Ye, Yi
(2014)
EVALUATION OF THE SIGNAL-TO-NOISE RATIO REQUIRED TO ACHIEVE THE SAME PERFORMANCE IN ENGLISH AND MANDARIN CHINESE.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
Difficulty communicating in noise is a common complaint for people with hearing loss. When communicating in noise, speakers increase the intensity level of their voice and alter the stress patterns of their speech not only to monitor their own voice but also to be heard by others. Speech that increases in intensity for the purpose of self-monitoring and being understood in noise is called Lombard speech. Few studies have assessed communication performance with Lombard speech in noise which closely reflects the real-life communication situation. In addition, the characteristics of Lombard speech may be different(among) languages with different characteristics and identifying features so the few results available for English listeners may not apply to listeners of other languages.
This study evaluated the performance of English speaking and Mandarin Chinese speaking individuals listening to English and Mandarin Chinese speech in corresponding babble noise. Speech materials were the IEEE sentences in English and translated into Mandarin Chinese while controlling for phonological, grammatical, and contextual predictability. The sentences and 4-talker babble were recorded in a conversational manner and at a Lombard speech level produced while listening to 80 dB SPL of noise. The performance of 18 native English speakers and 18 native Mandarin Chinese speakers was evaluated. The SNR-50, the signal-to-noise level required to produce 50% performance, was the same for conversational and Lombard English indicating that there is not a particular benefit in producing Lombard speech to be understood. The reason to produce Lombard speech in English is to improve the signal-to-noise ratio in order to facilitate improved communication. The results for the Mandarin Chinese listeners revealed a benefit when producing Lombard speech with the SNR-50 for Mandarin Chinese significantly different between conversational and Lombard speech. In noisy situations where increasing vocal intensity is expected, , Mandarin Chinese listeners appear to benefit from features preserved or enhanced through Lombard speech that English listeners do not access.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
10 September 2014 |
| Date Type: |
Publication |
| Defense Date: |
22 July 2014 |
| Approval Date: |
10 September 2014 |
| Submission Date: |
16 July 2014 |
| Access Restriction: |
No restriction; Release the ETD for access worldwide immediately. |
| Number of Pages: |
111 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
School of Health and Rehabilitation Sciences > Communication Science and Disorders |
| Degree: |
PhD - Doctor of Philosophy |
| Thesis Type: |
Doctoral Dissertation |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
speech perception, signal to noise ratio, babble noise, Lombard speech, English, Mandarin Chinese |
| Date Deposited: |
10 Sep 2014 19:24 |
| Last Modified: |
15 Nov 2016 14:22 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/22332 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}