Now I tell about the low-rate speech coding, namely the vocoders, since vocoders are primarily used at the moment in the speech communication and the quality in not very good, we will go to the linear predictive coders. The linear predictive coders have got the two excitation state impulse for the voice speech and the random noise or the unvoiced. The vocal track could be modelled with the following transform function. This is S and then these LPC parameters are quantized and sent onto the line.

---