Each audio file that does *not* start with "LF-" contains the vowel sounds synthesized with one particular resonator and the geometric glottis model for all five pressure values in a sequence (0.8, 1.0, 1.2, 1.4, 1.6 kPa), where the duration for each pressure value is about 1 s.
The audio files starting with "LF-" contain the vowel sounds synthesized with a closed glottis and the Liljencrants-Fant glottal flow pulses.
