Time and Frequency Domain Characteristics

1. Record your speech data for all the sound units listed in theory using 16 kHz sampling frequency and 16 bits/sample resolution.

 

2. Plot (a) speech data of each sound unit, (b) Plot 30 ms of it to observe expanded version and (c) log magnitude spectrum for the 30 ms segment.

 

3. Note down your observations.

Speech is an example of random signal

1. Generate two sinusoid signals each of 50 Hz and plot one on the other in different colors. Plot its log-magnitude spectra (Example-1).

 

2. Record speech data for vowel /A/, thrice by yourself (Example-2).

 

3. Collect speech data for vowel /A/ from three different friends.

 

4. Plot example-1 and example-2 of your recording, one over other, in different color. Note down your observations.

 

5. Repeat the above procedure for corresponding log magnitude spectra.

 

6. Plot example-2 of your recording and speech from friend-1 and also plot the corresponding log magnitude spectra.

 

7. Repeat the same procedure for example-2 of your recording and speech from friend-2.

 

8. Repeat the same procedure for example-2 of your recording and speech from friend-3.

 

9. Note down your observations for all cases.