Cryptography Reference
In-Depth Information
add 3 bits to each pixel. That is, it takes the red, green, and the blue
values for each pixel and changes the least significant bit of each one
independently. That means one color could quite possibly become
eight. This is quite likely to happen if that color is common in the
image because each pixel is handled differently. On average, each of
the eight slightly different colors should appear after ten to twelve
pixels of the same color are mapped.
The program
MandelSteg, developed
by Henry Hastur, hides
information in the least
significant bit of an
image of the
Mandelbrot Set. This
synthetic image is
computed to seven bits
of accuracy and then the
messageishiddeninthe
eighth. See page 319.
This means that it is impossible for the algorithm to predict the
final number of colors it needs. It might try to reduce the number
of final colors in the image to 64. Then, after the data is mixed in, it
might end up with 270 colors or 255. If it was 255, then it could save
the file. Otherwise, it would start the process again and reduce the
colors some more. The entire process is iterative. S-Tools attempts
to predict the correct number through extrapolation, but it has taken
several iterations every time I modified a file.
9.3.3 Sound Files
The simplest way to hide information in the noise is to use uncom-
pressed sound files in formats like the WAV. S-Tools can store data in
the least significant bits of aWAV fileā€”one of the standard sound for-
mats for Microsoft Windows. These files can use either 8 or 16 bits of
data to represent each instance. People with Sound Blaster cards will
have no problem generating these files from any source.
S-Tools hides one bit per either 8 or 16 bits and will also use a
random number generator to choose a random subset of bits. This
spreads the distortion throughout the sound file. The program will
display a graph representing the sound and also play it for you. After
data is hidden, the graph shows all of the changes made to the wave
form in red and leaves the unchanged parts in black. This is, in effect,
revealing where the pattern of ones and zeros in the hidden file
differed from the least significant bits of the sound file. Figure 9.9
shows a screen shot from the program.
9.4 RandomWalks and Subsets
This chapter has discussed hiding information in image or sound
files by grabbing all of the least significant bits to hold information.
There is no reason why all of them need to be used. Both Hide and
Seek and S-Tools use randomnumber generators to choose the bytes
that are actually drafted to give up their least significant bits to the
Search WWH ::




Custom Search