Crest factor and short-duration transients: influence of environmental background, event duration and measurement time weightings

摘要

Short-duration transients may typically be observed on representations of time signal histories.Crest factor (the unitless ratio of peak to rms magnitude over an interval) is a useful startingpoint for observing and quantifying impulsive acoustic transients in Information Technologydevices. With low-level sound sources, low-frequency ambient noise can dominate the rmsvalue, with two penalties: subjective importance of transients relative to machine sound may beunder-represented, and transients may “ride on” low-frequency content, offsetting them relativeto zero-crossing and complicating their crest factor measurement. Conveying crest factorinformation into level and spectrum measurement versus time ideally requires a zero timeweighting, feasible with level but a practical impossibility with spectral measurement. Thebehavior of different time weightings in level measurement vs. time will be assessed, first withsynthetic and then with product signals, and examples of different spectrum vs. time methodsand settings will be given. Considerations of subjective audibility, methods and settingsmeeting audibility tests, and validation by both listening and psychoacoustic measures will bepresented.