>>34
But encrypted data within steganography takes more analysis to detect.
Well, it would be pretty silly to use steganography not in conjunction with an encryption scheme. Now I think it should be possible to steganographically hide a public key encrypted message as long as there are no visible unencrypted headers (and that the public key structures have uniform distribution in all their bits, which I don't think holds for all ciphers although it might be possible to adapt for some). But yeah, for decryption, one would have to try every steg encoding scheme against every private key in storage.