Data hiding schemes using complete complementary codes have been developed for image and audio signals. It has been shown that a high payload and a low bit-error-rate can be achieved by such data hiding strategies because of the ideal auto- and cross-correlation properties of complete complementary codes. The authors have proposed a disaster prevention broadcasting system based on the data hiding scheme using complete complementary codes. In the proposed system, detailed information on disasters such as earthquakes and tsunamis are embedded into audio signals like siren sounds. The audio signals in sirens can be processed easily by smartphones and mobile tablets for faster dissemination of disaster information in the community aiding an effective and safe evacuation. This paper shows the results of the practical experiments using real siren sounds from a loud speaker. It is shown that the embedded information can be extracted successfully by choosing appropriate parameters. In this work, the authors have also developed an Android application operated on the smartphones and mobile tablets to extract the embedded messages from the stego siren sounds and evaluated its performances.