The effects of quantity and location of video playback errors on the average end-user’s experience

The end-user experience of a given platform or product with respect to video playback is an increasingly important aspect for engineers and product planners to understand. However, testing real world situations in an objective and repeatable fashion is complex. In an initial study, the video Gross Error Detector (GED) which is a tool that provides a quick and cost-efficient way to evaluate video playback, was mapped to end-user´s perception of video smoothness. Video errors such as dropped, repeated or out of sequence frames were evaluated in varying quantities to understand the impact on the average en-user´s perception of playback. A second study, discussed here, was done to understand the differences in the quantity and location of these playback errors. These end-user perception studies provide useful data so the video GED can be used to monitor errors in video for quality control, benchmark video processing and algorithms, and can be rooted into the video processing system to optimize algorithms and limit settings.