No abstract

Building services equipment often produces noise signatures with significant tones in them that

Noise with prominent tones from building mechanical systems is often detrimental to the environmental quality and leads to complaints. Previous studies have investigated the relationship between existing tonal noise metrics and human annoyance perception, but little is known about at what level the tones at assorted frequencies induce human annoyance. This paper investigates human annoyance responses due to noise with tones to produce a dose-response relationship for estimating the thresholds of annoyance to tones in noise. The subjective test is conducted using noise signals with varied loudness and tonalness thorough an Armstrong i-Ceiling system in the Nebraska indoor acoustic testing chamber. Binary logistic multiple regression models are used to predict the percentage of annoyed people or likelihood-to-complain with confidence intervals. This paper also examines the statistical performance of models with assorted noise metrics and non-acoustical variables to calculate the probability of occupants feeling annoyed for any given background noise with tonal components.

Audible tones in background noise as produced by mechanical equipment in buildings can lead to complaints from occupants. A number of metrics including prominence ratio, tone to noise ratio, tonal audibility, and Aures’ tonality have been developed to quantify the perception of tonalness, but it is not clear how these measures correlate with subjective annoyance response from listeners. This research investigates the relationship between tonality metrics and subjects’ annoyance. As reported in another paper, two tests were conducted to determine annoyance thresholds for tones in noise. One involved having subjects rate their annoyance after being exposed to background noises with differing levels of tones while working on a given task. In the second test, subjects were asked to select the minimum tone level above a set background noise condition at which they began to feel annoyed. The thresholds according to assorted tonality metrics are calculated based on the subjective results and compared to previous research.