Influence of vehicle-exterior image on sound quality evaluation of Japanese and German drivers

Transcription

1 Acoust. Sci. & Tech. 37, 3 (16) TECHNICAL REPORT #16 The Acoustical Society of Japan Influence of vehicle-exterior image on sound quality evaluation of Japanese and German drivers Junji Yoshida 1;2;, Florian Völk 2, Hugo Fastl 2 and Gerhard Rigoll 3 1 Osaka Institute of Technology, Omiya, Asahi-ku, Osaka, Japan 2 AG Technische Akustik, MMK, Technische Universität München, Arcisstr. 21, Munich, Germany 3 MMK, Technische Universität München, Arcisstr. 21, Munich, Germany (Received 31 August, Accepted for publication December ) Abstract: In this study, an influence of vehicle-exterior image on sound quality evaluation was investigated. Luxury- or sporty-vehicle image was presented to Japanese or German frequent drivers, while listening to an acceleration sound. Subsequently, loudness, luxury and sportiness for the sound were rated. As the result, significant influence was not observed in the loudness and luxury evaluation in both nationalities. However, sportiness rating to the presented sound was markedly increased when the sporty-vehicle image was presented with the sound. Consequently, it appeared that the influence of the vehicle-exterior image on the sound quality rating differs depending on the evaluation parameter and sportiness is markedly affected by the visual information. Keywords: Sound quality, Vehicle-exterior image, Auditory-visual interaction, Acceleration sound PACS number: Lj, Lj [doi:.10/ast ] 1. INTRODUCTION Quietness is one important aspect of vehicle performance. In recent years, not only the quietness but also the sound quality has been focused on in vehicle development. Various sound quality indices have been developed on the basis of auditory perception characteristics and some of them consider the nationality of the drivers [1 5]. In most of these studies, auditory evaluation tests were carried out by presenting sounds without designated visual stimuli. However, recent studies have revealed that non-acoustical factors such as visual stimuli can affect auditory evaluations [6 13]. In actual driving situations, when hearing interior sounds, drivers are aware of the vehicle design and have an impression of whether they are driving, for example, a sporty or a luxury model. This leads to the hypothesis that the sound quality evaluations of vehicle-interior sounds may be affected by the impression of the exterior design. In our previous studies, we investigated the influence of a vehicle-exterior image on the loudness evaluation for acceleration sound, and the results showed that the perceived loudness of a sound presented with a luxuryvehicle image was increased compared with the sound yoshida@med.oit.ac.jp associated with a sporty-vehicle image [12,13]. Considering the recent increase in the need to improve sound quality, investigating only the influence on loudness is not sufficient; it is also important to investigate the influence on sound quality evaluation. In addition, the influence may be different among different nationalities owing to differences in the cultural background regarding vehicles and sound [3,14,]. Accordingly, we investigated the influence of the vehicle-exterior image on sound quality evaluation by presenting exterior images of different vehicles to Japanese and German drivers in this study. 2. EVALUATION OF VEHICLE-EXTERIOR IMAGES Vehicle-exterior images were evaluated to determine the participants overall impression of the images and to select the images for the following sound quality experiments Procedure, Stimuli and Participants Thirty exterior images of vehicles sold in the EU, USA and Japan were employed as visual stimuli. All vehicles had similar colors (white or silver) in order to minimize the influence of color on the sound quality rating []. The evaluation parameters luxury and sportiness were 123

2 Sportiness VS01 VS03 VS02 VS05* VS04 VL04 VL01* VL05 VL03 VL02 Acoust. Sci. & Tech. 37, 3 (16) Table 1 Vehicle characteristics of selected images. (a) Selected images for Japanese drivers. Luxury-vehicle images Sporty-vehicle images Image Type Cntry cc Image Type Cntry cc VL01 Sedan Japan 3,500 VS01 Coupe U.K. 1,800 VL02 Sedan Japan 5,000 VS02 Coupe Japan 2,600 VL03 Sedan U.K. 6,700 VS03 Coupe Japan 3,400 VL04 Sedan U.S.A. 3,500 VS04 Sedan Japan 2,000 VL05 Sedan Japan 4,500 VS05 Coupe Germany 3, Luxury Sportiness (a) Impression map for Japanese drivers VS01 VS02 VS04 VL05 VS03 VS06* VL02 VL06* VL04 VL Luxury (b) Impression map for German drivers Fig. 1 Vehicle-exterior image impression maps for Japanese (a) and German (b) drivers. The axes represent the luxury and sportiness scales and small circles represent the ratings of the vehicle-exterior images (bold and filled circles are used for the sound quality evaluation). used as indicators of the impression. The participants rated these parameters on separate integer scales from 3 to +3 in seven categories, with negative values indicating negative