>>128
We are responding to the same point. But it is perfectly possible for stenography to use a key. For instance, a scattered sequence of bits that weaves through a large collection of home photos some of which were deleted and partially overwritten with audio recordings. If the amount of encoded information is sufficiently low, how will you see the signal through the noise? Suppose they guess the exact algorithm that you used, any associated keys, and they extract the data. How will they know they have obtained the encrypted data instead of bits of random noise?
>>125
steganography is related in the it satisfies a related purpose.