Since the last decade, data transmission has kept developing faster than ever. With the support of online systems and mobile devices, the development of data transmission is also followed by up-to-date or newer applications. However, this development of data transmission is not without risk. Many users are still negligent in maintaining their data privacy, leading to disclosing their private data. Therefore, data security is needed to ensure the safety of the private data being transmitted. Data hiding is one of the methods introduced in data security. In this research, private data are embedded inside an audio cover to be retrieved safely from the receiver. Nevertheless, the capacity of private data and the similarity level of embedded audio files become a significant problem. We focus on extending the capacity of private data by utilizing an interpolation-based approach and optimizing the similarity level by dispersing the embedding space and shifting the embedded value to be as close as possible to the original audio file. The experimental results show that this approach can obtain a similarity level of 103.5 dB when embedding 100 kb of data and fall to approximately 88.15 dB when embedding 300 kb of data.