Table 7 Performance for HD, for the training () and evaluation () data; audio and video quality and impairment factors are derived from subjective tests (Subj.) or predicted from audio and video quality models (Pred.); : content-blind Q-based model, : content-aware Q-based model, : content-blind IF-based model, : content-aware IF-based model; in italic: the respective model performs significantly better than the corresponding basic model ; in bold: significantly better performing model between and .