National Conference on Innovative Paradigms in Engineering & Technology (NCIPET-2012) Proceedings published by International Journal of Computer Applications® (IJCA) 20 An analytical approach for wideband speech coding with the help of filter bank Ujjwal v. Ramekar Lecturer Department of electronics and telecommunication Sipna’s college of engineering and technology,Amravati -444601 ABSTRACT Interpreting the need of high quality wideband speech coding some standardization events have been conducted. There are various methods and standard for speech coding. these paper present a new technique of speech coding. here interpretation is done with the help of gamma tone filter bank. The o/p of the system is analyzed and process to get the respective train of pulses. the positioning and amplitude and gain of pulses are coded using the process of quantization. KEYWORDS Wideband speech signal gamma tone filter bank, speech coding 1. INTRODUCTION In today’s era of technology there is no coding technique which is capable of providing efficient coding speech coding the art of creating the minimally redundant representation of speech signal that can be efficiently transmitted or stored in digital media. & decoded the signal with the best possible perceptual quality in these work analysis of speech is done with the help of gamma tone filter bank. here by applying the concept of filter bank we have done the coding. Speech coding is necessary for cellular phone which has limited data rates for each user it is necessary for voice IP, audio, visualtele conferencing to reduced the bandwidth consumption The organization of paper is as follows section 2 includes the concept and techniques utilized in the proposed work. Section 3 includes parameter of speech. Section 4 includes the synthesis of speech. Section 5 conclude the paper. 2.CONCEPTANDTECHNIQUSUSEDINT HEPRPOSEDWORK 2.1 Speech coding Speech coding (or digital speech coding) is the process by which a speech signal can be temporally compressed into less bits/second and then decompressed, while preserving its most important content. • The main objectives of digital speech coding are to lower the bit rate used to represent a speech signal while maintaining an adequate level of perceptual fidelity. In addition, for some applications we need to consider the complexity (computation required to encode/decode) (e.g. telephone line or cell network) and for storage (e.g. MP3) Basic Components in a Source Coding System Fig 1:basic coding system 2.1.1Classification of coding algorithms 1. Based on how the speech production and perception models are incorporated • Open-loop codec (also called vocoders): Extract an excitation signal + a vocal tract system to later re-synthesize the speech, without any attempt to preserve the waveform shape. These more or less correspond to the Model-based codec. • Closed-loop codec: Apply the source/system model of speech within a feedback loop to iteratively reduce the difference between the original and quantized signals. Based on how the waveform of the speech signal is preserved with the process Based on the nature of the encoded parameters Fig 2:close loop algorithm 2.2Gammatone filter bank Gamma tone filter can be implemented using the FIR or IIR filter or frequency domain technique here fir can be employed to implement linear phase filter. It has length of 2N-1& obtained by convolving sampled gamma tone impulse response g(n) Of length n=100 Fig3:Impulse response