Basic framework of vocoders for speech processing

Abstract

The main objective of this paper is to develop a basic framework of a vocoder for speech processing. The basic framework of a vocoder consists of voice preprocessing, encoder, decoder, and post-processing blocks and performing multithreading among these blocks. A GUI can be prepared which will take the input from the user and set the environment as required. The voice subsystem consists of transmission path (TX chain) and receiver path (RX chain) having various signal processing blocks. These blocks are executed in sequential order to process the voice signal. This basic framework allows user to completely analyze the effect of preprocessing blocks, post-processing blocks, and vocoders on the speech signal at different points. A thread is a process which takes samples as input, calls an appropriate block for processing, and gives processed samples as output.