Technical NoteUncertainties of the frequency response of wet microphone windscreens
Introduction
Environmental noise measurements often take place over long periods during which acoustic sensors and their outdoor protection are exposed to rainfall. Such situations are very common for example in noise monitoring stations around transport infrastructures. In standards, the influence of rain is only considered by the rain-induced noise that can strongly degrade the signal to noise ratio of the measurement, and common practice is to invalidate the data recorded during a period of rain. Unfortunately, very little information can be found in the literature on the influence of the water that can remain inside the foam of the windscreen, after a rain period, on the frequency response of the windscreen. Currently, this influence is ignored but it could lead to a larger uncertainty in the measurement results.
Very few authors have addressed this issue: Nelson and Godfrey [1] has concluded that there is no significant difference between the frequency response of a wet windscreen and of a dry windscreen. This conclusion was nevertheless based on an experimental study that was concerning only one type of windscreen, that no longer exists now, and only road traffic noise was considered. Moreover, no information on uncertainties is provided in this study.
The purpose of the present paper is to investigate: (a) if the presence of water inside the foam of a windscreen has a significant effect on the frequency response of a windscreen, (b) the significant parameters that could influence this effect (rainfall rate, amounts of precipitation, type of the windscreen …), (c) the uncertainties on an acoustic measurement result to be taken into account for this effect, and finally (d) how long this effect remains after a rainfall period.
An extensive experimental study has been made in the facilities of the Laboratoire National de métrologie et d’Essai (Trappes, F) in order to measure the change in system sensitivity for measurement conditions consistent with dry and wet windscreen. This campaign includes many types of currently available commercial windscreens. The results given here do not depend on a specific environmental noise source and can be applied for any kind of environmental noise measurement for frequencies from 80 Hz to 10 kHz.
The first section deals with the relative frequency response of a wet/dry windscreen and with parameters that can influence this response. The second section presents how to correct measurements in order to take into account the influence of water inside the foam, and also the associated uncertainties due to disparities in windscreens. Many experimental numerical values are provided for practical applications. The last section provides some information about the time span during which correction and uncertainties must be considered after a rain period.
Section snippets
Uncertainties
Depending on frequencies, the water that remains inside the foam can significantly change the frequency response of the windscreen (see Section 2.3). This change may induce a bias in the acoustic measurement result that may not be negligible when considering an uncertainty budget. This bias is also associated with an uncertainty coming from the variability on the frequency response for each type of windscreen, or from the (small) effect of the height of precipitation and of the rainfall rate.
How long is it necessary to take into account uncertainties?
After a rainfall period, the uncertainty must be considered as long as the water inside the foam significantly influences the acoustic response of the windscreen. This duration strongly depends on the drying time of the foam.
An experimental study of the drying behavior of every windscreen has been done with the help of an environmental chamber and of the standardized system of artificial rain generation. Simple empirical relations of the normalized mass of water remaining inside the foam along
Conclusion
An experimental campaign has been carried out in order to investigate the acoustic response of several types of wet windscreen microphones. The effect of the water impregnation of the foam is significant only for frequencies above 1 kHz and for amounts of precipitation above 1 mm. This effect leads to a bias between the wet and the dry situation, and also to a larger uncertainty of the acoustic response of a windscreen. This effect depends on the type of windscreen: the bias can be up to 2 dB for
Acknowledgements
This research was partly financially supported by the French Ministry of Sustainable Development (DGPR, SETRA, IFSTTAR). The authors are grateful to the members of the working group on environmental acoustics uncertainties of the French Association for Standardization (AFNOR) for their valuable comments and especially to M. Lecollinet for helpful discussion about uncertainties. The authors also thank Mr Durocher (LNE) for help on the experimental implementation of ISO/IEC 61672-2.
References (6)
- et al.
A note on the acoustic sensitivity of microphone windscreens
Appl Acoust
(1978) - French Standard Pr S 31-115. Guide pour l’évaluation des incertitudes de mesurage en acoustique de l’environnement....
- International Standard ISO/IEC 61672-2:2003. Electroacoustics – Sound level meters – Part 2: Pattern evaluation tests...
Cited by (7)
Design of MEMS microphone protective membranes for continuous outdoor applications
2021, Applied AcousticsCitation Excerpt :Attached membranes provide a sufficient protection to the microphones from water and dust over a long period of outdoor usage. Compared with these results of [4] the membranes provide a direct protection with no time dependencies due to drying windscreens. The question of the influences was answered in the last section.
Measurement-based auralization methodology for the assessment of noise mitigation measures
2016, Journal of Sound and VibrationInfluence of rainfall on the noise shielding by a green roof
2014, Building and EnvironmentCitation Excerpt :As a result, the potentially changed response by the windscreen would cancel out (to some extent). In addition, detailed analysis of the effect of rainfall on common windscreens [29] showed that below 1 kHz, effects could be neglected anyhow, even in case of strong rainfall events. During the artificial wetting experiment, care was taken not to moisten the windscreens, to fully exclude wet windscreen effects.
A Study on Robustness to Perturbations for Representations of Environmental Sound
2022, European Signal Processing ConferenceTransient acoustic detection for hostile fire indication for helicopters
2020, Proceedings of SPIE - The International Society for Optical Engineering