The cut is the test

A research preview points towards a familiar production problem.

ZERVUE Editorial · News analysis · AI SHIFT ·

A change must survive the sequence

Adobe’s November 11 research preview explores edits that persist across cuts. That makes a useful starting point for an editorial question: how should a small production team decide whether a new AI tool belongs in its workflow? A persuasive demonstration is a reason to investigate, rather than a delivery promise. [1]

Choose a difficult sequence

Our recommendation is to start with an owned or rights-cleared sequence containing several views of the same subject. Set one modest change, then inspect the joins, lighting and continuity. Write down which frames need repair and whether the result can still be revised. This is a proposed evaluation method; ZERVUE has not performed this test.

Price the whole revision

Include selection, correction, export and client changes when estimating the work. Keep the source sequence and an accepted reference so a later tool update can be compared fairly. A research preview does not guarantee a released feature, access or production reliability. The useful decision is whether the whole sequence survives the revision, not whether one frame looks impressive.