Effect of Reverberation Time on Vocal Fatigue Pasquale Bottalico, Simone Graetzer, Eric J. Hunter Voice Biomechanics and Acoustics Laboratory Department of Communicative Sciences and Disorders Michigan State University pb@msu.edu, sgraetz@msu.edu, ejhunter@msu.edu Abstract Vocal effort is a physiological magnitude that accounts for the changes in voice production that occur as vocal loading, which is the stress inflicted on the vocal folds when speaking for long periods, increases. It has been quantified in terms of Sound Pressure Level (SPL). In previous research, it has been shown that prolonged vocal effort can lead to vocal fatigue. An experiment was conducted to measure the effect of reverberation time on vocal fatigue, by means of an evaluation of variation in vocal effort over time. Twenty subjects were recorded while reading a text in anechoic, semi-reverberant and reverberant rooms in the presence of babble noise. Within- subject variation in SPL was measured per task. It was found that SPL tended to increase less over time in the semi- reverberant environment than in the anechoic or reverberant environments. This finding suggests that subjects experienced less short-term vocal fatigue in the semi-reverberant environment. Index Terms: vocal effort, vocal fatigue, reverberation time, room acoustics 1. Introduction In this study, the effects of room acoustics, voice style (corresponding to normal and raised levels) and vocal fatigue on vocal effort (quantified as Sound Pressure Level or SPL) and self-reported vocal effort, control, comfort and clarity, are considered. The interaction between a person, a room and an activity leads to different sensations relating to voice production. This interaction determines the acoustic comfort, which contributes to human well-being. It also determines vocal comfort, which is related to all aspects that reduce vocal effort [1]. Vocal comfort appears to decrease with the speaker’s perceived fatigue and the sensation of needing to increase the voice level [2]. The maximization of intelligibility, clarity, vocal comfort and control, and the minimization of vocal fatigue and effort, should be the priority of any professional talker. This is particularly important when the person is at elevated risk of vocal injury, such as in the case of teachers [3]. In classrooms, noise levels are typically high and acoustical conditions are not optimized for the talker but for the listener. Vocal control can be defined as the capacity to self- regulate vocal behavior, e.g., Sound Pressure Level (SPL), and resonance. The sensation of control relates to the ability to adjust the voice consciously. In a communication environment, in general, speakers try to control their voice production in order to increase speech intelligibility. For example, while considering a communication partner with hearing limitations, a talker (deliberately or inadvertently) uses “clear speech”, which has been characterized by a slower speech rate, a wider range of fundamental frequency (fo), and a higher temporal modulation index than conversational speech. Likewise, when talking in a noisy environment, people tend to raise the level of their voice (vocal effort) in order to maintain understandable communication [4]. Vocal effort is the exertion of the speaker as quantified by the A-weighted SPL (dB) at 1m distance from the mouth [5]. It is a physiological entity that accounts for changes in voice production when loading increases [6]. It relates to various factors such as the type of interlocutor, the speaker-listener distance, the background noise level, and other acoustic characteristics of the room [7, 8], linguistic factors such as vowel quality [8], and the speaker’s level of fatigue 10, 11]. The effect of reverberation time and speaker-listener distance on voice power level was investigated by Pelegrín- García et al. [7]. Thirteen male talkers were recorded in four different environments: an anechoic chamber, a lecture hall, a corridor, and a reverberant room with average reverberation time (T30, 0.5–1 kHz) of 0.04 s, 1.88 s, 2.34 s and 5.38 s, respectively. The voice power level was found to depend almost linearly on the logarithm of the distance (with slopes between 1.3 and 2.2 dB per doubling distance) and changed significantly among rooms (intercepts between 54.8 and 56.8 dB). With the exception of the reverberant room, voice power level decreased as reverberation time increased. Reverberation time was found to influence vocal intensity in continuous speech by Black [8]. He reported an analysis of SPL measured in the context of read speech produced by 23 males in 8 rooms differing in shape (rectangular and drum), for size (4.2 m 3 and 45.3 m 3 ), and reverberation time (0.2-0.3 s and 0.8-1.0 s). Greater vocal intensity was found in the dead (less reverberant) room than in the live rooms. Moreover, the speakers’ intensity was found to be lower during the reading of successive phrases (comparing the first 3 and last 3 phrases of a total of 12 phrases) in live rooms than in dead rooms. Voice power has been found to be related to room size and the magnitude of amplification by the room of the talker’s voice at his/her ears, compared to anechoic conditions (termed “room gain” and “voice support”). Brunskog et al. [12] investigated objectively measurable parameters of the rooms related to any increase of the voice sound power produced by speakers and to the speakers’ subjective judgments about six different rooms with different sizes, reverberation times and other physical attributes. They found that the speaker’s voice power level when teaching is positively correlated with room size and inversely correlated with room gain. Speech Prosody 2016 31 May - 3 Jun 2106, Boston, USA 494 doi: 10.21437/SpeechProsody.2016-101