Sound Tutorial
Sound Tutorial
Exercise 1
Time
2 °) If the sampling is at 9 KHz. How many times per second is sound analyzed?
4 °) A discretization of the amplitude encoded on 1 byte, over how many levels will the sample be
encoded?
5 °) The low amplitude signals will be more disturbed by the quantization noise (unfavorable S / N
ratio). It would be interesting to obtain, for this type of signal, an S / N ratio as favorable as that of
signals of greater amplitude. What solution do you suggest? Schematize?
6 °) Knowing that the minimum frequency of this signal is 3 Hz, and the maximum frequency is 22
KHz, what would be the minimum sampling frequency to preserve all the information of the original
signal? In this sampling case, calculate the size of the compression of this signal in stereo for a time
of 60s with an amplitude encoded on 2 bytes?
7 °) What are the parameters that determine the quality of the digitization of a sound signal?
Exercise 2
We have seen that in the classic method called pulse code modulation (PCM), each sample is
transformed into information encoded on a number of bits. If it were possible to reduce the number
of bits needed to quantize each sample without altering the sampling step, the size of the files could
be reduced. This is what differential modulations or DPCM (Differential Pulse Code Modulation)
propose which, instead of coding the absolute value of each sample, is content to encode the
difference observed between two samples.
1) The first technique called "Delta modulation" makes it possible to substitute for the value defining
the improvement in the improvement of the only sign (positive or negative) of the variation of the
signal compared to its previous value.
b) In some cases, this coding has faults. What defect is it? and in which case is it found?
c) Which technique allows to limit the warning of the fault of this coding in this case?
Schematize on an example.
amplitude on 2 bytes (stereo foomat). What is the gain in size expressed as a percentage of the
digitization of this signal:
b) In DPCM compared to a PCM? (Knowing that in addition to the variation sign, the expression of
the difference is coded on 4 bits)
3 °) What does ADPCM mean? What is the difference between this technique and the DPCM?