Sound-masking technique for combined noise exposure in open public spaces

A sound-masking technique was applied to minimize the annoyance in open public spaces exposed to combined noise sources. The noises were recorded binaurally through artificial head microphones from the building construction sites with nearby roads. Sound quality (SQ) parameters as well as spectral and temporal characteristics of the noises were investigated. For the auditory tests to evaluate the presence of the sound-masking system, sound maskers were manipulated and presented in the open public spaces. Psychoacoustical factors affecting subjective responses to the combined noise exposure were investigated and the effectiveness of the soundmasking system was examined. INTRODUCTION Demands on favorable acoustic environment in open public spaces have been increasing with the increase of outdoor activities. Overall sound pressure level in the area has been thought as the most important aspect affecting the amenity of the environment. However, most of the open areas are necessarily accompanied with heavy flux of transportation and people so that excessive noise threatens soundscape of the environments. Many studies investigated the acoustical environment of exterior spaces in order to define the relationships between sound characteristics and subjective responses (Anderson et al. 1983; Skanberg & Ohrstrom 2002). Although classifying the characteristics of soundscape in open public space is important, active treatment on the sound environment is often needed to improve quality of the acoustical amenity of the areas. As an actual method to control the soundscape, several nature sounds were introduced with speaker systems to the open public spaces (Jang et al. 2003; Lee et al. 2005; Jeon et al. 2007). As most of the studies applied nature sounds without manipulation, hence masking effects on the environmental noise have been deficient. In the present study, the characteristics of combined exterior noises from road traffic and construction site were investigated. The preferences on the soundscape were sought before and after the application of the sound-masking system. Sound quality attributes of the combined noise were determined and sound maskers (to enhance the acoustical amenity) were produced. MEASUREMENTS Sounds from building construction and nearby roads were recorded in eight open public areas which were chosen according to the volume of road traffic and the distance from construction sites (see Table 1). Measured noise level ranged from 55.8 to 78.0 dBA during 10 minutes. Sites B and F were chosen and, as shown in Figure 1, HATS (Head and Torso simulator) was located at 33 m, 15 m from the roads in Sites B and F, respectively. Therefore, in Site F, HATS was located at 50 m apart from the construction site. The height of microphone was set at 1.5 m from the ground. The Community: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT noises were recorded as ‘wav’ format using Adobe Audition software, through AD/DA converter. Table 1: Categorization of open public spaces for noise measurements (a) Site B (b) Site F Figure 1: Binaural recordings in Sites B and F As shown in Figure 2, frequency characteristics of the noise for both sites were compared. Both site B and F showed relatively high levels at low frequencies due to road traffic, but harmonic components appeared in Site F because of the noises from construction machinery. Site B showed lower levels at low and mid-low frequency ranges, and the level difference between Sites B and F was increased at mid-low frequency range. Figure 2: Frequency characteristics of recorded noise Place Road traffic volume Number of construction sites A Low 0 B High 0 C Low 1 D High 1 E Low 1 F High 1 G Low 2 H High 2 Community: 9th International Congress on Noise as a Public Health Problem (ICBEN) 2008, Foxwoods, CT EXPERIMENTS AND RESULTS In the laboratory and actual sound fields, three experiments were undertaken to evaluate the subjective attributes of the maskers. First, masker sound among various nature sounds was selected by preference tests. Second, signal to noise ratio between masker and maskee was determined and finally, the SQ characteristics of the masker were manipulated in order to enhance the effectiveness of masking. Experiment 1: Selection of maskers Sound-maskers were evaluated by an auditory experiment which was designed to select the sounds from the combined sounds with various maskers (nature sounds). Ahn (2002) found that nature sounds from fountain and bird are preferred in open public spaces. In the present study, nine nature sounds as shown in Table 2 were used to investigate the preferred sound masker. Paired comparison method was used: each of the stimuli pair consisted of a site noise with one sound masker and the noise from the same site with another sound masker. All the subjects were asked to select one in each pair. Duration of each sound was 7.0 s with an interval of 3.5 s (total 17.5 s). All the pairs were randomly presented to the subjects. Visual image of the actual site was presented to subjects before the auditory tests begin. Twelve subjects – aged from 20 to 30 – evaluated the sounds through headphones in a sound proof chamber. Presentation level of the sound stimuli was set to 58 and 60 dBA for road traffic and construction noise, respectively, by considering the actual sound level. Consistency tests indicated that 11 out of 12 subjects showed significant responses (p<0.05). As shown in Figure 3, ‘Stream’ and ‘Lake’ sounds were highly preferred in both cases of road traffic and construction noise. ‘Rainfall’, ‘Seagulls in port’ and ‘Wind sound’ showed relatively lower scale values of preference. Table 2: Stimuli in Experiment 1 No. Road traffic noise Construction noise 1 2 Waterfall Waterfall 3 Rainfall Rainfall 4 Stream Stream