Recipe of speech compression using coiflet wavelet

Speech compression is one of the most visible applications of wavelet. Here in this paper we examined the performance of Coiflet wavelet in continues Hindi speech coding. This is the second attempt carried out among the Indian language. In this experiment we used Coiflet wavelet to decompose the signal and applied to brige-Massmart strategy based Thresholding to make insignificant wavelet coefficient to zero and finally lossless run length encoding algorithm applied to encode the remaining coefficient. Performance of this process evaluate in term of Signal to Noise Ratio (SNR), Peak Signal to Noise Ratio (PSNR), Retained Signal Energy (RSE), and Compression Ratio (CR). For this experiment we use continues speak sentence from Hindi Language. We can successfully compress and reconstructed continues Hindi sentence with perfect audibility.