Continuous Assessment of the Unpleasantness of a Sound

The goal of the present study was to estimate the accuracy of a continuous assessment method used for sounds of long duration. Such sounds were created by mixing stationary low-frequency random noise and twelve events of shorter duration (1 s to 11 s). These events were obtained from three stationary unpleasant sounds, the amplitude of which being varied in four different ways. Long duration sounds were presented to listeners who had to continuously assess their unpleasantness by moving a cursor sliding along a five-levels scale (from 'not at all unpleasant' to 'extremely unpleasant'). A second experiment, using a paired comparison procedure, gave a reference of the unpleasantness of the three sounds and the background noise used to synthesize the whole stimuli. The comparison of the results of both experiments allowed to confirm the validity of the continuous assessment, though its results were slightly dependent on the duration of the events.