impressions, after looking at the vehicle image on a sheet or a PC monitor. Twenty Japanese drivers and 16 German drivers participated in this test. The average ages of the Japanese and German participants were 23 and 33 years, respectively, and all participants were male except for one female German participant. All participants had normal visual acuity Results To obtain the overall judgment (impression) of each vehicle image, the scores for each vehicle were averaged among all participants of each nationality and normalized to give a standard deviation of one for the distributions of the luxury and sportiness scores. Figures 1(a) and 1(b) show so-called impression maps, composed of the luxury and sportiness ratings for the Japanese and German drivers (small circles). (b) Selected images for German drivers. Luxury vehicle image Sporty vehicle image Image Type Cntry cc Image Type Cntry cc VL06 Sedan Germany 3,500 VS01 Coupe U.K. 1,800 VL02 Sedan Japan 5,000 VS02 Coupe Japan 2,600 VL03 Sedan U.K. 6,700 VS03 Coupe Japan 3,400 VL04 Sedan U.S.A. 3,500 VS04 Sedan Japan 2,000 VL05 Sedan Japan 4,500 VS06 Coupe Germany 4,0 As shown in the Japanese impression map (Fig. 1(a)), some vehicle images are located close to each other in the upper part of the map. This indicates that these vehicle images were perceived as sporty vehicles. The vehicle images located on the right side were regarded as luxury vehicles by Japanese drivers and the images located in the bottom left did not give an impression of luxury or sportiness. A similar tendency was obtained for the German drivers as shown in Fig. 1(b). There are several vehicle images giving impression of high sportiness or high luxury to the German drivers. However, the vehicle image positions in the German map were found to be more widely scattered than those in the Japanese map. Several vehicle images giving an impression of luxury in the German map were evaluated as being somewhat sporty, although most images giving an impression of luxury in the Japanese map were regarded as less sporty. It is considered that the driving situation in Germany, where vehicles are driven at a higher speed and are more sporty, is different from that in Japan owing to the different traffic conditions. Therefore, a slight difference in the impression may have been generated for some vehicle-exterior images. On the basis of these impression maps, five vehicle images giving high-luxury impression (filled circles) and five images giving a high-sportiness impression (bold circles) were used for the subsequent sound quality evaluation, because the impressions of sounds were considered to be clearly different between luxury and sporty vehicles. Eight images were common to both maps and the two remaining images were different (marked by stars in Fig. 1 and underlined in Table 1). The vehicle type (Type), the country of development (Cntry) and the engine displacement (cc) of the vehicles in the images employed for Japanese and German drivers are shown in Table

3 J. YOSHIDA et al.: INFLUENCE OF VEHICLE IMAGE ON SOUND QUALITY Table 2 L Aeq for the presented vehicle acceleration sounds. L Aeq L Aeq SL SS SL SS SL SS SL SS SL SS SOUND QUALITY EVALUATION (AUDITORY-VISUAL TEST) The sound quality of ten acceleration sounds was evaluated using the selected vehicle exterior images in this experiment Procedure, Stimuli and Participants The evaluation parameters luxury and sportiness were used as indicators of the sound quality as they were also used during the visual test [3]. In addition to these parameters, loudness was also used as a fundamental parameter of the sound quality. Regarding the experimental participants, it has been reported that participants who frequently drive vehicles can carry out more stable evaluations of presented vehicle sounds than infrequent drivers [12,13]. Hence, we selected the participants who drive vehicles frequently (at least twice a week) from participants in the first visual evaluation test. As a result, nine Japanese drivers and five German drivers participated in this test. All Japanese and German participants were male and their average ages were 23 and 39, respectively. All of them had normal hearing and visual acuity. The same ten acceleration sounds recorded binaurally (HEAD acoustics HMS) at a sampling rate of 44.1 khz inside vehicles at the passenger seat position were used for both nationalities. The vehicles used for the recordings were of a similar class to those selected in the visual test. Five of the ten sounds were sounds of luxury vehicles (SL01 SL05) the others were sounds of sporty vehicles (SS01 SS05). As the sounds were intended to provide the impression of an accelerating vehicle to the participants, acceleration sounds from low to high engine rotational speeds were used (about 1,000 to 6,000 rpm). Also, fade-in and fade-out processing was applied to the beginning and end of each sound to avoid click artifacts. The sounds were edited to durations of about s and presented dichotically at their original