Interactive vocal tract model

Screenshot of the interactive vocal tract model

The interactive vocal tract model is used to find a configuration that produces the correct sound of a Posture.

During the synthesis, the volume is normalized (to avoid clipping), but the gain is only reduced, never increased. To increase the gain if the sound is well below the threshold for the current configuration, the synthesis must be restarted.

Dynamic parameters can be changed without a restart. To change static parameters, the synthesis must be restarted.