Path: utzoo!attcan!utgpu!news-server.csri.toronto.edu!mailrus!cs.utexas.edu!samsung!uakari.primate.wisc.edu!nic.MR.NET!uc!shamash!mpe
From: mpe@shamash.cdc.com (2375)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Speaker Control/Music
Message-ID: <21857@shamash.cdc.com>
Date: 27 May 90 17:34:26 GMT
References: <1990May23.162547.26538@agate.berkeley.edu> <948@ashton.UUCP>
Reply-To: mpe@shamash.cdc.com
Distribution: usa
Organization: Control Data Corporation, Arden Hills, MN
Lines: 27

You can seemingly control the speaker volume by alternating the interval
between tones.  An example of a routine is:

      for i:=1 to count do
      begin
        sound(1000);
        nosound;
        delay(volume);
      end;

In this routine, the tone of the sound is the value 1000 and the volume
is somewhat controlled by the delay interval between sounds.  Please note,
the method outlined here only approximates the function you want and the
REAL output tone will be a rather complex combination of the actual sound 
tone and the delay time.  Find someone who knows something about FFT's 
to explain what I mean...

As for a TSR routine to implement this, I leave that to the reader as
an exercise.

The fastest method that I know to implement a solution to control the 
volume of a PC for the purpose of not waking everyone up during midnight
programming sessions is to splice-in a variable resistor is series with
the speaker.

cat *flame*  > /dev/null
.