Video quality measurement is the practice of turning the question of how good a compressed or delivered video looks into a reproducible, scalable, comparable number, rather than an opinion that shifts with the viewer, the screen, and the room. It is done in two ways: subjective testing, where a panel of viewers rates clips under controlled conditions to yield a Mean Opinion Score, and objective metrics such as PSNR, SSIM, and VMAF, which compute a score in software in seconds. The key catch is that the target is perceived quality, not pixel arithmetic, so every automatic score is a proxy with blind spots, and a single pooled number can hide a bad scene that the average smooths away. A measurement program does four jobs: compare encoders, set quality targets, catch regressions, and prove delivered quality. It underpins related concepts such as QoE, the reference setups, and the objective-versus-subjective split.
The handbook · 7 volumes
Fora Soft Video Engineering Handbook
Video quality measurement replaces arguments about how a stream looks with numbers you can target and monitor. Our seven-volume handbook covers the whole video stack; for video quality measurement start with Volume 1 on quality metrics and compression, then Volume 3 on QoE monitoring and ABR and Volume 6 on AI models in video. Written by Nikolay Sapunov, CEO at Fora Soft.
Setting up video quality measurement so you can target quality instead of guessing?
We will review your pipeline and send a measurement plan with metrics, thresholds and where to run them, from encoder to player. 250+ video products shipped since 2005.
