Speaking of Speech
Posted September 29, 2004
On a single time scale an oscillograph lacks sufficient resolution to display the essential characteristics of human speech. When an oscillograph shows the intelligible portions of speech, it distorts the vocal characteristics. Under these conditions insufficient resolution produces the illusion that voice is a series of pulses. At the opposite extreme, magnifying the time scale to show vocal characteristics hides the slow changes that contain the intelligible speech.

In simplest terms speech is a modulation of a voice. The modulation contains the intelligible sounds while the voice carries the tonal qualities. Oscillographs of "You Got Mail" illustrate these concepts.

When viewed on a time scale of 900 millisecond, the oscillograph shows three distinct regions. The leftmost region corresponds to the first word. This region shows an increase in amplitude, a double peak and slow decay. These changes take about one-quarter second. The following words have similar characteristics and individual differences.

The oscillograph shows the individual regions differ over intervals ranging up to tenths of a second.

Expanding the time scale by a factor of fifty reveals qualities of the voice. This level of resolution shows voice is an excellent approximation of a short term periodic signal. The period of this voice is about 4.5 millisecond and corresponds to a frequency of 220 Hertz.

Short term periodicity of voice enables a viewer to immediately recognize a transient signal on an oscillograph. The transient signal disrupts the repetitious pattern of the voice. When several transient signals are present, they have different pulse characteristics and occur at irregular intervals. These differences contrast sharply with the regularities of voice and enhance visual detection of transient pulse patterns.


Return to Problems of the Acoustic Evidence

Return to New Leads in JFK Assassination Research


Last Updated on December 5, 2008 by Herbert Blenner