jbn@wdl1.UUCP (John B. Nagle) (05/12/84)
You don't get as high a density on media copied by analog means as
you can get on digitally-copied media, because there is no way to deal
with bad spots dynamically, and thus data must be recorded in very redundant
fashion. Schemes for recording data on photographic film, analog phonograph
records, and video tape have already run into this problem. The Compact
Disk people use some incredible kludges to deal with the high error rate
they face from their production process; there is electronics in the player
to fill in reasonable music sections over gaps! And this is performed
in addition to forward error correction; error correction isn't enough.
Mass-produced optical disks (the audio Compact Disk and the larger
analog video disk) are produced by photolithography, like ICs. But the size
is huge by IC standards. The yield would be terrible if they had to be
error-free, so they put up with huge error rates and apply techniques to
make the result tolerable. This problem will probably force optical data
disks to be of substantially lower effective density than density than their
audio and video counterparts.
People have been predicting the debut of optical storage for about
fifteen years now; so far the magnetics guys have always been able to stay
ahead. Just because IBM likes the idea doesn't mean it will go anywhere;
IBM has its duds too. Remember the IBM 8100 personal computer, the IBM
3101 ASCII terminal, the IBM Displaywriter, the IBM 96 column punched
card, the IBM Data Cell Drive, the IBM Hypertape Cartridge, and the IBM
Dark Trace Terminal.