> I have some large data sets to store. It would be nice to write tham in
> audio mode with larger sectors than the standard ISO9660 format. Since
> these will be written and read with an application, the format can be
> anything I desire, as long as it's physically portable. I would love to
> put the length in the first four bytes and then just read data from CD
> to a disk file (and obviously write as well). It would reduce the media
> count and improve the transfer rates as well.
>
> Has anyone done anything with this? And is there any reason why it would
> not work. Yes, obviously I have to pad to the correct length before
> writing, to fit the larger sectors.
It can certainly be done, but I wouldn't do it for any data where I had
the least concern about its correctness.
Two issues:
- You only get the two-level C1/C2 Reed-Solomon error correction, rathern
that the three levels available in Mode 1. If a scratch or dust blemish
on the media overwhelms the C1/C2 error coding (which typically happens
several times per CD even when they're in good shape) the low-level
electronics will automatically "conceal" the error, via a sample-and-
hold or via interpolation of some sort. The application will receive
*no* notice that this has occurred. You'll end up with bad data points
in your dataset. Unless you wrap some sort of CRC or checksumming
around the blocks, you'll never know that the data was bad.
- Red Book audio sectors cannot be addressed precisely - this is the
"jitter when ripping" problem. Unless you've got a Plextor-grade
CD-ROM drive with its own built-in jitter compensator, you'll find
it difficult to read through the same disc twice and get the same
set of data. You'd have to wrap your own block headers, counts,
etc. around the data, add sync headers, and so forth in order to
be able to get the data back reliably.
Unless you're willing to hand-craft this sort of additional addressing
and error detection (or correction) yourself, I think you'll find that
using raw Red Book mode just isn't going to cut it for data. That's
the whole reason Mode 1 was invented in the first place.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
