While VMAF (Video Multimethod Assessment Fusion)—developed in part by Netflix—is seen by many as the king of the video quality metrics, use cases for live are more challenging. This prompted Synamedia to create a metric very close to the de-facto standard, dubbed predictive VMAF (pVMAF), which it has subsequently integrated into its vDCM encoder software. First introduced at Mile High Video in 2023, pVMAF is Synamedia’s machine learning-based answer to the complexity scale problems posed with VMAF, which relies on calculating multiple metrics like VIF (variance inflation factor) at different scales and combining them into one score. This works well for VoD applications, where many CPU cycles can be allowed to make the calculation. In the live space, that is…