Seedance 2.5 grey-peg test: chain-link fence run past a foreground arm (10 s)
Ten-second shot through a fence with a large foreground arm and shoulder; the runner crosses, stops, swings and exits at frame edge, consistent identity throughout.
Task: One take through a chain-link fence: a runner crosses a lawn, a foreground figure's raised arm dominates frame right, the runner reacts and exits.
Outcome: succeeded
0.63 s: Runner in grey tee crosses the lawn left to right behind the fence wire; brick building behind.
1.88 s: A large raised arm and shoulder of a figure in a black tank top enters frame right and stays in the foreground.
3.14 s: Runner has stopped and stands square to camera; fence mesh crosses the frame.
5.65 s: Face details change slightly (glasses-like shapes at this size); uncertain at 320 px.
6.90 s: Runner swings an arm toward the foreground figure.
8.16 s: Runner turns away and bends; the foreground arm has left frame.
9.41 s: Runner exits at the left screen edge; hedge and tree remain.
Intent vs result: The peg video (9.0 s vs 10.04 s output) shows a grey runner at the left crossing the lawn and a large grey foreground mass on the right from about 2 s; the output keeps that geography and replaces the mass with a person's arm and shoulder, retimed slightly longer.
Publication update (2026-10-29): the founder has now cleared public use of this output, including incorporated input likenesses and soundtrack. Earlier rights/likeness holds described above are resolved by that confirmation. Reference-input files remain private. Audio remains not listened to; visual observations cover sampled frames only.
Structured fields
{
"inputs": [
{
"description": "8 linked image input(s) recorded in the generation metadata (details withheld)."
},
{
"description": "1 linked video input(s) recorded in the generation metadata (details withheld)."
}
],
"run_at": "2026-09-25T09:11:34Z",
"metrics": [
{
"name": "output_duration",
"unit": "s",
"value": 10.042
},
{
"name": "resolution",
"value": "854x480"
},
{
"name": "frame_rate",
"value": "24/1"
},
{
"name": "scene_detector_transitions_thr0.30",
"value": 0
},
{
"name": "black_intervals",
"value": "none"
}
],
"procedure": "Seedance 2.5 text-and-reference video generation with a grey-peg motion video (a rendered blocking pre-visualisation that fixes camera path and positions), several approved scene stills and per-character face/front/back reference photos, plus a long structured prompt (look statement, reference-use rules, cast, camera, ordered choreography, continuity, sound). Historical founder test period using an older mannequin-based blocking pipeline. Sampled inspection: 8 timestamped frames from 0.63 s to 9.41 s of a 10.04 s file, plus ffmpeg scene detection at threshold 0.30 and blackdetect.",
"parameters": {
"seed": "unset",
"model": "seedance_2_5 (label in job metadata)",
"multi_shots": false,
"aspect_ratio": "16:9",
"audio_requested": true,
"resolution_label": "480p",
"requested_duration_s": 10
},
"limitations": [
"Sampled visual inspection from timestamped frames (contact sheets), not full real-time playback; motion smoothness, micro-timing and lip-sync are unrated.",
"Audio was not listened to; the file has an audio track but sound, dialogue and music are unassessed.",
"Prompt intent is taken from private job metadata and summarised here only in general terms; the founder's taste and intent for each test are unknown.",
"Single historical generation without a fixed seed; not a controlled benchmark and not validation of any current product or pipeline.",
"Face detail at 5.65 s is uncertain at the sampled resolution."
],
"artifact_ids": [
"ast_01M64VHC1141AM9ZE45DKY0T3W"
]
}Evidence claims
- [sourced_fact] Timestamped observations above come from viewing 8 sampled frames of the output (sampled visual inspection, not full playback).
- [upstream_claim] The generation was run by the founder in an earlier test period; H01 did not run or reproduce it.
- [inference] Foreground occluders defined in the peg video survive as real people; lawn, fence and exit direction track the pre-visualisation.
Rights
| component | license | hosting | redistribution | modification | training | Evidence and notes |
|---|---|---|---|---|---|---|
| documentation | unknown | permitted | permitted | unknown | unknown | Original analysis written by H01 (an AI agent) for Veyrul from private viewing; no third-party text, frames or audio reproduced. |
| media | unknown | permitted | permitted | unknown | unknown | Founder confirmed public use of all these Higgsfield outputs on 2026-10-29, including incorporated reference inputs, likenesses and soundtrack. Audio has not been listened to; permission does not establish listening or independent review. |