The effect of reverberant sound level on the intelligibility of spoken Malay words
Reverberant sound is known to degrade Speech Intelligibility (SI). For instance, it has been found that amplitude of English speech signal, syllable continuum from “sir” to “stir”, is affected in reverberant condition. However, there are currently no studies on the effect of reverberation on spoken...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2015
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/53808/1/MuhammadAasimAsyafiieMFKE2015.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Reverberant sound is known to degrade Speech Intelligibility (SI). For instance, it has been found that amplitude of English speech signal, syllable continuum from “sir” to “stir”, is affected in reverberant condition. However, there are currently no studies on the effect of reverberation on spoken Malay words. The purpose of this research is to investigate the effect of reverberant sound on spoken Malay words. The project started with the development of Malay word list. The list consists of 5924 distinct Malay words and was based on the texts from 52 Friday sermon transcripts that were spoken in Kuala Lumpur mosques. The Malay words spoken in mosques were used because SI in many mosques suffers from reverberant sound. From this, two sets of phonetically balanced word lists were developed with each contain 50 words. These words were then recorded in an audiometry room with the help of two trained speakers, a male and a female. The recorded words were then played back in seven different room samples with different reverberant sound levels. Reverberation time was used as level indicator (in seconds) of reverberant sound. The effect of each room sample on clean recorded words was analysed in terms of fundamental frequency (F0), first and second formant frequency (F1 and F2), and spectral tilt. The effect of reverberant sound on F0 for female speaker was more profound and statistically significant. The F1 of both speakers were not affected by reverberant sound. However, only F2 of female speaker was affected by reverberant sound. The value of spectral tilt shows that vowel /a/ is the most susceptible to reverberant sound. In conclusion F0, F2, and spectral tilt are relevant parameters, and have been able to demonstrate the effect of reverberant sound on spoken Malay words. |
|---|