A ranking procedure, called random access, which has proven very successful for questions of sound quality evaluation1,2 (e.g. Fastl 2000, 2002) is illustrated in figure 1. In the example displayed, six sounds denoted A through F have to be ranked with respect to their sound quality. When clicking on the loudspeaker icon, the respective sound, e.g. an idling motor, is heard. The task of the subject is to shift the icons A through F in the empty fields denoted 1 through 6 in such a way
Proceedings of the Institute of Acoustics
that the sounds are ordered with respect to their sound quality. The subjects are free to listen to each sound as often as they like and to correct the sequence until a final status is reached. This large freedom of the subject, who has "random access" to the sounds to be ranked is one of the reasons that this procedure is nowadays preferred for ranking of sound quality.
1