Steganograhy is the technique of hiding messages in a signal such that an unindended receiver of the signal doesn't come to know about the presence of the hidden message. Image steganography has been studied deeply by people. We study an application of concepts from image steganography to audio steganography.

The audio signal is converted to an image by taking the high frequency coefficients of the wavelet transform of the audio. A standard image steganography technique is applied to this image signal to hide the message in it. The image with the hidden message is then converted back to audio signal. A similar method is used for extracting the the message from the audio. This method gives us multiple parameters that could be used as a key during message extraction.

Long messages could be hidden into the audio without any perceptible change in it. We analysed the message reconstruction error as a function of the message length and concluded that even long messages could be securely transmitted and faithfully reconstructed.