 One of the most powerful options within Pratt concerns the analysis of spectrographic information. Let us illustrate this on the basis of one word which I will add to the object window by means of direct recording. I will choose new, record mono sound, select 22,050 Hz mono and I will record the word kiss. Let us listen to it, kiss. And let us give it the name kiss. Save to list at close will create a sound object with the name kiss. Now here I have my new sound object kiss and I can now apply the view and edit option. So what I will do first, I will zoom into the selection, view, zoom to selection. Now here I have my recording kiss, kiss, kiss. There we are. Now let us interpret this spectrogram. On the vertical axis we find the frequency, the default is a range from 0 to 5000 Hz. That is the standard analysis range for human speech. On the horizontal axis we find the time in seconds. The default maximum is 10 seconds but that is normally not needed because we want to inspect shorter sounds. If sounds are longer, we can either change the show analysis settings or we can zoom in as I have just done now. Well then we have the third dimension in spectrograms, the degree of darkening which displays the intensity of certain frequencies as a so-called, if you wish, third dimension. Let us change the intensity settings within a spectrogram first since this influences the visibility to a large extent. What we have to do is activate the menu option spectrum, spectrogram settings. And then let us change the dynamic range, that is the decibel value to a lower value. Let us take 30 decibels instead of 50 decibels, the default. And you see what is happening is that many additional soft noises are no longer displayed and now the formant frequencies are displayed more distinctively. If you click into the spectrogram, a red horizontal line appears and you can read off the frequency value on the vertical axis to the left. So here my cursor is at 2224 Hz. More details about the formants can be retrieved with a few settings. With the menu option formant, show formants, we can create an overlay of dots indicating the formant contours. Now if we just mark our vowel in kiss, we have a very interesting option to display the formant value, namely the option formant listing. And now you see the formant values of the selected portion in a text window from where you can either copy these values into other programs or you can save them into an external file. Now you see for example the vowel starts with an F1 value of 288 Hz, F2 is almost 2,000, F3 is 2,627 and the fourth formant if you need it has a value of 3,703 Hz. Another interesting option is the option pitch. And again with a few mouse clicks you can show the pitch analysis. Now this analysis displays a blue line, the movement of pitch. And the corresponding scale is displayed on the right hand side ranging from 75 Hz to 500 Hz, the default value of pitch analysis. Now normally pitch is associated, as it is here in this part, with the fundamental frequency, that is the vibration of the vocal folds. Sometimes the lowest part could also be something like this, but here of course we know there is no fundamental frequency involved at all. So this is just an analysis which we do not want to comment any further. Now if you click into the window, again a horizontal line appears and now you can read off the pitch of each of these points. However there is also another option which is quite similar to the formant option. If you mark a certain portion again and you want to list the pitch, well here you are. The pitch F0 ranges from 213 to 122, so clearly a falling voice. Let us finally make use of the picture window, which always shows up on program start, but is often closed manually for quick examinations. The picture window allows us to extract certain portions of the sound analysis and generate external graphics. For example we can convert the spectrogram into some sort of image. Now here is the menu option, Spectrum and then Paint Visible Spectrogram. And if we now say OK, then the picture window will open and you will see the spectrogram as an image, as a graphics, which you can now save. Let us say as a Pratt picture file or you can save it in this case as a Windows meta file, so into an external format, even as an encapsulated post script file. You can do the same with the formants. If you apply the menu option formant, draw visible formant contour, well this is what you get. And again you can save this as an external file or use it elsewhere. And finally you have the same possibility with pitch, draw visible pitch contour and here you are. Now these options are important options for sound analysis. For example the pitch analysis is very important for tone languages, for the analysis of intonation, formant frequency analysis is very important for the distinction of particular vowels. Let's say in sound shifts and the spectrographic analysis in the Pratt sound editor is the basis of all these options. In addition to these possibilities there are several possibilities of fine adjustment. For example try the spectrogram settings here and let's change these values and you see immediately the resolution of the spectrogram will change from a narrow band to a white band and vice versa. Or you can apply pitch settings and many more to evoke changes in the display. These fine adjustments cannot be dealt with in this clip but they should be applied whenever they are required.