L Aeq using Sennheiser HD 600 headphones [16]. L Aeq for each presented sound is shown in Table 2. The auditory-visual test was conducted using software that sequentially displayed the images, selected as described in the previous section, on a PC monitor and instructed the participants to imagine driving the vehicle. After clicking a button on the display, the acceleration sound was presented while the vehicle image was still displayed. By clicking another button, the participants rated the sound quality using various categories. For instance in the loudness evaluation, the participants rated the loudness within the categories very soft, soft, relatively soft, neither soft nor loud, relatively loud, loud and very loud. Each of these major categories was subdivided in four subcategories, resulting in 31 levels. For the analysis, integers from 1 to 31 were assigned to the categories from very soft to very loud. The luxury and sportiness ratings were obtained using the same procedure. After the rating had been given, the next vehicle image was presented and so on. The image number, sound number and rating of each parameter were automatically saved in a file. All participants practiced the evaluation procedure before the test was performed Auditory-Visual Stimuli In the auditory-visual test, changes in the sound quality ratings of the ten acceleration sounds were evaluated when the presented vehicle image was changed from luxury to sporty and vice versa. Hence, two stimulus patterns composed of the images and sounds were presented. In pattern 1 (matching), each of the luxury-vehicle sounds SL01,..., SL05 was combined with one of the luxuryvehicle images, and each of the sporty-vehicle sounds SS01,..., SS05 was combined with one of the sportyvehicle images. In pattern 2 (conflicting), sporty-vehicle sounds were presented with luxury-vehicle images and luxury-vehicle sounds were presented with sporty-vehicle images. The resulting combinations for both nationalities are shown in Table 3. In the luxury-vehicle sound group, the difference in the sound quality ratings from pattern 1 to pattern 2 shows the difference in the sound quality evaluation when the vehicle image is changed from luxury to sporty. In the sportyvehicle sound group, the difference in the ratings from pattern 1 to pattern 2 shows the difference in the sound quality evaluation when the design is changed from sporty to luxury. A single sound quality evaluation including loudness, luxury and sportiness ratings was carried out as a trial, and a series of trials was performed in a session in both countries. A number of sessions were conducted for each participant to evaluate the significant influence of the vehicle images on the sound quality ratings in the evaluation variance. In a session, all sounds in a single test pattern were evaluated twice in random order, resulting in judgments. For each pattern, each participant performed five separate sessions. Hence, each subject performed 0 trials (two test patterns trials five iterations), resulting in a total of 1,800 trials for the 1

4 Acoust. Sci. & Tech. 37, 3 (16) Table 3 Presentation patterns of sounds and images. (a) Patterns for Japanese drivers. 1 VL01 VL02 VL03 VL04 VL05 VS01 VS02 VS03 VS04 VS05 Luxury-vehicle image Sporty-vehicle image 2 VS01 VS02 VS03 VS04 VS05 VL01 VL02 VL03 VL04 VL05 Sporty-vehicle image Luxury-vehicle image (b) Patterns for German drivers. 1 VL06 VL02 VL03 VL04 VL05 VS01 VS02 VS03 VS04 VS06 Luxury-vehicle image Sporty-vehicle image 2 VS01 VS02 VS03 VS04 VS06 VL06 VL02 VL03 VL04 VL05 Sporty-vehicle image Luxury-vehicle image Loudness rating Loudness rating (a) Japanese drivers (b) German drivers Fig. 2 Loudness rating results for different vehicleexterior designs. SL: luxury-vehicle sound, SS: sportyvehicle sound. Japanese drivers (0 trials nine participants) and 1,000 trials for the German drivers (0 trials five participants). About min was required to complete one experimental session Results Loudness ratings Figure 2 shows the average loudness ratings of the Japanese and German drivers for the ten different sounds (SL: luxury-vehicle sound, SS: sporty-vehicle sound). The white and gray bars indicate the loudness ratings while exposed to images of sporty and luxury vehicles, respectively. The Japanese drivers (Fig. 2(a)) rated the loudness of most sounds presented with luxury-vehicle images slightly higher than that of the same sounds presented with sportyvehicle images, similarly to in our previous studies [12,13]. However, the difference between most pairs was not significant ( p = 0:05, t-test) except for sound SL03 (double asterisks in Fig. 2(a)), although the difference between most pairs was significant in the previous studies [12,13]. If the luxury ratings of the employed luxury-vehicle images had been significantly larger than those of the sportyvehicle images, the loudness difference would have been expected to be large as observed in the previous studies. This is because the loudness inside the car is considered