FORMANT CHARACTERISTICS OF HUMAN LAUGHTER
Rationale for the Study:
Laughter is a daily part of people’s lives and we take notice that some people have really high pitch laughs, full blown belly laughs, or a nasally laugh, or a low laugh. What we don’t realize is what goes into laughing. The shape of one’s mouth, how opened one’s mouth is, gender, or the frequency of one’s laughter. Extreme articulation during laughter production, such as a wide jaw opening, suggests that laughter can have different frequencies. This study looks at frequencies of F1 to F5 in determining differences in laughter among individuals, but specifically genders.
Methods:
A notable characteristic of laughter is that it is sometimes produced with extreme voice characteristics. For instance, fundamental frequency values up to 1245 Hz for males and 2083 Hz for females. For instance, laughter can occur in various voice qualities such as a throaty voice. A throaty voice is typically accompanied by a constriction of the lower pharynx and widely opened mouth, and increases the frequency of the first formant (F1). This study measured the fundamental frequency and the frequencies of the first five formants (F1-F5) of vocalic segments of laughter produced in a soundproof booth by actors.
They looked at sound recordings of eight professional actors (3 males and 5 females), who produced laughter appropriate for various contexts including: joy, tickle, and sneering. Next, they looked at the sound material involved. They chose to use only laughter sequences with a minimal duration of 3 seconds, which gave a good emotional expression. The stimulus set consisted of 125 laughter sequences with 10-22 sequences per speaker.
Lastly they focused on the acoustical analysis. They discovered the ceiling of the pole search range was 5000 Hz for males and 5500 Hz for females, Measurements were not taken from segments which were unvoiced, produced with closed mouth, or where spectral measurement extraction was uncertain. Laughter sequences were segmented in the time domain according to individual vocalic segments, which we defined as burst of energy of (un)voiced exhaled breath having a single peak. The boundaries of these vocalic segments were determined visually in the amplitude-time spectrum as a distinct rise of energy from background noise into a single vocal peak. For inverse filtering, the number of poles was set to 6, the lip radiation to .99, and the low cutoff frequency to 80 Hz. Inverse filtering was regarded as being successful when (1) the peak magnitude of the harmonics of the glottal flow displayed in the spectrum was monotonically decreasing.
Results:
Statistical tests revealed that in all six acoustical parameters, female speakers had higher frequency values than male speakers.
Discussion:
The analysis of estimated formant frequencies of vocalic segments of laughter revealed that some vocalic segments showed very high F1 frequencies, whereby F1 could be almost 1100 Hz for male speakers and 1500 Hz for female speakers.
Reflection:
This article was took what seems to be common sense, regarding that females would have higher frequency laughter than males. They took this common sense and looked at it in a more scientific approach to give exact figures of research. This article did not seem fully organized and put together. From what they made their thesis, they took it a lot beyond that, and made it hard to follow.