Re: Bitmap Steganography freeware

On Jan 18, 12:30 am, jmorton123 <jmorton...@xxxxxxxx> wrote:
Furthermore, the normalization process only changes about half the
original
LSBs and does so at random.  Think about this:  if a bitmap image has
a LSB
bias and you randomly flip all of the LSBs, have you altered that
bias?  I'm thinking
not.

If you flip all of the LSB's then the bias is reversed.

If you XOR all of the LSB's with an unbiased random stream
(i.e. "randomly flip all of the LSBs") then the bias is removed.

If you XOR half of the LSB's with an unbiased random stream
and leave the other half intact then the bias is halved.
.