Assume you have a speech audio samples with bit depth of 13 bits. It is a common setting to produce a digital representation of the full range speech signals. However, this is too expansive for many transmission systems, wire or wireless. Thus, you are going to transmit the signal at a bit depth of 8 bits and expand it back to 13 bits at the receiving end. (a) Consider the companding scheme with linear quantization method in Figure Q3(a) is used to transmit the speech signal. Determine the sample value at the receiving end (i.e. output) for a sequence of audio samples (55,-315, -1500, 3650). (b) (c) Input Compressor [7 marks] Channel Expander Output Figure Q3(a) Apply a non-linear quantization method in the companding scheme in Figure Q3(a). Determine the samples value at the receiving end for the same sequence in Question 3(a). You may use the current standard of -law encoding at 8 bit depth with u=255. [11 marks] Compare your output samples' values in Question 3(a) and 3(b). Analyse your results.