Analyses of speech recordings from an acoustic head and torso simulator with and without face coverings using both spectrogram and transcription tools
| Autor: | Laura Ruhala, Richard J. Ruhala, Guillermo Brito, Britt Cumbee |
|---|---|
| Rok vydání: | 2023 |
| Předmět: | |
| Zdroj: | The Journal of the Acoustical Society of America. 153:A373-A373 |
| ISSN: | 1520-8524 0001-4966 |
| DOI: | 10.1121/10.0019218 |
| Popis: | The COVID-19 pandemic caused the need for a wide assortment of face coverings to be adopted all over the globe. While reducing the transmission of the virus, the use of such face covering also reduced speech sound levels and intelligibly. Sound levels and recordings were made at a 2-mdistance from an acoustic head and torso simulator (HATS) with and without masks, including a KN95 mask, a surgical style mask, and a transparent plastic face shield. In this study, WAV recordings were made using short phases of pre-recorded American English female and male talkers from the ITU-T P.50 database. Adobe Audition software was used to improve the playback quality of the speech recordings, trim the modified files into 10 sof speech, ensure that they all started with the same spoken words, and create spectrograms to identify the mask-related attenuations or amplifications. Adobe Premier Pro software was used for the transcription of these files. Documenting which words were transcribed correctly or incorrectly provided further indication of the effects of various face coverings on speech intelligibility. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|