Speech Corpus
Acoustic database management system
| First version: | Last version: 2000 |
| Application: | ||||
| The system is
used by the team and its partners, in their R&D
projects. Thus, the study of the sound material managed by the system has made it possible to acquire new data about the allophonic realisation of the Russian phonemes in different phrasal positions. |
||||
| Description: | ||||
| The system is designed for the storage of
the phonetically representative sound material (the text). The storage and the description unit is the syllable. The description includes a graphical representation of the syllable in the text, an ideal phonemic (broad) and phonetic (narrow) transcription, a real transcription, statistical characteristics, peculiarities of the syllable realization connected with the prosodic position and the name of the sound file in which the syllable is stored. The user can listen to the whole text or any fragment of the text, though not shorter than the syllable. He can find all the realisations of the syllable in the text, using its graphical representation or transcription. Using graphical representation of the selected fragment, he can get its transcription and sound output in the realisation of the indicated speaker(s). To provide the possibility for the audio output of the selected fragment and obtaining its transcription the following preliminary procedures (see below) were performed. The text in the orthographic form independent of the speaker’s realization was used as the invariant. The audio output of the selected fragment comes as the result of the formation of the sound stream from the sound files in the order they appear in the text. The transcription of the fragment is formed in a similar way: transcriptions of the syllables which form the fragment (in the order they appear in the orthographic text ) are taken from the corresponding fields in the database. The audio output of the fragment can be saved into a sound file for further analysis of the acoustic characteristics. |
||||
| Type of processing: | ||||
| Preliminary text segmentation into open syllables; each syllable is presented as a separate file; the file name contains the ordinal number of the syllable in the text. | ||||
| Hardware/software requirements: | ||||
| IBM PC or compatible, Windows. The system
implementation is based on MS Access 2.0. The sound material is recorded in 16 bit Raw PCM format at 20 KHz sample rate. |
||||
| Distribution: | Internal use | |||
| Developer team: | Contact person: | |||
| Department for phonetics and methods of teaching foreign languages, SPbGU | Pavel A. Skrelin Phone: (+7 812) 328 95 65 E-mail: paul@PS1089.spb.edu, paul@phonet.lang.pu.ru |
|||