HVAC and Refrigeration (HVAC&R) equipment designers would like to have better noise criteria to guide their designs to reduce customer complaints and to understand what characteristics of sounds people dislike. To develop a robust sound evaluation model, the main sound characteristics that influence overall judgments of the sounds need to be determined. A semantic differential test was conducted using twenty-two sounds that were a mixture of recordings and modified recordings of two types of HVAC&R units. Sound descriptions provided by subjects in another test were used to develop seventeen rating scales. A factor analysis was performed on the ratings of thirty-nine subjects with normal hearing. Three strong factors were identified that are related to loudness, tonal content or sharpness, and fluctuation or irregularity. The correlation between sound metrics and average responses on each scale was examined. The highest correlations were always for the metric that is supposed to measuring the attribute associated with a particular rating scale. Models to predict annoyance ratings from sound metrics were also examined, and models that included a loudness and sharpness term gave very good results, but these need further development and need to be tested on a greater variety of HVAC&R sounds. The results of this test have also inspired development of signal modification techniques to separate tonality and sharpness attributes so that their individual influences on overall judgments of the sounds can be examined.
HVAC, Sound Quality, Semantic differential, HVAC&R
Acoustics and Noise Control
Date of this Version