Abstract
Sound environments in public spaces such railway stations and airports are often reported for problems in Japan. Japanese society is aging rapidly and listening difficulties during announcements at a railway stations are becoming a serious problem in Japan. For elderly in whom hearing loss progresses with age, it is difficult to comprehend announcements at major railway stations, especially in the crowded Tokyo metropolitan area. The author has studied and reported on the ability of elderly people to reproduce announcements at railway stations. However, that study did not give much consideration to the influence of the listening impression of elderly people, which appears to differ depending on changes in the speaking rate of an announcement. Therefore, this paper focused on the speaking rate of announcements made at a railway station as an acoustical factor of phonatory system in a speech transmission process.
To clarify the influence of the speaking rate of announcements on auditory impression at a railway station, a physical model of a station was built to conduct a subjective experiment with elderly people and young adults enrolled as participants. About 16-second general announcement for a railway station was used in this experiment. The background noise used was the same as that heard in the concourse of a real station in the Tokyo metropolitan area. The relationship of the speaking rate of an announcement to subjective evaluations such as “Listening difficulty”, “Noisiness”, and “Unnatural utterance” was investigated. The optimum speaking rate for announcements at railway stations was discussed based on the results.
In this experiment, the speaking rate of announcements was based on varied parameters, and subjective evaluations of announcements were made under three sound settings for back ground noise conditions (LA 65 dB, 70 dB, 75 dB). Speaking rates of announcements were set at 6 levels from 4.5 to 9.5 syllable/s, at 1 syllable/s interval. The signal-to-noise ratio condition was +10 dB. With regard to “Listening difficulty” and “Noisiness”, evaluations were made on a 5-point Likert scale (“1: Not good at all”, “2: Not very good”, “3: Somewhat good”, “4: Very good”, “5: Extremely good”) was used. To evaluate “Unnatural utterance”, a 7-point Likert scale (“-3 and 3: Feel very slow (or fast)”, “-2 and 2: Feel slow (or fast)”, “-1 and 1: Feel a little slow (or fast)”, “0: Appropriate”) was used.
In this study, speaking rate condition, background noise level conditions and whether or not the participant had hearing loss were used as factors, and a three-way analysis of variance was conducted. As a result, with regard to “Speaking rate”, it was suggested that “Listening difficulty” has the lowest evaluation value when the speaking rate was 6.5 or 7.5 syllable/s. “Listening difficulty” with a background noise of 75 dB has a higher evaluation value with background noise 65 dB and 70 dB. Furthermore, within this speaking rate range, “Unnaturalness utterance” for announcements at the railway station was not felt. These results demonstrated that the optimum speaking rate of an announcement at the railway station is in the range of 6.5 to 7.5 syllable/s.
