Reinhard Weber

Semantic evaluations of noise with tonal components in Japan, France, and Germany : A cross-cultural comparison

An evaluation of tonal components in noise using a semantic differential approach yields several perceptual and connotative factors. This study investigates the effect of culture on these factors with the aid of equivalent listening tests carried out in Japan (n=20), France (n=23), and Germany (n=20). The datas equivalence level is determined by a bias analysis. This analysis gives insight in the cross-cultural validity of the scales used for sound character determination. Three factors were extracted by factor analysis in all cultural subsamples: pleasant, metallic, and power. By employing appropriate target rotations of the factor spaces, the rotated factors were compared and they yield high similarities between the different cultural subsamples. To check cross-cultural differences in means, an item bias analysis was conducted. The a priori assumption of unbiased scales is rejected; the differences obtained are partially linked to bias effects. Acoustical sound descriptors were additionally tested for the semantic dimensions. The high agreement in judgments between the different cultural subsamples contrast the moderate success of the signal parameters to describe the dimensions.

Interior car sound quality—Assessment of acceleration noises

Investigation of booming noise effects was the reason for the present study. Interior noises of six different cars accelerating from 1500 to 6000 rpm at third gear (full load) were presented over headphones. Seventy normal‐hearing subjects (approximately one‐third males and two‐thirds females) aged from 20 to 60 years (at least 15 subjects per decade) took part in two experiments. After an initial orientation phase, preference judgments were made using a paired comparison paradigm. In experiment 2, the noises were assessed on a bipolar categorical scale (disagreeable/agreeable) with nine subdivisions. Immediately after the experiments, an interview was conducted to learn about the subjects’ impressions about the noises. The preference data were analyzed using MDPREF and compared to the categorical judgment data for different subsets of subjects. The results on subjective data will be presented, together with a correlation analysis on their relationship with a number of acoustical and psychoacoustical sign...

Noise sensitivity and loudness derivative index for urban road traffic noise annoyance computation

Urban road traffic composed of powered-two-wheelers (PTWs), buses, heavy, and light vehicles is a major source of noise annoyance. In order to assess annoyance models considering different acoustical and non-acoustical factors, a laboratory experiment on short-term annoyance due to urban road traffic noise was conducted. At the end of the experiment, participants were asked to rate their noise sensitivity and to describe the noise sequences they heard. This verbalization task highlights that annoyance ratings are highly influenced by the presence of PTWs and by different acoustical features: noise intensity, irregular temporal amplitude variation, regular amplitude modulation, and spectral content. These features, except irregular temporal amplitude variation, are satisfactorily characterized by the loudness, the total energy of tonal components and the sputtering and nasal indices. Introduction of the temporal derivative of loudness allows successful modeling of perceived amplitude variations. Its contribution to the tested annoyance models is high and seems to be higher than the contribution of mean loudness index. A multilevel regression is performed to assess annoyance models using selected acoustical indices and noise sensitivity. Three models are found to be promising for further studies that aim to enhance current annoyance models.

Comparison of detection threshold measurements and modeling for approaching electric cars and conventional cars presented in traffic and pink noise

This study investigates the difference in audibility of an approaching conventional car with internal combustion engine and an electric car at various velocities. The goal was to compare the risk that pedestrians do not hear the approaching car in time. Binaural recordings of each of these approaching cars were presented together with either a traffic noise masker or a pink noise masker. In the first detection experiment, the threshold level was determined for which the cars could just be detected. In a second reaction time experiment, the moment was determined at which the approaching car was first detectable. This measured reaction time should give an indication about how much time a person has to evade an impending collision. Results indicated that slowly approaching electric cars where less audible than cars with a conventional engine. The results also showed that the decrement of reaction times as a function of SNR was halved when pink noise was used instead of traffic noise. A psycho-acoustic maskin...

Perceptual space, pleasantness and periodicity of multi-tone sounds

Technical sounds often contain several tonal components, forming a multi-tone sound. The present study investigates the perception of multi-tone sounds consisting of two harmonic complexes with different fundamental frequencies and combination tones with frequencies that are equal to the sum of multiple integers of the two fundamentals. The experimental parameter is the ratio between the two fundamental frequencies ρ. A total of 15 synthetic multi-tone sounds are rated by 37 participants. In the first experiment, the perceptual space is assessed based on 16 adjective scales using categorical scaling. The resulting perceptual space has the four dimensions (i) pleasant, (ii) power, (iii) temporal structure, and (iv) spectral content of the sounds. In the second experiment, the pleasantness is measured with a paired comparison test. The data consistently show that sounds based on ratios of small integers (e.g., ρ=4:3) are significantly less pleasant than sounds with ratios based on large integers which were constructed by a slight detuning from a ratio of small integers. The repetition rate derived from an autocorrelation analysis of the stimuli turns out to be a good predictor of the (un-)pleasantness sensation.

