提示词 / 授权来源
Prompt 535 · @tanabe_fragm
@tanabe_fragm · 文本生成视频
24歳・スポーティな女性インフルエンサーの1日 Vlog(日本の8月)
No reference material. Text-to-video.
■ BEFORE YOU GENERATE (set these in the Dreamina UI)
Aspect ratio : 9:16 (vertical)
Resolution : 720p
Duration : 25s
Frame rate : 24fps
■ NOT IN THIS PROMPT — added later in post
Japanese captions (burned in with ffmpeg)
Background music at 120 BPM (mixed in with ffmpeg)
See setup.md.
【SUMMARY】
25 seconds, vertical 9:16, 24fps. A fast-cut day-in-the-life vlog that a 24-year-old
Japanese fitness influencer is filming of herself on her own phone, in AUGUST IN JAPAN: a
morning smoothie, a sunrise run along a river, and an afternoon out with friends. She is
166cm with a lean model's build. Bright, hot, humid, alive, and QUICK — ten short shots,
none of them lingering, cut on the beat. The film is about how good her body feels to move.
She is the only face in the film.
IT OPENS ON HER FACE, LARGE AND CLOSE ON A PILLOW, WAKING AND FINDING THE LENS.
IT ENDS ON HER RIGHT HAND REACHING OUT AND TOUCHING THE LENS.
【PACE — the governing rule of this film】
This is a fast vlog, not a slow one. TEN shots in twenty-five seconds.
Every shot begins with its action ALREADY UNDERWAY — nobody walks in, settles, or gets
ready on camera. The first frame of each shot is already the middle of the movement.
Every shot ends the instant its action completes. No shot holds on a settled pose, and no
shot has dead air at either end.
Cuts land on the beat of a 120 BPM track, so every cut falls exactly on a whole second.
Two of the ten shots are ONE SECOND LONG — tight inserts with no face in them, which
snap the rhythm and give her face a rest between close-ups.
【DIALOGUE — the complete spoken content of this film】
SHE SPEAKS EXACTLY FOUR TIMES IN TWENTY-FIVE SECONDS AND AT NO OTHER MOMENT. These four
lines are the only words spoken in the film by anyone:
at 3.0s 「おはよう」
at 16.2s 「5キロ!」
at 17.1s 「かんぱーい!」
at 22.6s 「またね」
Together they take under three seconds. FOR THE OTHER TWENTY-TWO SECONDS HER LIPS ARE CLOSED
and the only sounds she makes are breathing and wordless laughter.
She improvises nothing. She adds no extra greeting, no filler, no muttering, no commentary
on what she is doing, no sign-off, and no words of thanks to the viewer.
LAUGHTER IS A SOUND, NOT A SENTENCE — when she laughs her mouth opens and air comes out,
but no words form and her tongue and lips make no consonants.
Shot by shot: shot 1 has one line and is otherwise silent. Shots 2, 3, 4, 5 and 6 are
completely wordless with her lips closed. Shot 7 has one line. Shot 8 has one line in its
first half second and then only laughter. Shot 9 is completely wordless with her lips
closed. Shot 10 has one line and is silent after it.
Her mouth is closed and still in the last two seconds of the film, from 23.0s to 25.0s.
【GLOBAL SETUP】
The conceit — hold this for all 25 seconds:
The camera is HER OWN PHONE. Every shot is footage she shot herself for her channel. It is
carried only three ways:
SET DOWN — she has propped the phone on a stand, a step, a table or the ground and moves
within a COMPLETELY LOCKED frame. It never tracks, pushes, pans or drifts, because
nobody is holding it. Both her hands are free.
SELFIE — she holds the phone in one hand, arm visible at the edge of frame, talking into
the lens. ONE FREE HAND ONLY, doing one thing at a time. Any change of framing is her
arm extending or bending, never more than an arm can reach.
FRIEND — ONLY in the two afternoon shots where her friends are present, a friend holds
the phone. Both her hands are free, the frame sways naturally, and it may walk.
