🚀 go-pugleaf

In <199211101157.AA24155@alchemy.cs.ruu.nl>
Piet van Oostrum <piet@cs.ruu.nl>
wonders about my hints:

> > If you are interested, I suggest you consider [trumpet triumphal]
> > The Braid [drum roll] yes, my own baby! Look it up for details on
> > how it is done (just add low bias).  
> Can you give some info on this?

Given:
    a random bit stream,
    a bit stream of data to be hidden,
    a bit stream of digitized material.

Recipe:

    Say we have a digitized sound with a 16 bit sampling range.
    We read the digitized material one word at a time (16 bit chunks).

    We read the random bit stream 4 bits at a time, in order to generate
    decimal values between 0 and 15.  This number is the order of the bit
    we will borrow within the digitizing word.  Had we had 32 bit sampling
    we would have picked our random bits 5 at a time... etc...

    As we output the bits of the digitizing word to destination, we ignore
    the true value of the selected bit and replace it with the actual value
    of our next bit of sensitive data.

    If we have a lot of digitized material and relatively little sensitive
    data, we could pull an extra number from the random bit stream such
    as 0 (we don't substitute any bit in this word) or 1 (we do).  The
    variations and the gradations are infinite.

    If we want to change the obvious white noise into a less obtrusive
    tape hiss, we generate the decimal number two or more times and pick
    the lowest value of the two or three as the borrowed bit pointer.
    The least significant bits will be diddled more often than others.
    
    Et voila...  a hissy sound track wich contains seditious material,
    which is best kept secret, like the Constitution maybe...



--
 Alain
 [ ... usual disclaimer... ]