Multimodal quality perception: the effects of attending to content on subjective quality ratings

Research on the quality measurement of communication systems has traditionally required subjects, either implicitly or explicitly, to focus attention on quality at the expense of content. However, information content is likely to be the primary focus of attention for the home or business user. Therefore, a potential problem for traditional quality testing procedures is that the reported quality ratings may not be representative of more natural viewing environments in which the focus of attention is on content rather than quality. This study required two groups of subjects to evaluate the overall transmission quality of audio-video sequences containing varying levels of degradation. Half the subjects were asked to assess quality only for each test sequence, while the remaining subjects were asked to recall the audio-video content in addition to a quality evaluation. The results found no difference in quality ratings between the groups, indicating that quality ratings are independent of content recall. This result shows that quality ratings are robust to a content recall task, suggesting that quality opinions obtained under controlled laboratory conditions may not be too dissimilar from opinions made under more natural viewing conditions.