Sora Video Archiveformat: 9:16 vertical, 10.0 s, photoreal phone-cam.
lens/sensor: 40mm ... by @subhamgupta
format: 9:16 vertical, 10.0 s, photoreal phone-cam.
lens/sensor: 40mm eq, shutter 1/90 s, ISO 1000, AWB 4800K.
capture feel: chest-height, front-facing, locked stance with micro-sway; handheld micro-jitter 2–3 px; rolling-shutter skew ~1.5° on fast edges; slight AF focus-hunt ±2% every 0.6 s; mild H.264 compression; fine sensor grain.
location: hotel lobby elevator, brushed steel doors closed at frame 0. Marble floor scuffs, directory LED panel on right, floor arrow decal ghosted, warm tungsten overhead.
one rule inversion: the mirror-steel reflection lags my real motion by 0.30 s (exact), then re-synchronizes.
action/beats:
- 0.00 s HOOK: I’m centered; reflection is 0.30 s behind.
- 0.20 s: I raise right hand; reflection hesitates then raises at 0.50 s.
- 0.90 s: doors begin to part; visible interior shows the same lobby but lit like noon daylight while our side stays tungsten.
- 1.90 s: elevator “ding” plays; doors nearly meet again.
physics: reflections obey correct perspective and brightness; reflection autonomy lasts exactly 2.00 s then snaps in-sync as door edges align; no secondary impossible effects.
micro-detail: steel hairline scratches, smudged fingerprints, marble veining, faint HVAC whoosh, distant roller-bag wheels, directory LED flicker at 50 Hz (visible as rolling banding on reflection only), tiny chromatic aberration on door edges.
audio: lobby bed + soft ding at 1.90 s; no music; phone-mic color.
loop engineering: end with doors almost closed, my hand back down, body posture matching frame 1; directory LED flicker phase and AF position identical first/last frame; copy prop positions so final frame = frame 1.
branding: tiny creator stamp “clipalchemy” bottom-left, 50% opacity, unobtrusive; keep unobstructed.
output: seamless loop; no captions burned in; keep exposure stable; no extra VFX beyond the single lag rule.
SoraStats is an independent archive and analytics site, not affiliated with OpenAI. Video availability reflects public archive data and may differ from source platform availability.
SoraStats interpretation
Archive analysis
This stable archive record combines playback availability with the latest stored public engagement snapshot, so the video can be compared with other Sora-era work even when source availability changes.
- Likes per view
- 6.5%
- Remixes per view
- 0.0%
Ratios are descriptive, not quality scores. Counts may lag the source platform, and deleted or private activity is outside archive coverage.