---
name: ad-spot
description: "A 15 s product commercial — 3 GPT Image 2.5 keyframes with the product (macro hero, the action, the end card with the slogan text), a historical ElevenLabs Music bed (120 BPM, a drop), six shots on the beat; MiniMax H3 Max reference-to-video 480P (9 s, cuts up to 2 s off in this test) for drafts, Seedance 2.5 reference-to-video 720p (7 min, cut deviations up to about 1 s, its own synced sound effects) for the source's preferred final. Part of the ad reference set."
---

These creative references describe source experiments with external tools. Use your own supported account and tools; this download installs no automation. Settings and timings are experimental observations, not current FlashBoards availability. Use only media and voices you have permission to use.

# ad-spot · the commercial

> **Alternative external route:** Seedance 2.5 reference-to-video (`bytedance/seedance-2.5/reference-to-video`, references as @Image1… / @Audio1…, 4–30 s, native sound) was preferred in the source experiments. The measured side-by-side comparison was a product commercial, not every format; compare it with the H3 method below for your material.

> **Source author's sound preference (9.10.2026), not FlashBoards policy:** avoid ElevenLabs Sound Effects (`/v1/sound-generation`, `fal-ai/elevenlabs/sound-effects`) for effects, room tone, ambience and jingles; the author found them strange and louder than the voices and preferred ElevenLabs for voices (TTS v4, Speech-to-Speech, Audio Isolation). Historical demonstrations below and in related guides still include ElevenLabs Music beds and an earlier Sound Effects ambience test; those are separate experiments, not a universal voices-only rule. For your own mix, use permissioned film sound (line-free gaps, Demucs without vocals), the video model's native sound or supported video-to-audio tools; keep effects under the voice.

1. **Keyframes** (GPT Image 2.5 Flare edit, medium, 1920×1088, picture 1 = the product sheet): the macro hero shot (dark studio, wet glass, rim light), the action (pour over ice), **the end card with the slogan text** — the text is drawn here, not in the video.
2. **Music (historical experiment)**: ElevenLabs `POST /v1/music` (`music_length_ms: 15000`) — "dark minimal electronic, 120 BPM, a riser, one big bass drop"; measure the beat phase and the drop, put the cuts on bars. This Music bed is distinct from the Sound Effects preference above; for your own ad, use music you have permission to use.
3. **Six shots**: macro droplet → the can cracks open → the pour → an orbit on the drop → a splash into ice water → the end card holds.
4. **H3 Max** (`minimax/h3-max/reference-to-video`, 16:9, 480P, 15 s): Image 1 product sheet, Image 2–4 keyframes ("shot 1 starts exactly like it", "shot 6 is exactly this frame, with exactly this text"), Audio 1 music. 9 s of generation. Cuts landed 2.4 / 4.0 / 6.0 / 10.7 / 12.0 instead of 2.5 / 4.5 / 8.0 / 10.0 / 12.0. Audio played as given → lay the music back and add sound effects from licensed footage or a supported video-to-audio tool where the actions happen; do not use ElevenLabs sound effects, following the source preference above.
5. **Seedance 2.5** (`bytedance/seedance-2.5/reference-to-video`, `task: reference`, 16:9, 720p, `duration: "15"`, `generate_audio: true`): `image_urls` the same four, `audio_urls` the music, refer to them as **@Image1…@Image4, @Audio1**; plain text prompt, one paragraph per shot with its seconds. 428 s of generation. Cuts 2.4 / 4.75 / ~7 / 10.0 / 12.5; finer liquids and ice; the sound is a re-composed version of the music (corr 0.4–0.57) with its own hiss, pour and splash in sync — kept as is.

The largest listed cut deviation is 2 s on H3 (6.0 versus 8.0) and about 1 s on Seedance (~7 versus 8.0; the final cut was 12.5 versus 12.0). These measured values take precedence over the source synopsis's smaller bounds; they are not guarantees for a new request. No example keyframes, music or clips are delivered or required.