It is currently accepted that noise is one of the most important annoyance factors in open-space offices. However, noise levels measured in open spaces of the tertiary sector rarely exceed 65 dB(A). It, therefore, appears necessary to develop a tool that can be used to assess the noise environment of these offices and identify the parameters to be taken into consideration when assessing the noise annoyance. This article presents a questionnaire to be filled by people working in such environment, and a case study in different open plan offices. The majority of the 237 respondents consider that the ambient noise level in their environment is high and that intelligible conversations between their colleagues represent the main source of noise annoyance. This annoyance was significantly correlated with their evaluation of sound intensity, which could not be represented by A-weighted level measurements.

No abstract

The perception of the noise coming from a car's door closure has been analyzed, the focus being put on the image of the quality of the car that the listener can have in mind while hearing the sound. Different experiments have been realized: a free sorting experiment for reducing the number of stimuli without any loss of generality, paired comparisons with similarity and preference ratings and, finally, free verbalizations analysis. The results have agreed on the importance of two timbre parameters, the frequency balance of the sound and its cleanness (only one temporal event should be audible). In particular, even if loudness had appeared as the most important sound feature in previously published studies, it did not in this one; the reason is probably that previous studies had focused on annoyance creating by sounds.In a more general way, this study has proved the stability of the perceptual space derived from two different methods: a multi-dimensional analysis of similarity ratings and the analysis of free verbalizations.

Ambisonics and higher order ambisonics (HOA) technologies aim at reproducing sound field either synthesised or previously recorded with dedicated microphones. Based on a spherical harmonic decomposition, the sound field is more precisely described when higher-order components are used. The presented study evaluated the sound field reproduced with four microphones of order one to four, perceptively and objectively in terms of localisation accuracy. A localisation test was performed on four microphones and a synthetic fourth order system. Results showed an improvement of the localisation with higher order ambisonic microphones. Reproduced localisation indices were calculated for the four microphones and the respective synthetic systems of order one to four. The perceptive and objective analysis revealed the same conclusions. The localisation accuracy depends on the ambisonic order as well as the source incidence. Furthermore, impairments linked to the microphones were highlighted.

Even for low noise levels, employees working in open-plan offices declare they are disturbed by different sound sources and especially by intelligible speech. This paper presents two experiments that aim at studying the effects of sound sources on task performance and cognitive load. In the first experience, thirty-five individuals were assigned a serial memorization task in four noise conditions: printers, intelligible speech, non-intelligible speech and phone rings. Noise annoyance was assessed by measuring task performance and cognitive workload (NASA-TLX). Although there was no significant difference between intelligible speech, unintelligible speech and phone ring tone signals in terms of annoyance, the performance in a memory task decreased in the presence of intelligible speech. In the second experiment, the relevance of a new intelligibility index was evaluated. This index was derived from SIIt as defined by Rhebergen et al. (2006) to deal with fluctuating noise as speech babble. Fifty-five subjects performed a serial memory task in four STIt (non-stationary STI) conditions (from 0.38 to 0.69) while STI values were kept constant (around 0.36). No significant difference appeared between the decreases in performance observed with multiple levels of intelligibility. However, an inter-individual variability appeared in performance measurements. The best performers seemed to be less affected by the high level of intelligibility condition than the other ones.