Visual assessment of HDR video

Abstract

The future of Multimedia can be depicted in a scenario where almost all the fruition of audiovisual content will be delivered to the final user by means of streaming over any kind of transmission media. This migration towards streaming will change drastically most of today's scenarios, ranging from TV receivers down to any kind of mobile terminals, including 5G devices. On the other side the content itself will be of better quality; in fact it will be in HD and UHD format, it will be scalable and it will also use a Wider Color Gamut (WCG) and High Dynamic Range (HDR), to improve the User Experience. Furthermore the users will be able to access the Broadcasting services (i.e. TV) also through the new generation 4G/5G networks. This paper deals with the formal subjective assessment evaluation of WCG and HDR video content, and it is located in the framework of the standardization activities of MPEG (ISO-JEC/SC29/WG11) that, together with ITU-T (through the joint committee JCTVC) is active in the definition of new technologies for WCG and HDR video coding. Here the results of the comparison between a new test method designed to evaluate HDR video content, i.e. the HDR-Side By Side (HDRSBS), and a classic subjective video test method, the DCR (Degradation Category Rating), are described. The new HDR-SBS test method, was designed to evaluate HDR and WCG video material produces by the Submissions made in response to the Call for Evidence (CfE) recently issued by MPEG [1]; the HDR-SBS method is basically a category rating method based on a side-byside approach. The paper demonstrates that, despite the effort made by MPEG to design a new test method devoted to the assessment of HDR content, and despite what suggested in other papers describing similar tests [2] [3] [6], the DCR method shows better performances. Furthermore the DCR test method results to be less stressing for the human subjects participating to a formal subjective assessment experiment.