--- name: avatar-walk-talk description: "The avatar walks while talking — room tour, show-around, walk and talk — on the avatar-long-speech chain: a selfie start frame with the whole place visible, a route split by the voice pieces (one stop per piece, gestures tied to words), the handheld-selfie prompt. Part of the avatar 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. # avatar-walk-talk · room tour, show-around Inputs are the person's own: {look sheet} or {picture}, {speech} with their {voice}, {place}, {route}. ## Steps 1. **Start frame** — every stop of {route} visible behind the character, so extend keeps one place: `openai/gpt-image-2.5/flare/edit`, quality low, 1152×2048 (32 s): ```text Picture 1 is a character sheet of one . A vertical 9:16 live-action selfie frame shot with a phone's front camera held at arm's length; the phone itself is never seen. The same stands of <{place}>, the place visible behind : . A medium close-up from chest up, slightly high angle, look(s) straight into the lens with a smile, about to speak. face, hair and clothes exactly as in picture 1. Natural phone camera look, . ``` 2. **{speech}** — written along {route}: one stop per ~10 s, a line about each thing. Then split the speech at pauses (see `avatar-long-speech`) — **assign the stops to the pieces by their transcribed words**, not by plan. 3. **Piece prompt** — for every piece: ```text A vertical 9:16 selfie : the films with the phone's front camera held at arm's length in one hand; the phone itself is never seen. Audio 1 is voice: say(s) exactly these words, lips exactly in sync with Audio 1: "" walk(s) and what do(es) on which word>. The camera moves with as walk(s), the framing bobs and sways a little with steps like a real handheld phone, and the place passes behind . face, hair and clothes stay exactly as in the video, and it is the same as in the video: . . One continuous shot, no cuts, no music, no text on screen. ``` - `` example shape: " walk(s) over to , the camera following face, and on '' tap(s) / point(s) at / hold(s) ." — one physical action per thing, a laugh / sigh / whisper where the speech has it. - Repeat the same list of things in every piece: the only memory of the place extend has is the clip so far. 4. Generate the stub and continuations with the settings in `avatar-long-speech`, assemble the new footage, and remove the stub. ## Measured (9.10.2026, one minute) - 56.6 s speech, 5 pieces 10.4–12.2 s → 60.3 s clip, 26–42 s per piece, ≈ 4 min in all. - Speech is less faithful than standing: correlation median 0.67, ≥ 0.9 in 5 of 19 three-second windows (steps / room sound mixed in or re-voiced — not separated); drift up to +2.9 s. **Listen to the whole clip before publishing.** ## Example run (reference only — never reuse its files, prompts, voices or images for a person) The route and filenames below describe the historical experiment, not delivered or required material. Use your own place, start frame, permissioned speech, per-piece prompts and outputs; no local room-tour runner is supplied. - Demo character Mara Vale, her small bedroom at night: start frame from the worked example (door, unmade bed with fairy lights and photos, desk with a warm lamp, big leafy plant, window with city lights, clothes rack). - Route: door → bed → fairy lights → desk lamp ("five bucks") → plant "Gerald" (a salute on "Respect") → clothes rack (lifts a sleeve: the same jacket she wears) → window ("the whole city just glows"), a wave on "See you tomorrow". - Speech `room_line.mp3` + `room_line.parts.json`, steps `room/stub.mp4 … ext5.mp4`, result `room.mp4` + `room.json`.