Three Audio CODECs Using the LSM Interpolation and Comparison with PCM
Sameh Abdelwahab Nasr Eisa *
Department of Mathematics, New Mexico Institute of Mining and Technology, 801 Leroy Place, PO #3696, Socorro, NM, 87801, USA.
*Author to whom correspondence should be addressed.
Abstract
In this paper we take advantage of the known interpolation Least Square Method (LSM) to construct audio coding/decoding (CODEC) algorithms. The purpose of this algorithm is to compress audio data, maintain high quality audio, and enable sending/storing audio as serial data through digital transmission systems. Our proposed algorithms can be an efficient replacement for the quantization process used in many CODECs and modulations like PCM. We clarify our research by explaining the reasons, the assumptions, and the experiments' results for each step individually. We applied the algorithms to 20 audio files and introduced three algorithms that approach the most efficient compression ratio in addition to best signal to noise ratio. We showed Pseudo Codes, Flowcharts, and complete results of some experiments, and also a comparison with PCM used in telephony system.
Keywords: Audio, Speech, Processing, CODEC, Interpolation, Polynomials, Modeling