El podcast no existe.El prompt sí.

SEEDANCE 2.5
DURATION: 17 SECONDS
ASPECT RATIO: 16:9

REFERENCE ASSIGNMENT:
@video1 = the authoritative performance. Lock the host's exact speech, timing,
mouth movement, head motion, hand gestures and body language to this video,
frame for frame. Nothing about his delivery changes.
@image1 = exact identity of the HOST: face, curly dark hair, build, black
crew-neck t-shirt, forearm tattoo.
@image2 = the authoritative SET and the exact identity of the INTERVIEWER, plus
the authoritative wide two-shot composition for SHOT 01 and SHOT 05.

SET (from @image2):
White seamless studio, white walls meeting at a soft corner behind the table.
Long light wood table running across the lower frame. Two black boom-arm
broadcast microphones angled in from either side. A small black audio mixer with
coloured pads at the centre of the table, a laptop lying flat beside it, black
over-ear headphones resting on the table in front of each man, a cable running
across the surface. Black mesh Eames-style office chairs. This set is identical
in every single shot.

INTERVIEWER (from @image2):
Man in his early fifties, salt-and-pepper hair swept back, short greying beard,
no glasses, charcoal grey knit sweater. He sits at the LEFT side of the table,
in profile facing right toward the host. This is an invented character — he must
not resemble any real or recognizable public figure.

HE NEVER SPEAKS. His mouth stays closed for the entire clip. He is listening.

HOST:
The man from @image1 and @image2 — black t-shirt, curly dark hair. He sits at the
RIGHT side of the table, facing left toward the interviewer. He is the only
person who speaks.

WORLD LOCK:
Bright white podcast studio, soft even diffused key light with gentle falloff,
clean natural skin tones, no colour cast, no cinematic grade. Photorealistic:
real skin pores and texture, visible knit weave on the sweater, cotton texture on
the t-shirt, realistic specular highlights on the microphone bodies and the wood
grain of the table. Modern professional podcast production look.

CAMERA LANGUAGE:
This must read as a genuine multi-camera podcast edit, not as a moving AI shot.
Every angle is a locked-off tripod camera with only the faintest breathing drift
— no handheld shake, no push-ins, no orbits, no zooms, no dolly. All energy comes
from HARD CUTTING between static cameras, like a real podcast vision switcher.
35mm on the wide, 85mm on the singles with shallow depth of field.

SHOT 01 | 00:00 - 00:03
WIDE TWO-SHOT, exactly the composition, camera height and framing of @image2 —
symmetrical side-on view, both men in profile, microphones and mixer between
them, table across the lower frame. The host is ALREADY mid-sentence as the clip
opens — no silence, no settling in, no establishing beat. The interviewer sits
still, hands loosely clasped on the table, head angled toward the host.

SHOT 02 | 00:03 - 00:08
HARD CUT. MEDIUM SINGLE on the HOST, 85mm, locked tripod, framed from mid-chest
up, his microphone entering frame from the left, white wall softly out of focus
behind him. Three-quarter angle, speaking toward the interviewer off-frame left.
His hands move naturally as he talks, exactly as in @video1.

SHOT 03 | 00:08 - 00:10.5
HARD CUT. MEDIUM SINGLE on the INTERVIEWER, 85mm, locked tripod, framed from
mid-chest up, his microphone entering frame from the right. He listens: one slow
blink, a single small nod, eyes steady on the host off-frame right. Mouth closed
and still. Restrained, genuine attention — nothing theatrical. The host's voice
continues uninterrupted over this shot.

SHOT 04 | 00:10.5 - 00:15
HARD CUT. TIGHTER SINGLE on the HOST, 85mm — head and shoulders, closer than
SHOT 02, microphone in the lower left of frame. This is the point of the clip.
Hold here without cutting away until his final word is completely finished. Mouth
articulation perfectly synchronized to the audio of @video1.

SHOT 05 | 00:15 - 00:17
HARD CUT back to the WIDE TWO-SHOT — identical camera position and framing to
SHOT 01 and @image2. The host has finished speaking and settles slightly back
into his chair. The interviewer gives one small final nod. Two seconds of held
silence, room tone only. No dialogue.

CONTINUITY:
The interviewer stays SCREEN LEFT and the host stays SCREEN RIGHT in every angle
— their singles must respect this, the host always facing screen left, the
interviewer always facing screen right. Both men keep the same clothes, same
hair, same faces and the same seating positions throughout. Microphones,
headphones, mixer, laptop, cable and chairs never move between cuts. Every shot
is the same room, same light.

AUDIO:
Only the host's voice, taken exactly from @video1 — same voice, same language,
same intonation, same pacing. No second voice. No voice-over. No narrator. No
music, no score, no sound design. Only quiet studio room tone underneath.

NEGATIVE:
no camera movement, no handheld shake, no zoom, no push-in, no dolly, no orbit,
no slow motion, no speed ramps, no subtitles, no text, no captions, no on-screen
graphics, no logos, no watermarks, no split screen, no music, no second voice, no
interviewer dialogue, no interviewer mouth movement, no extra people, no windows,
no coloured lighting, no lens flare, no cinematic teal-orange grade, no beauty
smoothing, no glasses on the interviewer, no identity drift between shots, no
wardrobe changes, no crossing the line — screen directions never flip.