Acoustic design for speech privacy in high-speed train cabins

Abstract

Speech privacy, which is commonly referred to as overhearing unwanted speech sounds, is of great importance to improve the quality of the acoustic environment inside high-speed train cabins. Not only the train noise but also the noise generated from conversations of other passengers is a major disturbance. In order to provide an acoustic environment amenable to speech privacy, the acoustic conditions were measured in travelling train cabins to characterize the train noise and in stationary train cabins to investigate the room acoustic conditions. The frequency content of train noise was varied according to the operating speed and driving conditions such as the type of tracks and condition of the fields. Train noise levels were measured while increasing speed to calculate the speech transmission index (STI) for the quantification of speech privacy. The room acoustic conditions of stationary trains were characterized using room acoustic parameters suggested in ISO 3382-3. Single-number quantities (D2,S, Lp,A,S,4m, rD and rP) were calculated in train cabins based on the spatial decay of sound pressure level (SPL) and speech transmission index (STI). For the evaluation of speech privacy to improve the acoustic condition, a geometric acoustic simulation was used to reproduce the acoustic conditions of train cabins from in situ measurement data. The scale models were also conducted to explore the effects of changing interior surfaces on the single-number quantities. Results indicate that the seating designs and arrangements which cover passenger’s ears such as when there is no gap between the backrests of adjacent seats and seats with high backrests, show a significantly enhanced speech privacy. An absorptive ceiling was found to be the most effective type of interior surface among all finished surfaces. In addition, a regression model was suggested based on the evaluation results of design elements such as seating design, absorption coefficients of each interior surface and background noise levels, to predict the privacy distance (rP). These regression models can be used in practice to assist the design of acoustic environments promoting speech privacy in high-speed train cabins.|고속열차 객실의 성가신 소음원으로 철도차량 소음뿐만 아니라 다른 승객의 대화 소리도 고려해야 하며, 대화 음의 전달을 차단하는 스피치 프라이버시는 고속 열차 객실의 음환경 만족도를 향상시키는데 중요한 요소이다. 객실의 스피치 프라이버시를 향상시키기 위하여 고속 열차 객실의 주행중 소음 특성과 정차시 실내음향 특성을 조사하였다. 주행 중 주파수 대역별 소음 특성은 속도 및 주행 조건에 따라 주파수 특성이 변화하며, 속도별 열차 소음레벨은 스피치 프라이버시 특성을 정량화하기 위한 음성 전달 지수(STI) 계산에 사용되었다. 정차시 고속열차 객실의 음환경은 위치별 음압 레벨(SPL)과 STI의 거리 감쇠량을 바탕으로 ISO 3382-3의 실내음향 파라미터의 단일 수치 평가량(D2,S, LP,A,S,4m, rD, rP)을 측정하였다. 음환경 개선을 목적으로 하는 스피치 프라이버시 평가를 위하여, 현장 평가 결과를 바탕으로 기하학적 컴퓨터 시뮬레이션을 구축하였다. 또한 축소모형은 내부 마감 재료 변화에 대한 단일 수치 평가량의 영향을 조사하기 위해 구축되었다. 그 결과, 의자의 형상과 높이가 음성전달이 되는 귀 주변의 음성을 차단해 주기 때문에 스피치 프라이버시 개선에 가장 효과적인 부위로 밝혀 졌으며, 마감재 중에서는 천장 마감면이 스피치 프라이버시 개선에 가장 효과적인 설계 요소로 평가되었다. 또한, 프라이버시 거리 (rP)를 예측하기 위한 회귀 모형을 설계 요소인 의자 등받이 높이, 내부표면의 흡음 계수, 배경 소음 레벨 등의 음향 파라미터에 기초하여 제안하였다. 제안된 회귀 모형의 설계 적용을 위하여 철도 소음과 대화 음성 음원의 Signal-to-noise ratio에 따른 음원을 제작하여 주관적 청감평가 기반의 음향적 만족도 평가를 진행하였다. 음향적 만족도 평가 결과는 측정된 STI 및 rP의 분포와 매칭하여 스피치 프라이버시의 음환경 등급을 4단계로 제시하였다. 제안된 회귀식과 만족도 등급화 결과는 고속 열차 객실의 스피치 프라이버시 개선을 위한 음향 환경 설계에 사용하여 고속열차의 전반적 만족도를 향상 시키는데 활용될 수 있을 것으로 사료된다.; Speech privacy, which is commonly referred to as overhearing unwanted speech sounds, is of great importance to improve the quality of the acoustic environment inside high-speed train cabins. Not only the train noise but also the noise generated from conversations of other passengers is a major disturbance. In order to provide an acoustic environment amenable to speech privacy, the acoustic conditions were measured in travelling train cabins to characterize the train noise and in stationary train cabins to investigate the room acoustic conditions. The frequency content of train noise was varied according to the operating speed and driving conditions such as the type of tracks and condition of the fields. Train noise levels were measured while increasing speed to calculate the speech transmission index (STI) for the quantification of speech privacy. The room acoustic conditions of stationary trains were characterized using room acoustic parameters suggested in ISO 3382-3. Single-number quantities (D2,S, Lp,A,S,4m, rD and rP) were calculated in train cabins based on the spatial decay of sound pressure level (SPL) and speech transmission index (STI). For the evaluation of speech privacy to improve the acoustic condition, a geometric acoustic simulation was used to reproduce the acoustic conditions of train cabins from in situ measurement data. The scale models were also conducted to explore the effects of changing interior surfaces on the single-number quantities. Results indicate that the seating designs and arrangements which cover passenger’s ears such as when there is no gap between the backrests of adjacent seats and seats with high backrests, show a significantly enhanced speech privacy. An absorptive ceiling was found to be the most effective type of interior surface among all finished surfaces. In addition, a regression model was suggested based on the evaluation results of design elements such as seating design, absorption coefficients of each interior surface and background noise levels, to predict the privacy distance (rP). These regression models can be used in practice to assist the design of acoustic environments promoting speech privacy in high-speed train cabins.