She is alone all morning, so the morning is set-down shots and one selfie only.
Shot by shot: 1 set down, 2 set down, 3 set down, 4 set down, 5 set down, 6 set down,
7 selfie, 8 friend, 9 friend, 10 set down.
The recording is ALREADY RUNNING when the film begins — no shot ever shows her starting one.
She touches the phone twice: to tilt it at 4.0s, and to stop it at 25.0s.
Season and weather — AUGUST IN JAPAN:
Deep summer. Hard bright sunlight with sharply defined shadows. Towering white cumulus
clouds stacked high in a saturated blue sky. Humid air. Cicadas singing in every outdoor
shot — thin and distant in the morning, thick and loud in the afternoon. Sweat on her skin
and heavy condensation on every cold glass.
Visual style:
Photorealistic modern vlog footage, bright and clean. Natural light in every location: low
strong sunrise light in the morning half, high hard afternoon light in the second half,
low warm orange light at the end. Shallow depth of field, roughly f/2.0. Emphasize extremely
realistic physical texture: skin, sweat, linen weave, cotton rib knit, technical lycra,
condensation running on glass.
THE ONLY PERSON — a 24-year-old Japanese woman.
Warm honey-beige skin, lightly sun-tanned, with a faint dusting of freckles across the
bridge of her nose. Keep the real fine pores and the fine downy hair on her skin.
Almond-shaped eyes with a clear double lid, a straight fine nose bridge, full lips, high
cheekbones, and a sharp defined jawline. Her gaze is direct and strong; when she laughs her
outer eye corners drop and her whole face joins in.
Height 166cm with a lean athletic model's build: long arms, long legs, broad-ish shoulders,
a narrow waist with the faint line of abdominal muscle, and visible definition in her calves
and shoulders. Her proportions read as a runner, not as a bodybuilder.
Her hair is dark brown, glossy, and long enough to reach below her shoulder blades. The
COLOUR, LENGTH and TEXTURE of her hair are identical in every shot; only the way she has
tied it changes, and it changes exactly three times:
HAIR DOWN, SLEEP-TOUSLED — loose, unbrushed, flattened on one side. Shots 1 and 2.
TIGHT HIGH PONYTAIL — gathered high and pulled tight, swinging free below her shoulder
blades, loose strands stuck to her temples by sweat. Shots 3 to 7.
HAIR DOWN IN LOOSE WAVES — washed and dried after the run, soft waves through the lower
half, pushed back behind one ear. Shots 8, 9 and 10.
IDENTITY ANCHORS — identical in all ten shots, never changing:
1. Her face: almond eyes with a clear double lid, straight fine nose bridge, full lips,
high cheekbones, sharp jawline.
2. The faint dusting of freckles across the bridge of her nose.
3. A fine plain gold chain necklace, worn against her skin in every shot, morning to
night. She never takes it off.
4. A thin black hair elastic. When her hair is tied it is in her hair; when her hair is
down it is around her right wrist. It is always somewhere on her body.
Because her hairstyle changes during the day, these four anchors carry her identity across
the cuts. Render them the same way every time.
Wardrobe — light summer clothing a Japanese woman actually wears in August. Every garment is
a single solid colour with no pattern, no logo and no lettering:
morning : a thin white ribbed camisole, light grey cotton shorts, bare feet, and an
oversized white linen shirt worn open with one shoulder slipping down
running : black sports bra, short black running shorts, white ankle socks, white
running shoes
afternoon: white ribbed tank top, a beige linen maxi skirt that moves in the air, white
sneakers
The skin she shows is the skin you show because it is hot. It reads as a real August
morning, relaxed and unposed.
Other people:
Two friends appear in the afternoon. Their faces stay outside the frame — out-of-focus
shoulders at the edge of frame and hands holding iced-latte glasses. She is the only person
whose face the camera ever sees.
