Fwd: unknown data tape - test-drb@ccmp.vtda.org

11 Dec 2005

...

Subject: Fwd: unknown data tape
   From: John Foust <jfoust at threedee.com>
   Date: Sun, 11 Dec 2005 11:01:07 -0600
     To: <cctalk at classiccmp.org>

Speaking of audio cassette storage, I received this inquiry
to my web.  Click the link to hear the data he has...  What's
odd to my ear is that the data seems to be packeted between
carrier.  Might this be someone's custom data, as opposed to
ordinary program load/save? 
No and yes.  Most of the cassette schemes were either FM or FSK
and both required a carrier to establish sync before the data.
there were a quite a few modulation schemes but they can be
described as either frequency shift keyed or phase shift keyed
and in some ways similar.  The important factors were cassette
recorder players had poor speed control (both short and long term)
high background noise (only soso signal to noise) and lousy
bandwidth (rolls off below 300hz and the upper liimit maybe 8-11khz).
So the recording schemes had to compensate for that.

Allison

...

As for why MP3s work fine and .WAV is unnecessary, MP3's
compression divides the signal into ~32 frequency bands, then
encodes the amplitude of those over time.  The typical 
two-frequency sine modulation used by cassette data recorders
probably fits quite well into those two bins, and MP3 would
recreate the simple sine waves quite well.  

After all, some of the hardware/software decoders within the 
classic PCs (as well as their emulators) may have only used 
the timings of zero-crossings, so they aren't very sensitive 
to perfect reproduction of the waves.

- John

>Date: Sat, 10 Dec 2005 16:11:34 +0100
>From: Paul Geisler <paul.geisler at web.de>
>To: jfoust at threedee.com
>Subject: unknown data tape
>
>
>hello.
>
>i found some old tapes i know nothing about ecept by title it may contain interesting
information, and came from merely professional use.
>
>the tapes are recorded on standard compact casette but with some dedication to data
recording for example switchable write protection etc.
>the label reads "T 300 Certified Data Casette".
>
>i capture the contents by stereo tape recorder, found both channels nearly equal so
guess this mono, mixed them. the tapes are two-sided like normal audio ones.
>
>a strange thing is, that the recording is in small "packets", with empty
tape inbetween. i first guessed it is only "empty formatted" tape, but the data
packets seem to contain a varying length of information. still strange to waste more than
a half of material this way?  also strange the packets are ending with a simple tone,
instead are headed with, so maybe this is used backward at original drives.
>
>here is a seven packet sample some minutes after beginning of one tape; all tapes
seemed to have this same "format" from start to some end location which is not
the real tape end:
>http://hirnsohle.de/test/datenSnipplet.wav  (16bit mono microsoft-wav 624kb)
>
>i would be very happy if you had any idea how to look at the contents (if any)..
i'm currently searching for some sofware "tape modem". i know such things
for C64 datasette tapes, so maybe you know some universal program, where different
emulations or flexible parameters can be used to find a matching demodulation/decoding?
>
>thanks very much & best regards
>Paul Geisler 