“A group of close friends embarks on an impulsive winter escape to Japan, finding warmth in shared meals and laughter amidst the icy exhilaration of the ski slopes.”
Reelly · the gate
Same footage, same template, same music, same 47 cut points — the template’s durations were lifted from your own edit, so all three share its rhythm exactly. The only thing that differs is which moment landed in each cut. Watch the first one. Post it, or don’t — deciding not to counts, and is just as informative.
111.4s · 46 cuts · 32 distinct clips · 0 repairs, 0 fallbacks
“A group of close friends embarks on an impulsive winter escape to Japan, finding warmth in shared meals and laughter amidst the icy exhilaration of the ski slopes.”
The control. Sorted by capture time, entering every clip at zero.
The original. Every slot duration in the template was extracted from this edit — so all three videos share its cut rhythm exactly.
Worth knowing while watching: this one has things the renderer does not yet do. The template preserves three text overlays, per-clip pan and scale keyframes, a timeline fade-out and the music fade — none of which are applied yet. Some of the gap you see is missing features, not missing judgement.
| Version | Framing | Interest | Variety | Story | Total |
|---|---|---|---|---|---|
| Pipeline | 7.8 | 8.4 | 7.6 | 8.0 | 31.8 |
| Your own edit | 7.2 | 7.4 | 7.8 | 7.0 | 29.4 |
| Head-trim baseline | 5.4 | 4.4 | 6.0 | 3.4 | 19.2 |
Judged from contact sheets sampled at identical timecodes, with presentation order randomised across five runs. The pipeline beat head-trim 5/5 and your own edit 3/5 — which is a coin flip, so read it as indistinguishable from a human edit, not better than one. A single run had it winning outright; reversing the order flipped the answer, which is why this is an average. It also judges still frames, so your edit’s timing and motion are invisible to it.