Discussions

I produce short vertical product ads (9:16, usually 15-20s) and use Genny for the voiceover. My current order is: generate the visual first, then write the script to fit the cut, then record VO in Genny. Two things keep tripping me up. First, when the visual is locked at 15s and the generated VO lands at 17s, is it better to rewrite the script shorter or to adjust the speaking rate? Does pushing the rate noticeably degrade quality on the newer voices, or is there headroom? Second, for muted autoplay on paid social I need burned-in captions. Is there any way to export a timing or transcript file from Genny that I can use for caption timing, or do people just re-time by hand in the editor? Context on why the timing is fixed: I generate the visuals from a single product photo with [Nereo](https://www.nereo.io/), so the cut length is decided before the VO exists. Curious how others sequence this - do you write the script first and cut the visual to it instead?