means reading from it is equivalent
to reading from the first LBA on the disk onwards.

Thanks to all for the help and suggestions. There's a lot of tools out
there; it's just finding out what they are, and I appreciate the
pointers.

Bart's SCSItool was able to determine that the session on the disk has
206,000 blocks written (about 402Mb), but was unable to read any actual
data using the sector hexdump.

ISObuster was able to determine that the disc has one session and one
track of 402Mb (correct) but was taking a very long time. I'll come
back to ISObuster if readcd (see below) doesn't work.

I have downloaded and built dd_rescue but not yet tried it.

Also tried polishing the disc with silicone polish to see if masking
scratches helped. For information, the disc is a no-name one with silver
top and blue dye. It has the serial number 25E805241232C09. It had a
paper label attached to the top surface which I removed using label
remover. The top is undamaged.

Using Linux 'cdrecord -atip' says the disk type is "Long strategy type,
cyanine, AZO or similar", the manufacturer index is 22, and the maker is
Ritek.

I'm now processing the disc with Linux readcd (part of Joerg Schilling's
cdrecord suite) and an AOpen CRW-4850 drive. This is reporting "error
on sector n corrected after x tries", and the sector number is
incrementing slowly, so it appears to be getting somewhere. I'll leave
it running and report back. The command used is "readcd dev=2,0,0
f=/tmp/wos.iso -v -noerror".

Thanks again to all for the suggestions.
 
Neil Maxwell said:
I've done a lot of work with deteriorating CDRs, and was surprised and
pleased to find that the Liteon LTD-163 DVD reader was able to easily
read CDRs (including blue dye Verbatims from around '98) that several
Plextor CD readers and burners and NEC DVD burners were unable to
touch.
Click to expand...

That's a useful data point, thanks. I had tried a LG DRD8120B DVD
drive, but this didn't want to know. When faced with an unreadable disc
(floppy or optical), the first thing I do is try it in several drives.

Other drives tried: Teac CD-R66S scsi, Plextor 12/20 PleX scsi, Pioneer
DR-U06S scsi, LG GCE-8480B atapi, Teac CD-W54E atapi, LG GCE-8320B
atapi, no-name CDM-T531A2 atapi, AOpen CRW4850 atapi.

Of the above, the last seems to be the most accommodating of this
particular disc.
 
That's a useful data point, thanks. I had tried a LG DRD8120B DVD
drive, but this didn't want to know. When faced with an unreadable disc
(floppy or optical), the first thing I do is try it in several drives.

Other drives tried: Teac CD-R66S scsi, Plextor 12/20 PleX scsi, Pioneer
DR-U06S scsi, LG GCE-8480B atapi, Teac CD-W54E atapi, LG GCE-8320B
atapi, no-name CDM-T531A2 atapi, AOpen CRW4850 atapi.

Of the above, the last seems to be the most accommodating of this
particular disc.
Click to expand...

I've got similar models to several of your drives (the Plextors and
Teac), and they're not much good for dying disks, I've found.

A discussion some time back suggested it might be the Mediatek chipset
that contributed to the success of this particular Liteon and an
LG8525B reader one poster had similar luck with. It would be
interesting to see if your unsuccessful LG reader used the same
chipset. Not much hard data on this kind of thing so far, I'm afraid.

http://groups-beta.google.com/group..._doneTitle=Back+to+Search&&d#4af4637b7037a323

Be sure to post any results you get if you try these drives.
 
Mike Tomlinson said:
/dev/cdrom is a block device, which means reading from it is equivalent
to reading from the first LBA on the disk onwards.

Thanks to all for the help and suggestions. There's a lot of tools out
there; it's just finding out what they are, and I appreciate the
pointers.

Bart's SCSItool was able to determine that the session on the disk has
206,000 blocks written (about 402Mb), but was unable to read any
actual data using the sector hexdump.
Click to expand...
ISObuster was able to determine that the disc has one session and one
track of 402Mb (correct) but was taking a very long time. I'll come
back to ISObuster if readcd (see below) doesn't work.
Click to expand...

I have found that successfully reading bad sectors is speed dependent
(one sector may read best at 4-speed while an other reads best at 8-
speed) and that there are 2 (or 3) speeds that work best. Also, chan-
ging speeds during retries sometimes gets a badly readable sector.
You can change speed using Joerg Fiebelkorn's CD Throttle.
Also, a hot drive reads worse and the more retries are needed and the
more the drive is slowing down and speeding up again during retries,
it hots up faster. I get the best results when the drive is cold.
I have also found that what one drive can't read, another might, and
vice versa. When a drive that reads the TOC succesfully but can't
read a specific file (using ISOBuster) it may be possible to read it
on another drive (even though that drive won't read that TOC), by
using Extract From-To and the file info (begin sector, end sector)
provided by the TOC on that other drive.
 
Previously Mike Tomlinson said:
[...]

I'm now processing the disc with Linux readcd (part of Joerg Schilling's
cdrecord suite) and an AOpen CRW-4850 drive. This is reporting "error
on sector n corrected after x tries", and the sector number is
incrementing slowly, so it appears to be getting somewhere. I'll leave
it running and report back. The command used is "readcd dev=2,0,0
f=/tmp/wos.iso -v -noerror".
Click to expand...

This is likely easier than using dd_rescue. And likely better
adjusted to 2k sectors.

Arno
 
Back
Top