The poster contains descriptions of several procedures for analyzing vocal aspects of communicative interaction. The task of these procedures is to find deviations from individual phonetic and prosodic standards. No attempt has been made to develop a fully automatic system, but digital audio signal processing algorithms are used to support the human perception of vocal variations (e.g. by contrast enhancement or feature extraction). Results from these procedures for vocal analysis are compared with results from other measurement and analysis procedures [1, 2, 3]. The audio data on which the vocal analyses are based are recorded in parallel and on the same time scale as are the data on which the other procedures are based.

The audio demonstrations contain examples for two types of audio processing for facilitating the perception of vocal changes: sequencers enable the listener to compare different sections of the speech signal by reducing the temporal distance between them and filters enable him to concentrate on specific phonetic and prosodic features by emphasize or even isolate these features. The corresponding algorithms are implemented in the graphical data flow language of the Laboratory Virtual Instrument Engineering Workbench (LabVIEW).

References

1. Altorfer, A.; Jossen, S.; Käsermann, M.L.; Foppa, N.; Zimmermann, H. (1998). Measurement and analysis of head movement behavior during conversation. This volume.
2. Jossen, S.; Altorfer, A.; Käsermann, M.L.; Foppa, N.; Zimmermann, H. (1998). Computer-based analysis of physiological data in natural situations. This volume.
3. Käsermann, M.L.; Altorfer, A.; Foppa, N.; Jossen, S.; Zimmermann, H. (1998). Measuring emotionalisation in everyday face-to-face communicative interaction. This volume.