QoE in 10 seconds: Are short video clip lengths sufficient for Quality of Experience assessment?

Abstract

Standard methodologies for subjective video quality testing are based on very short test clips of 10 seconds. But is this duration sufficient for Quality of Experience assessment? In this paper, we present the results of a comparative user study that tests whether quality perception and rating behavior may be different if video clip durations are longer. We did not find strong overall MOS differences between clip durations, but the three longer clips (60, 120 and 240 seconds) were rated slightly more positively than the three shorter durations under comparison (10, 15 and 30 seconds). This difference was most apparent when high quality videos were presented. However, we did not find an interaction between content class and the duration effect itself. Furthermore, methodological implications of these results are discussed.