Institute for Natural Language Processing

Experimental Phonetics

Phonetics Lab at IMS

• IMS research tools
• Standard hardware and software
• Acoustics lab

IMS research toolstop

• German text-to-speech synthesis system
  (M. Barbisch, B. Möbius, G. Möhler, B. Säuberlich, A. Schweitzer)
• Speech enhancement with subspaces
  (W. Wokurek)
• Wigner distribution in the ESPS/Waves environment
  (W. Wokurek)
• Diseg: Discourse segmentation and XML-style annotation of discourse relations (written in Perl/TK)
  (J. Mayer)
• Aligner for German
  (S. Rapp)
• Automatic recognition of pitch accents
  (S. Rapp)
• PSOLA-based resynthesis for ESPS/Waves
  modelling of F0, duration and intensity
  (G. Möhler)
• Automatic generation of intonation contours
  (G. Möhler)
• Automatic classification of voice quality using the EGG signal (online tutorial)
  (K. Marasek)
• WWW-based speech perception tool
  (G. Möhler, M. Jessen)

Standard hardware and softwaretop

• Linux and Windows PCs
• ESPS/Waves speech analysis package
• Praat speech analysis package
• Wavesurfer speech analysis package
• KPE formant synthesis environment
• PCLX and Electrolaryngograph
• MatLab (mathematical modelling software)
• HTK - Hidden Markov Modeling Toolkit (automatic speech recognition)
• Laryngograph - Electroglottography (EGG)
• SPM - Statistical Parametric Mapping (fMRI processing)

Acoustics labtop

The acoustics lab is equipped with an anechoic chamber, a Tascam DA-45HR DAT recorder, a Yamaha 03D digital mixer, a linux computer with an S/PDIF sound card (Midiman/M-Audio Delta DIO 2496) which allows for digital recordings with a resolution of up to 24 bit and sampling rates up to 96 kHz, a NAD 302 analog stereo amplifier, several microphones, headsets and headphones.

top