Performance core:
A good ordinary summer day, moving fast. Bright in the morning, focused and strong on the
run, loud and delighted in the afternoon. She is comfortable with the camera because it is
hers.
Micro-life rule:
Something on her moves every second — her chest rises, her nostrils flare on a breath, her
brow lifts and releases, a strand of hair shifts, a bead of sweat slides.
Hold across the whole film:
The picture carries no text of any kind — no captions, no titles, no watermarks. Every
sign, poster, label and package in the background stays soft and out of focus so that no
lettering is legible anywhere.
The soundtrack is her voice and the true ambience of each location only, with cicadas
under every outdoor shot. Sound effects only. No music.
She speaks only the words inside quotation marks below, and stays silent the rest of the
time. The friends never speak; they laugh.
Exactly one person's face in the film — hers. Exactly three outfits, in the order above.
Every drink she takes is clean and contained — her lips close on the rim, the liquid goes
into her mouth, and her chin and throat stay dry.
【TIMELINE — ten shots, every cut on a whole second】
── MORNING ──
0.0 - 4.0s 【wake / SET DOWN / 4.0s】
Setup: the phone stands on the bedside table beside her pillow. The frame is locked.
Character: the woman described above, bare-faced, HAIR DOWN AND SLEEP-TOUSLED, thin white
ribbed camisole, gold chain, black elastic on her right wrist.
Action: THE FIRST FRAME IS HER FACE, LARGE AND CLOSE ON A WHITE PILLOW, filling most of the
vertical frame — the camera is only a forearm away. Individual eyelashes, the freckles on
her nose, the downy hair along her jaw, the fine pores. A hard bar of morning summer light
lies across one cheek. Her eyes open immediately — one slow blink, two quick ones — find
the lens, and a smile starts.
At 1.2s she pushes up onto one elbow and sits up in one quick movement, so her face lifts
away and the white bed, the window and the curtains come into view. She pulls the loose
white linen shirt on over the camisole without fastening it, one shoulder slipping down,
AND RAISES BOTH ARMS STRAIGHT OVERHEAD INTO A LONG STRETCH in the same motion — both hands
free because nobody is holding the phone — back arching, neck lengthening, ribs lifting,
fingers spreading. She drops her arms.
At 2.6s SHE LEANS IN TOWARD THE CAMERA. Her face grows until it fills the frame again and
she speaks straight into the lens. The word lands, her smile opens all the way, and the
shot cuts on that smile.
Dialogue (Japanese, bright and warm, at about 3.0s):「おはよう」
Speech: that one word is everything she says in this shot. From the first frame to 2.9s her
lips are closed and she only breathes.
Camera: SET DOWN. COMPLETELY LOCKED for all 4.0 seconds — no move, tilt, pan or push.
Everything that changes is her moving closer to the lens or further from it.
Audio: cicadas starting up thin and far away, the rustle of linen, her voice. Sound effects
only. No music.
4.0 - 6.0s 【smoothie / SET DOWN / 2.0s】
Character: the same woman, HAIR STILL DOWN AND UNBRUSHED, same white camisole and open white
linen shirt, same gold chain, black elastic on her right wrist.
Action: HER HAND IS ALREADY ON THE PHONE IN THE FIRST FRAME — she grips its top edge, tips
the angle down and lets go, and the whole frame tilts once and settles. The recording was
already running; she is only fixing the framing. She steps straight back behind a kitchen
counter where a tall glass of thick deep-pink smoothie stands, its outside beaded with
condensation. She lifts it in both hands and drinks without pausing. Her throat moves as
she swallows. She lowers the glass, a faint pink line on her upper lip, licks it away and
grins at the lens. Cut on the grin.
Camera: SET DOWN. The frame tilts once at the very start as her hand adjusts it, then is
completely static and does not move again.
Light: hard white August morning light through a kitchen window, throwing a sharp-edged
shadow of the glass across the counter.
Speech: SHE SAYS NOTHING IN THIS SHOT. Her lips stay closed except to drink and to grin.
