Analysis of suprasegmental features and manipulation of speech signals.
Task 1
Please extract the pitch contour (F0) of the utterance “Jenny gave Peter instructions to follow” (file recording2.10.wav). First, determine the pitch range (60-200 Hz), then display the contour and correct the errors of automatic pitch extraction. Secondly, smooth the contour, view it again and again correct the errors (if there are any).
Save the original and modified contour in text files (with .Pitch extension).
Illustrate both contours, together with corresponding spectrogram and TextGrid, in the Praat Picture window. Save the picture.
Combine the wave file (Sound) with the modified Pitch and save it.
Task 2
Please manipulate the pitch in the utterance “Jenny gave Peter instructions to follow” (file recording2.10.wav) in such a way as to obtain:
a) main stress (accent) on the word Jenny
b) main stress (accent) on the word Peter
c) interrogative mode (-> question)
Publish each modified sound and save it in a wave file.
Please extract the pitch contour (F0) of the utterance “Jenny gave Peter instructions to follow” (file recording2.10.wav). First, determine the pitch range (60-200 Hz), then display the contour and correct the errors of automatic pitch extraction. Secondly, smooth the contour, view it again and again correct the errors (if there are any).
Save the original and modified contour in text files (with .Pitch extension).
Illustrate both contours, together with corresponding spectrogram and TextGrid, in the Praat Picture window. Save the picture.
Combine the wave file (Sound) with the modified Pitch and save it.
Task 2
Please manipulate the pitch in the utterance “Jenny gave Peter instructions to follow” (file recording2.10.wav) in such a way as to obtain:
a) main stress (accent) on the word Jenny
b) main stress (accent) on the word Peter
c) interrogative mode (-> question)
Publish each modified sound and save it in a wave file.
Task 3
Create a text file named list.txt containing files from this directory.
Next, run the script get_pitch.psc in Praat and create automatically files called Pitch and containing pitch contours of the utterances from the downloaded directory. In some utterances the automatic pitch extraction will require manual correction – please introduce changes and continue the script.
When you are ready (there should be .Pitch files corresponding to each wave & textgrid file in the directory), please run the script pitch_analysis.psc. It will collect information about features in the pitch (F0) domain: F0max i F0min (plus their position in the time axis), F0mean and standard deviation of F0 mean, and save it in a file called pitch.txt
Task 4
Please run the script play.psc.
On the basis of the script create a new script called textgrid_check.psc.
The script should open each sound file from a given directory and its corresponding textgrid.
It should also enable manual verification and correction of the displayed textgrids and save introduced changes.
Task 5
Please try the function of voice and duration conversion in Praat (button CONVERT, change gender & lenghten); use files from the previous exercises (e.g. file recording2.10.wav).
Task 6
Sample recordings from the emotional speech database are available here. Our reference for the experiment is the paper from K. Scherer et al. "Vocal expression of emotions".
Create a text file named list.txt containing files from this directory.
Next, run the script get_pitch.psc in Praat and create automatically files called Pitch and containing pitch contours of the utterances from the downloaded directory. In some utterances the automatic pitch extraction will require manual correction – please introduce changes and continue the script.
When you are ready (there should be .Pitch files corresponding to each wave & textgrid file in the directory), please run the script pitch_analysis.psc. It will collect information about features in the pitch (F0) domain: F0max i F0min (plus their position in the time axis), F0mean and standard deviation of F0 mean, and save it in a file called pitch.txt
Task 4
Please run the script play.psc.
On the basis of the script create a new script called textgrid_check.psc.
The script should open each sound file from a given directory and its corresponding textgrid.
It should also enable manual verification and correction of the displayed textgrids and save introduced changes.
Task 5
Please try the function of voice and duration conversion in Praat (button CONVERT, change gender & lenghten); use files from the previous exercises (e.g. file recording2.10.wav).
Task 6
Sample recordings from the emotional speech database are available here. Our reference for the experiment is the paper from K. Scherer et al. "Vocal expression of emotions".