This is an exercise using the Praat application, which is available for free on the internet. It will allow you to visualize some of the phenomena that we have gone over in class, but you will be using your own voice. You will submit the following by email to a special email address:
1. Two audio files, saved as .wav files, as specified below.
2. One MS Word (.doc or docx file) showing four views of the utterances in those two audio files. If you wish to send more, you have the option to do so.
Create a "Praat Exercise" folder in some convenient place on your computer hard disk. This folder may be on your Desktop or in your Documents folder. Just make sure you know where it is, so you can save files to it and retrieve files from it
Plug in your headset.
Launch Praat. Two windows will open: "Praat Objects" and "Praat Picture". You may close the "Praat Picture" window.
In the Praat objects window, under the "New" tab, select "Record mono sound". If your system can only record stereo sounds, then select "Record stereo sound" instead. This will open the "Sound Recorder" window. Select a sampling frequency of 44,100 samples per second by clicking in the corresponding button in the "Sampling Frequency" menu on the right side of the window.
How to record and save sound files:
In the Sound Recorder window, click the "Record" button on the "Sound Recorder" window. The volume meter will show activity when you speak.
To test your recording ability, say a test phrase like "Testing 1 2 3" into your microphone. Then click on "Stop". Now listen to your recording by clicking "Play".
Naming and Saving Recordings:
To give your recording a name, type the desired name in the "Name" field near the lower right corner of the "SoundRecorder" window. The name should be descriptive, such as "ah ee eye" or "beat, bead, bee". You can then click "Save to list", which will save a temporary copy of the recording to the list on the PraatObjects page. This copy of your recording will be discarded when the Praat application is closed. To permanently save your recording as a wave (-.wav) file on your computer's hard disk, select File > Write to WAV file... and navigate to your Praat Exercise folder (which you created; see above) and then click "Save" to save the file with an appropriate name.
Viewing your sound files. In the Praat Objects window, select the name of the sound file you wish to view, and then click on "Edit". A new window will appear, which will display the time-domain and/or spectrogram representations of your file, depending on the setting of Spectrum > Show Spectrum.
You can eliminate unwanted sections of the sound display by highlighting the undesired section and typing control-X (Command - X on Macs).
To see a close-up of several cycles of one vowel, just highlight a small portion of the vowel and click on "sel' (select). You may need to highlight and select several times to get down to just a few cycles.
Saving your Screen Captures
For Windows users:
In order to be able to print out your Praat representations of your files, use the "Print Screen" button on your keyboard. You'll find it above the “backspace” button on a keyboard, and it says something like “PrtSc” or "Print Scrn", which means “Print Screen.” This puts a picture of the screen (a screen capture) into the clipboard.
For Mac users:
On the Mac, you can use Apple key + Shift + 3, then click on the screen to get a picture file to appear in your desktop.
To make a document that contains the four Praat representations of your utterances, open a new document in Word (or, on a PC, WordPad). Use the "paste" function (cntrl-v) place each screen image into the Word Document. Make sure you label each of the four Praat representations in your Word file.
[ On the Mac, you can insert each of the picture files you created.]
After you have inserted all four screen images into your Word document, save it with your name, such as "Jane Doe's Praat Utterances.doc".
When you have finished making the two .wav files called for in the assignment and a Word or WordPad document with the four screen captures, you are to email them as attachments to the following address:
Make sure the subject field of your email has this text: [CSD232F10]_Praat_Exercise_Due_February 11,_2010
In the message portion of your email, state your name and secret number. You will be given your secret number in class or by email.
You are to send two .wav files and one Word (.doc) or WordPad .(.rtf) file. The Word or WordPad file will have four images on it.
Here are the two utterances:
(For each utterance, submit its .wav file.
Utterance #1. [ɑ] and [i] (Just say the vowels in the words "pa" and "key", respectively, followed by the word "eye")
Utterance #2. "beet", "bead", "bee" [Just say the three words] .
Here are the images I want in the Word or WordPad file:
1. An image of the time-domain representation of Utterance #1 (the two vowels, [ɑ] and [i] followed by the diphthong in "eye" )
2. A close-up, showing two to four cycles of the time-domain representation of the vowel [ɑ] from Utterance #1
3. A close-up, showing two to four cycles of the time-domain representation of the vowel [i] from Utterance #1
4. An image of the time-domain representation of Utterance #2 ("beet", "bead", "bee" )
Here are some things to notice. You don't have to write about these considerations in your homework, but be prepared to talk about them:
1. What is the shape of the wave form for each of the vowels? Can you see the darkened bands representing the formants (formant frequencies) for the two vowels? Can you see the way the formants form a C-shaped pattern on the spectrogram?
2. In Utterance #2, each word has the same vowel, [i]. Does the length of the [i] vowel differ from word to word? Where is it short, and where is it long?
Please feel free to contact me if you need any help.