Study of audio AAC encoding and developing an optimized Encoder

Abstract

For high-quality, multichannel, and digital audio compression, MPEG advanced audio coding (AAC) and its successors become the most important pillars of the Moving Picture Experts Group (MPEG) family. MPEG AAC offers a wide range of applications, from low bitrate Internet audio to multichannel broadcasting services, thanks to its numerous working modes and configurations. In this thesis, we are going to study the different blocks of an AAC encoder based on the standard ISO/IEC 13818-7 such as the psychoacoustic model, Gain control, Transform block, Spectral processing, Quantization and Entropy coding. Then From the FFMPEG Codec which is based on the standard we will propose some Optimizations regarding the Encoder, like Removal of block switching, Fast MDCT, and Optimized TNS. And then test their implementation on C Code to compare the compilation speed between the original Codec and the optimized one.