to be related to whether or not the vehicle is a luxury car in general. When the vehicle is high luxury, the interior sound is expected to be quiet. The participants were considered to evaluate the presented acceleration sound by comparison with the expected quiet sound of a luxury vehicle. Consequently, the sound with the luxury-vehicle image tended to be evaluated as louder in the previous studies. However, the difference between the luxury scores of the luxury- and sporty-vehicle images was small in this study as shown in Fig. 1(a). As a result, the loudness differences between the sounds with the luxury-vehicle images and those with the sporty-vehicle images were small, in contrast to the previous studies. The impression of the presented vehicle images was considered to be different according to the participants interest and knowledge of vehicles. The participants in this study were not the same in those of the previous studies. This may be one of the factors generating the difference in the loudness evaluation between our previous tests and the present test. In the loudness evaluation by the German drivers (Fig. 2(b)), they tended to judge the loudnesses of many of the sounds while looking at the images of luxury vehicles as being slightly lower than those when looking at the sporty-vehicle images. The differences were also insignificant except for SL04 (double asterisks in Fig. 2(b)). From these results, the loudness ratings were not changed markedly by the impression of the vehicle-exterior image for both nationalities Luxury ratings Figure 3 shows the average luxury ratings for the Japanese and German drivers for the ten different sounds. 126

5 J. YOSHIDA et al.: INFLUENCE OF VEHICLE IMAGE ON SOUND QUALITY Luxury rating Sportiness rating Luxury rating (a) Japanese drivers Sportiness rating 5 (a) Japanese drivers (b) German drivers 5 (b) German drivers Fig. 3 Luxury rating results for different vehicle-exterior designs. SL: luxury-vehicle sound, SS: sportyvehicle sound. The white and gray bars indicate the luxury ratings while exposed to images of sporty and luxury vehicles, respectively. The results indicate that the Japanese drivers (Fig. 3(a)) rated the luxury for many sounds presented with luxury-vehicle images slightly lower than that for the same sounds presented with sporty-vehicle images. However, the differences for most pairs were also not significant, similarly to in the loudness evaluation, except for SL02 and SS03 as indicated by double asterisks. The small difference in the luxury scores between the luxuryvehicle images and the sporty-vehicle images is considered to be one of the reasons for this. On the other hand, a uniform tendency was not observed in the luxury evaluation of the German drivers; four pairs of audio-visual stimuli exhibited significant differences. The luxury ratings for SL01 and SL04 associated with luxury-vehicle images were significantly higher than those with sporty-vehicle images (Fig. 3(b)). However, the luxury ratings for SL05 and SS05 with luxury-vehicle images were significantly lower than those with sporty-vehicle images. These results suggest that neither the loudness rating nor the luxury rating was markedly or uniformly influenced by the impression of the vehicle-exterior image for both nationalities Sportiness ratings Figure 4 shows the sportiness ratings for the Japanese and German drivers. The white and gray bars indicate the ratings while exposed to images of sporty and luxury vehicles, respectively. Fig. 4 Sportiness rating results for different vehicleexterior designs. SL: luxury-vehicle sound, SS: sportyvehicle sound. In contrast to the loudness and luxury ratings, the sportiness ratings of most sounds were markedly increased when participants of both nationalities were presented with sporty-vehicle images, as shown in Fig. 4. The differences in the sportiness ratings were much larger than those in the luxury and loudness ratings with significant differences observed for most sounds. As the reason for these results, the difference in the ratings of vehicle images may have had an important role. The sporty-vehicle images employed in this auditory-visual test were rated much sportier than the luxury-vehicle images as shown in Fig. 1. Therefore, the sportiness ratings for the acceleration sounds presented with the sporty-vehicle images are considered to be evaluated significantly higher than the sounds with the luxury-vehicle images. For the German participants, the differences in the sportiness ratings between the sounds presented with sporty-vehicle images and those with luxury-vehicle images were relatively smaller than those for the Japanese participants, although the general tendency was the same for both nationalities. This result seems to be affected by the large variance of the sportiness ratings of the employed vehicle images as shown in Fig. 1(b). In contrast to the close positions of the sporty-vehicle images in the Japanese vehicle impression map (Fig. 1(a)), the positions in the German map vary widely. For instance, the sportiness rating for the VS04 image was lower in the German map than in the Japanese map. Accordingly, the sportiness ratings for sounds SL04 and SS04 did not increased 127