Audio: the scrape of the phone on its stand, ice in the glass, the base of the glass meeting
stone, cicadas outside. Sound effects only. No music.
── THE RUN ──
6.0 - 7.0s 【INSERT — she ties her hair / SET DOWN / 1.0 SECOND, no face】
Character: her hands, forearms and the back of her head only. Her face is not in this shot.
Action: TIGHT CLOSE-UP, ALREADY IN MOTION IN THE FIRST FRAME. The black elastic comes off
her right wrist, both hands sweep her hair up off her neck, and she binds it into a tight
high ponytail — elbows lifting wide, the nape of her neck exposed, the ponytail dropping
and swinging once as she lets go. One second, one action, then cut.
Camera: SET DOWN. Tight and completely locked.
Speech: SHE SAYS NOTHING IN THIS SHOT. Her face is not even in it.
Audio: the snap of the elastic, cicadas. Sound effects only. No music.
7.0 - 9.0s 【stretch / SET DOWN / 2.0s】
Character: the same woman, TIGHT HIGH PONYTAIL, black sports bra, short black running shorts,
white ankle socks, white running shoes, same gold chain.
Action: SHE IS ALREADY FOLDED FORWARD IN THE FIRST FRAME — one heel planted on a step, her
body folded over the straight leg, both hands past her toes. The long line of the leg runs
almost the full height of the frame. She rises fast, pushes an earbud into her ear with one
finger, and is already turning away to go as the shot cuts.
Camera: SET DOWN. Static low angle from near the ground, full body, so her legs read long.
It does not move.
Light: early hard sun, long sharp shadows on concrete, heat shimmer in the air.
Speech: SHE SAYS NOTHING IN THIS SHOT. Her lips stay closed; she only breathes.
Audio: cicadas, the click of the earbud, a car far away. Sound effects only. No music.
9.0 - 14.0s 【THE RUN — the centrepiece, five seconds / SET DOWN】
Setup: she has put the phone down on the path ahead of her and run back up it, so the shot
opens on a riverside path lined with trees with her already running toward the lens.
Character: the same woman, same tight high ponytail, same black sports bra and shorts.
Action: SHE IS ALREADY AT FULL STRIDE IN THE VERY FIRST FRAME, mid-step, in the middle
distance — she never starts from standing and she is never far away for long. She runs
straight down the path toward the camera and grows quickly in the frame. Her whole body
stays visible: her ponytail lifting and dropping with every step, her arms swinging in
clean rhythm, her calves tightening on each landing, her chin level. As she gets close the
sweat at her temple and along her collarbone catches the sun. She meets the lens, grins
without breaking stride, and passes the camera closely on one side and out of frame. Cut
the moment she leaves the frame — hold no empty path at the end.
Camera: SET DOWN, low on the path and locked. It does not move, track, pan or push at any
point. She is the only thing that moves, from middle distance to right past the lens.
Light: low sunrise sun behind her, rim-lighting her outline in white. Huge white cumulus
clouds stand up in a deep blue sky. Tree shadows cross her body at a steady rhythm.
Speech: SHE SAYS NOTHING IN THIS SHOT. She is running; her mouth is open only to breathe and
to grin, and no words come out.
Audio: her footfalls on asphalt getting louder and closer, her steady breathing, the river,
cicadas. Sound effects only. No music.
14.0 - 15.0s 【INSERT — her feet / SET DOWN / 1.0 SECOND, no face】
Character: her lower legs and shoes only. Her face is not in this shot.
Action: GROUND-LEVEL CLOSE-UP, ALREADY IN MOTION. Two white running shoes strike the sunlit
asphalt and push off, one then the other, calves flexing, a fine spray of dust lifting off
the ground. One second of pure movement, then cut.
Camera: SET DOWN, on the ground and completely locked.
Speech: SHE SAYS NOTHING IN THIS SHOT. Her face is not even in it.
Audio: two hard footfalls, close and loud, cicadas. Sound effects only. No music.
