Acoustic analysis of speech - basics.
Please download the zip folder containing wave files needed for the exercises.
SAMPA transcription for English is available here.
For IPA-SAMPA conversion see this document and this website.
Handout is here!
Task 1.
Run the programme Wavesurfer (you can download it from: http://sourceforge.net/projects/wavesurfer/files/latest/download).
a) open the file recording2.3 and display the waveform of the recording
b) play the whole recording and then play its selected parts
c) add the spectrogram
d) add additional panel and read in the transcription and segmentation contained in the file recording2.3.words
e) save the configuration as "my_configuration"
f) open the files recording2.8 and recording2.10 in this configuration and create word-level transcription and segmentation for both and save the annotations
Task 2.
Run the programme Praat (select the link: 32-bit edition: praat5341_win32.zip; versions for the other operating systems can be found here: http://www.fon.hum.uva.nl/praat/)
a) open the file PLtranscription1 and display the waveform and spectrogram
b) play the whole recording and then play its selected parts
c) read in the textgrid file containing the phonetic transcription and segmentation
d) in the pane "syllables" insert syllable boundaries (mark syllables with a dollar - $)
e) save annotations in a text file
f) provide annotations for files PLtranscription2 and PLtranscription3.
Task 3. Sound source analysis.
a) read in the sound file and its corresponding textgrid (from the previous exercise)
b) create Pitch object and convert it to PointProcess object
c) click on To TextGrid (uvu)
d) merge the textgrid with the previous one
e) for each sound labeled in the pho pane, mark whether its source is periodic (P), aperiodic (A) or mixed (M).
f) save changes and annotations
done :)
SAMPA transcription for English is available here.
For IPA-SAMPA conversion see this document and this website.
Handout is here!
Task 1.
Run the programme Wavesurfer (you can download it from: http://sourceforge.net/projects/wavesurfer/files/latest/download).
a) open the file recording2.3 and display the waveform of the recording
b) play the whole recording and then play its selected parts
c) add the spectrogram
d) add additional panel and read in the transcription and segmentation contained in the file recording2.3.words
e) save the configuration as "my_configuration"
f) open the files recording2.8 and recording2.10 in this configuration and create word-level transcription and segmentation for both and save the annotations
Task 2.
Run the programme Praat (select the link: 32-bit edition: praat5341_win32.zip; versions for the other operating systems can be found here: http://www.fon.hum.uva.nl/praat/)
a) open the file PLtranscription1 and display the waveform and spectrogram
b) play the whole recording and then play its selected parts
c) read in the textgrid file containing the phonetic transcription and segmentation
d) in the pane "syllables" insert syllable boundaries (mark syllables with a dollar - $)
e) save annotations in a text file
f) provide annotations for files PLtranscription2 and PLtranscription3.
Task 3. Sound source analysis.
a) read in the sound file and its corresponding textgrid (from the previous exercise)
b) create Pitch object and convert it to PointProcess object
c) click on To TextGrid (uvu)
d) merge the textgrid with the previous one
e) for each sound labeled in the pho pane, mark whether its source is periodic (P), aperiodic (A) or mixed (M).
f) save changes and annotations
done :)