Pitch strength, tone segregation, and frequency difference limen.

For pure tones in narrowband noise, two percepts are considered. At a sufficiently high tone‐to‐noise ratio, the tone can be clearly distinguished from the noise: the tone and the noise segregate. At a low signal‐to‐noise (S/N) ratio, slightly above the masking threshold, only a tonal noise is perceived: the tone is fused with the narrowband noise. With respect to these phenomena, two hypotheses are tested relating pitch strength, tone segregation, and frequency difference limen. The first hypothesis is that with the aforementioned stimuli, pitch strength alone is capable to serve as a segregation cue. The second hypothesis tested is that frequency difference limen is the psychophysical correlate of this sensation. Three experiments were conducted. The first experiment was an identification experiment. Listeners had to identify a segregated tone at five center‐frequencies (250–4000 Hz, octave‐wise) in the presence of narrowband noise. The second experiment measured the detection threshold of the tone. The...

Subjective evaluation of loudness and preference for noise containing audible tones

Noise in living environments often contains unwanted tonal components that contribute to the unpleasant and intrusive character of a sound and usually lead to an increased annoyance. In the measurement and assessment of noise immissions after the German DIN 45681:2005-3 standard, the higher annoyance due to audible tonal components is covered by so called tone adjustments. Depending on the SNR of a tonal component within each critical band, the level of a sound can be charged with a penalty of up to 6 dB. In this study, the psychoacoustic penalty levels for sounds containing tonal components are directly determined in listening tests. The points of subjective equality (PSEs) for loudness and preference are determined using a matching procedure. The level of the test sound with tonal components is varied until it is equally loud / preferred as a tone-free reference sound that is constant in level. The level difference between test and reference sound at the PSEs is a direct measure of the penalty level. In...

Acoustical aspects of travel comfort in the aircraft cabin

In several European projects tests were conducted during real flights and in aircraft simulators in order to identify important parameters of travel comfort. Physical environmental, physiological and questionnaire data were measured and investigated with respect to human perception. Besides numerous intrinsic quantities aecting health, well-being and awareness of ambience defined measurable environmental parameters influence flight and cabin crews’ as well as passengers’ perception, psychology and physiology. Sound and vibration are relevant environmental parameters with impact on passenger and crew. Other important influence is caused by e.g. air quality, pressure, local climate. Statistical analysis of the collected data reveals significant correlations between environment and human response for selected groups of test persons. A specific acoustic comfort is not well defined but part of common well-being and comfort. The demand for a general human response model is discussed, which relates ambience and perception.

Cross‐modality matching of loudness and whole‐body vibration strength

In every day live humans are often exposed to noise and vibration simultaneously. Regarding comfort issues inside a car, it is of interest to know, whether the simultaneous perceptions of noise and vibration interfere. Laboratory tests have been carried out with temporally overlapping, partially overlapping and non overlapping presentations of acoustic and whole-body vibration stimuli. Sitting on a rigid chair on a vibration test bench, participants are asked to judge the strength of the excited whole-body vibrations in comparison to the loudness of noise presented via closed headphones. An adaptive method is employed to determine the subjective point of equality (PSE) between both sensory channels. The acoustic stimulus is a 1/2-octave band-pass noise centred at 100 Hz with a fixed noise level, the vibration stimulus is a narrowband noise, also 1/2-octave broad with a centre frequency of 31,5 Hz. The signals have an equal duration of one second. The signal parameters are chosen with respect to specific situations in a car. The PSE’s of the loudness and the vibration strength as a function of the temporal order of the acoustic and vibration stimuli exhibit no dependency on the degree of temporal overlap of the stimulus presentation.

Assessment of impact of acoustic and nonacoustic parameters on performance and well‐being

It is of interest to estimate the influence of the environment in a specific work place area on the performance and well‐being of people. Investigations have been carried out for the cabin environment of an airplane and for class rooms. Acoustics is only one issue of a variety of environmental factors, therefore the combined impact of temperature, humidity, air quality, lighting, vibration, etc. on human perception is the subject of psychophysical research. Methods for the objective assessment of subjective impressions have been developed for applications in acoustics for a long time, e.g., for concert hall acoustics, noise evaluation, and sound design. The methodology relies on questionnaires, measurement of acoustic parameters, ear‐related signal processing and analysis, and on correlation of the physical input with subjective output. Methodology and results are presented from measurements of noise and vibration, temperature and humidity in aircraft simulators, and of reverberation, coloring, and lighti...