h3-layered-windsurfing-fashion-mv

H3 Layered Windsurfing Fashion MV

Give every reference one authority and build each 15-second segment around a musical energy curve.

Route the Request

• For prompt writing or refinement, return text only.
• Read references/style-blueprint.md and use $h3-prompt-writing.
• Invoke $h3-video-generation only when the user explicitly asks to render a named segment. Never submit all three segments implicitly.
• The full design is three separate 15-second jobs. Use native Ref2VA for identity/equipment references in each job, preserve the same ordered references and prompt labels across all three, then assemble the approved outputs; the Broker still does not expose a single multishot job.

Assign Reference Authority

1. Picture 1 is the key visual and opening-state authority: adult identity, face, hair, costume, palette, and first frame.
2. Picture 2 is the turnaround and equipment authority: body proportions, board, sail, mast, boom, rigging, fasteners, and color relationships. It is not a storyboard or camera guide.
3. Picture 3 is the segment-specific cinematography guide: framing, lens, movement, subject scale, and composition only. It cannot change identity, costume, equipment, or environment logic.
4. For the climax segment, omit Picture 3 and direct the giant wave, takeoff, airborne hero hold, landing, and fashion finish entirely in text.

Shape Each Segment

• 00:00-00:03: build anticipation.
• 00:03: musical drop and decisive action onset.
• 00:03-00:10: sustained high energy.
• 00:06-00:07: optional compact turnaround that proves equipment integrity.
• 00:09-00:10: absolute visual peak with a stable hero camera moment.
• 00:10-00:12: release.
• 00:12-00:15: fashion ending or bridge to the next phrase.

Keep water physics, rig tension, hand placement, wind direction, and board contact coherent. Use a trained adult athlete, controlled fictional conditions, and no injury.

Render Rules

• Default orientation for the full fashion campaign is 9:16, one 15-second segment at a time. The modal demo is a 10-second 16:9 T2VA style preview.
• Default to 864x480; allow 960x544 when requested. Explicit 768p means 1344x768 and may never be downgraded. For vertical delivery, use the Broker-supported vertical equivalent only if available and explicitly requested.
• Exclude adult content, nudity, sexualized framing, minors, celebrity likeness, named characters, real brands, logos, unsafe weather, injury, broken rigging, changing board geometry, extra limbs, unreadable text, subtitles, and watermarks.