Digital Signal Processing Reference
In-Depth Information
factor and R the delay of the echo. The echo signal is generated by the sum of a delayed
version of sound with the attenuation of a and the nondelayed version.
However, a single echo generator may not be useful, so a multiple-echo generator using an
IIR filter is usually applied, as shown in Figure 6.38 .
FIGURE 6.38
A multiple-echo generator using an IIR filter.
a. Assuming the system has a sampling rate of 8,000 Hz, plot the IIR filter frequency
responses for the following cases: a ¼ 0 : 5 and R ¼ 1; a ¼ 0 : 6 and R ¼ 4; a ¼ 0 : 7 and
R ¼ 10, and characterize the frequency responses.
b.
Implement the multiple-echo generator using the following code:
y ¼ filter([1], [1 zeros(1, R-1) alpha], x)
Following that, evaluate the sound effects of the speech file (“speech.dat”) for the following
cases: a ¼ 0 : 5 and R ¼ 500 (62.5 ms); a ¼ 0 : 7 and R ¼ 1000 (125 ms); a ¼ 0 : 5,
R ¼ 2000 (250 ms); and a ¼ 0 : 5, R ¼ 4000 (500 ms).
 
Search WWH ::




Custom Search