I had posted some musings about this on usenet a couple of weeks ago as
well. Maybe there are enough of us thinking about this now to make
something happen.
Many years ago, there were early standards for data on cassette tape.
One of the earliest was the "Kansas City" standard, which as I recall
worked at 300 bps, or about 30 async bytes per second. Then Tarbell
came out with a different system, and I think it pushed the rate up to
1200 bps (post corrections if it turns out that my memories of around 25
years ago are a bit fuzzy). Tape systems moved forward, and a company
called Braemar Computer Devices moved things closer to how "big" tape
data decks worked, by selling small drives that supported true NRZ
encoding. This boosted speed and reliability even more, as going to NRZ
helped make up for the wow and flutter of tape transports.
Here we are today. Wow and flutter have faded away with the MiniDisc's
use of digital processing locked to quartz crystals. Now, the big
question is what encoding scheme will permit the highest data rate
possible after passing both ways through the lossy ATRAC compression
algorithm...
If a MiniDisc holds around 140MB of uncompressed data, the acutal amount
stored by an audio encoding scheme will be somewhat less due to losses.
The loss amount would most likely vary somewhat with different
implementationas and versions of ATRAC. A good standard would take this
into account for cross-hardware compatibility.
Let's do some "wild guessing." If there was a one third loss of
capacity through the modulator -> ATRAC encode -> MiniDisc -> ATRAC
decode -> demodulator sequence, we would get about 94MB of useable
capacity, or more that a mebagyte per minute. This would mean taking
into account using both audio channels. Switching to monaural mode
would not result in any overall capacity gain, it would only cut the
available bandwidth and data rate in half. While not a huge amount, if
the modulation scheme were truly standardized, it would be a nice way to
store data on a small and stable media format. It would also make it
easy to back up and restore files on both desktop and laptop machines
that include sound cards. Using a sound card to encode and decode sound
is a pretty popular technique with Amatuer Radio operators, as well as
radio weather FAX enthusiats.
Maybe someone out there who has developed a software based "sound card
modem" would be interested in this. Working out the modulation scheme
and data rate is the first step. Quesstions include:
1) Is a single bit two tone stream best, or is a multiple simultaneous
frequency stream best given the ATRAC scheme's operation and data rate?
2) How do we label files, or do we just use the MiniDisc's TOC, and if
so, do we try to interface the TOC data from the MiniDisc Audio recorder
somehow to a PC? If we do, it makes it less standard and more custom as
there are many different MiniDisc recorders...
3) Can we come up with "sound card modem" software that will work with
PCs and Macs? This would be nice to keep things more interchangeable.
4) How would we promote this as a standard? The chances of getting an
audio MiniDisc recorder are a lot better and easier to accomplish than
getting a MiniDisc Data drive in my opinion. Also, using audio MiniDisc
recorders lets us use inexpensive audio MiniDisc media which is
available everywhere.
In my opinion (for whatever that's worth) it would be worth the tradeoff
of only using a fraction of the MiniDisc's actual capacity to be able to
store data on inexpensive and readily available MiniDisc audio discs.
Using a MiniDisc audio recorder for this would only require a sound card
and software, unlike a "regular" external data drive which would most
likely require a SCSI adapter, or in the
-----------------------------------------------------------------
To stop getting this list send a message containing just the word
"unsubscribe" to [EMAIL PROTECTED]
