The invention provides a novel way to generate audio data files with embedded messages. In particular, the system generates a plurality of permutations of audio input data provided by a user. The system then modifies the permutations of the audio input in various manners as described herein. As a result, the system is able to generate an audio file which includes a much greater number of embedded messages than audio files generated by traditional systems.