Loudness and annoyance of impulsive office noise

A study has been made on the evaluation of natural impulsive noises by subjective and objective procedures. One special example was used to get an indication of whether existing objective calculation methods might give sufficient agreement with subjective tests. The office noise sample studied bursts from a single typewriter operated at repetition rates between 1.5 and 15 strokes per second. A forced-pair comparison procedure of the constant stimulus type was used for the subjective evaluation. This choice has been prompted by the good results which we obtained with a corresponding method for speech quality measurements. In the subjective tests reported here, we asked for loudness and for annoyance comparisons. Two types of broad-band noise were utilized as reference signals. In addition to this, narrow-band noise centered at 1000 Hz was used to calibrate the two broad-band noises subjectively, and was also compared directly to the test noise. The subjective measurement results of this study are compared with those obtained by several well-known objective methods for noise rating and loudness evaluations.