Re: [ARSCLIST] CD writing speed

[Don Cox <doncox@xxxxxxxxxxxxxx>]

On 22/01/06, seva wrote:
> At 1:01 PM -0500 1/22/06, Charles Lawson wrote:
>> The
>> key is what type they are and whether they can be transparently
>> correctable.
> 
> 
> audio CDs have many errors and are not corrected, per the protocol. 
> they *may* be corrected by the player (hence the better players have 
> less trouble, less audible problems).
> 
> CD-ROM for data also has errors, but it has much more robust error 
> correction, and if the disk can be read, it is bit for bit accurate, 
> otherwise software could not be delivered nor stored on optical 
> media.  if it can't be corrected, your op system will say 'disk can 
> not be read'.
> 
> this is why audio data should be stored (archived) as data files on 
> CD-ROM, not CD-DA (audio disks).  if you can play it as an audio 
> disk, i don't count it as an archive at all.

I think it should be stored in both forms. If your data disk develops a
fault, the computer will probably refuse to deliver the whole file or
even anything at all. But assuming it is faulty, you will
get a bit-accurate copy.

An audio CD that develops a fault will generally play to some extent and
most of the audio would be retrievable, possibly with some distortion
from interpolation. For oral history research, that could be enough. 

Regards
-- 
Don Cox
doncox@xxxxxxxxxxxxxx