NetNews Usenet Archive 1992 #26/NN_1992_26.iso/spool/comp/sys/cbm/4431

Text File | 1992-11-08 | 1.7 KB | 36 lines

Newsgroups: comp.sys.cbm
Path: sparky!uunet!seas.smu.edu!utacfd.uta.edu!news.uta.edu!hermes.chpc.utexas.edu!news.utdallas.edu!corpgate!crchh327!crchh484!faurote
From: [e-mail] (Joe Faurote)
Subject: Re: Just an idea I had!
Message-ID: <[e-mail]>
Keywords: sampling,assembly
Sender: [e-mail] (News on crchh327)
Nntp-Posting-Host: crchh484
Reply-To: [e-mail] (Joe Faurote)
Organization: Bell-Northern Research Ltd.
References: <[e-mail]> <[e-mail]>
Distribution: comp.sys.cbm
Date: Fri, 06 Nov 1992 18:36:42 GMT
Lines: 20

Actually, this has been done before and it's not as difficult as the
last post seemed to suggest.  It's true, for good quality and in order
to reproduce the sample you have to sample at the Nyquist rate of two
times the maximum audio frequency you are expecting, but there's no
reason if you are just trying to reproduce voice that you would really
have to sample over 6 thousand times / second - telephones work on the
principle that voice will not be broadcast over 3 kHz, which works out
o.k. for the majority of users and saves money by using simple twisted
pair wires, but I digress.

The A/D converter in the cassette should be adequate for what you are
doing.  If I remember correctly, the program was written in assembly and
was very small.  You also had the option of how good of quality you wanted
which basically determined the amount of time your sample could be as
the data was stored in RAM.  I believe the article was in RUN or Compute's
Gazette, a long time ago (at least 5-6 years, I'm thinking).  If I can
come up with more specifics, I will repost.

Good-luck,
Joe Faurote