6 Acoust. Sci. & Tech. 37, 3 (16) Fig. 5 Influence of sporty-vehicle image on sportiness rating of sound. Showing the primary process occurring for Japanese and German drivers when evaluating the sportiness for a sound. significantly when they were presented with VS04 as the sporty-vehicle image. From these experimental results, the sound quality ratings were found to be affected by the impression of the vehicle-exterior images. Among the three evaluation parameters investigated in this study, only the sportiness rating of the sound was significantly increased by presenting participants with a vehicle image conveying a high level of sportiness. In addition, the influence of the vehicleexterior image on the sound quality rating was observed to be similar for the Japanese and German drivers for which the sound quality rating was affected significantly when the impression of the vehicle image was strong Summary and Discussion In this study, we investigated the influence of the vehicle-exterior image on sound quality ratings for Japanese and German drivers. The results showed that the sportiness rating of sounds was markedly affected by presenting a vehicle-exterior image that gave a strong impression of sportiness for drivers of both nationalities, although the influence on the loudness and luxury ratings was insignificant when the difference in the luxury impression of the vehicle images was small. As the reason why participants gave increased sportiness ratings for sounds presented with images of vehicles giving a stronger impression of sportiness, they seem to have integrated the impression of a sporty-vehicle elicited by the sporty vehicle image into their sound rating, as shown in Fig. 5. From these results, the sound quality ratings were found to be affected by the vehicle impression for drivers of both nationalities. In particular, when the visual impression was strong, the sound quality rating changed significantly as observed in the sportiness ratings. Therefore, it is important to take into account this influence in the development of vehicles with sporty sounds. If we develop a vehicle with a sporty sound and the vehicleexterior design is very sporty, we can expect that the design enhances the perception of a sporty sound. On the other hand, if we develop a sporty sound for a sedan-type vehicle, the impression of sportiness of the acceleration sound may be decreased by the visual information, and a more sporty sound should be designed to compensate for the visual influence. REFERENCES [1] R. Bisping, S. Giehl and M. Vogt, A standardized scale for the assessment of car interior sound quality, SAE Trans., 6, (1997). [2] W. Biermayer, S. Thomann and F. Brand, A software tool for noise quality and brand sound development, Proc. SAE 01 Noise & Vibration Conference & Exposition, , pp (01). [3] N. Noumura and J. Yoshida, Perception modeling and quantification of sound quality in cabin, SAE Trans., 112, (04). [4] H. Fastl and E. Zwicker, Psychoacoustics Facts and Models, 3rd ed. (Springer, Berlin, 07). [5] P. A. Jennings, G. Dunne, R. Williams and S. Giudice, Tools and techniques for understanding the fundamentals of automotive sound quality, Proc. IMechE, Part D: J. Automob. Eng., 224, (). [6] J. Blauert and U. Jekosch, Sound-quality evaluation A multi-layered problem, Acta Acust. united Ac., 83, (1997). [7] S. Namba, S. Kuwano, A. Kinoshita and Y. Hayakawa, Psychological evaluation of noise in passenger cars The effect of visual monitoring and the measurement of habituation, J. Sound Vib., 5, (1997). [8] U. Jekosch, Meaning in the context of sound quality assessment, Acta Acust. united Ac., 85, (1999). [9] C. Patsouras, T. G. Filippou and H. Fastl, Influences of color on the loudness judgment, Proc. Forum Acousticum Sevillia, pp. 1 6 (02). [] D. Menzel, H. Fastl, R. Graf and J. Hellbrück, Influence of vehicle color on loudness judgments, J. Acoust. Soc. Am., 123, (08). [11] L. Humphreys, S. Giudice, P. Jennings, R. Cain, W. Song and G. Dunne, The influence of company identity on the perception of vehicle sounds, Ergonomics, 54, (11). [12] J. Yoshida and T. Igata, Influence of impression of vehicle styling on loudness of acceleration sounds in cabin, J. Acoust. Soc. Am., 1, EL19 EL24 (11). [13] J. Yoshida and T. Igata, Dependence of loudness evaluation by drivers on vehicle styling, J. Acoust. Soc. Am., 132, (12). [14] S. Namba, S. Kuwano and A. Schick, Cross-cultural study on noise problems, J. Acoust. Soc. Jpn. (E), 7, (1986). [] T. Rader, M. Morinaga, T. Matsui, H. Fastl, S. Kuwano and S. Namba, Crosscultural effects in audio-visual interactions, Trans. TC Noise and Vibration Acoust. Soc. Jpn., N-04-31, pp. 1 6 (04). [16] F. Völk, Interrelations of virtual acoustics and hearing research by the example of binaural synthesis, Dr. Hut, München (13). 128