15.0 - 17.0s 【water / SELFIE / 2.0s】
Character: the same woman, SAME TIGHT HIGH PONYTAIL now damp with sweat and loose strands
stuck to her temples, same black sports bra and shorts, same gold chain, skin flushed.
Action: She has the phone in one hand, so only her other hand is free. A water bottle stands
on a bridge railing beside her. SHE IS ALREADY DRINKING IN THE FIRST FRAME — her lips
closed cleanly on the bottle, her throat moving, chin lifted. She lowers the bottle onto
the railing, and that same hand sweeps her damp fringe back off her forehead. She laughs,
looks into the lens and speaks. Cut on the line.
Dialogue (Japanese, out of breath and pleased, at about 16.2s):「5キロ!」
Speech: that is everything she says in this shot. Before it she is drinking and breathing
hard with her lips closed, and after it she says nothing more.
Transition out: on the final beat that free hand snaps a white towel up to her face, and the
towel comes toward the lens and covers it completely — the frame blooms to soft white, and
the next scene opens out of that white. This is a mask transition, quick and clean. THE
WHITE COVERS A GAP OF SEVERAL HOURS — on the far side she has showered and changed and her
hair is down. Nothing is cut hard, and nothing appears out of empty air.
Camera: SELFIE. Medium close-up, small natural drift.
Audio: her hard breathing, the river below, cicadas, her voice, the muffled brush of cloth on
the lens. Sound effects only. No music.
── AFTERNOON ──
17.0 - 19.0s 【friends / FRIEND / 2.0s】
Character: the same woman, HAIR NOW DOWN IN LOOSE WAVES, washed and dried, pushed back behind
one ear, the black elastic back around her right wrist. White ribbed tank top and beige
linen maxi skirt, same gold chain.
Action: A café terrace in hard afternoon light with deep shadows. ONE OF HER FRIENDS IS
HOLDING THE PHONE, so both her hands are free. THE GLASSES ARE ALREADY SWINGING TOGETHER IN
THE FIRST FRAME — three tall iced lattes, beaded with condensation and running with water,
meeting from three directions with a knock of ice. She calls out over the noise, pulls her
glass back, laughs out loud with her head tipping back and her shoulders shaking, and cuts
on the laugh.
Dialogue (Japanese, called out brightly, IN THE FIRST HALF SECOND OF THE SHOT):「かんぱーい!」
Speech: that one word is everything she says in this shot. AFTER IT SHE ONLY LAUGHS — her
mouth opens for the laugh but forms no words, no muttering and no comment to her friends.
This shot ends at 19.0s and she has stopped making any sound with words in it well before
then.
Camera: FRIEND. Handheld medium shot with a natural sway. It stays where it is and does not
walk.
Other people: two friends as out-of-focus shoulders at the left and right edges and two hands
on glasses. Their faces stay outside the frame.
Audio: ice knocking, her call, both friends laughing, terrace chatter, cicadas loud in the
trees. Sound effects only. No music.
19.0 - 22.0s 【the turn / FRIEND / 3.0s】
THIS IS ITS OWN SHOT AND IT MUST APPEAR. It is a different place from the café: a public
pavement under street trees, not a terrace, with no table, no glasses and no seating
anywhere in frame. Cut to it hard at 19.0s and give it its full three seconds.
Character: the same woman, SAME LOOSE WAVY HAIR WORN DOWN, same white tank top and beige linen
maxi skirt, same white sneakers, same gold chain.
Action: A friend walks backwards ahead of her holding the phone, so both her hands are free.
SHE IS ALREADY WALKING TOWARD THE LENS IN THE FIRST FRAME, under street trees in hard
afternoon light. Without breaking pace she SPINS A FULL TURN ON THE SPOT, arms opening
outward. The long linen skirt lifts and flares wide, then falls and settles. HER LOOSE HAIR
FANS OUT AWAY FROM HER SHOULDERS as she turns and lands back across her collarbone a beat
after her head stops. She comes out of the spin facing the lens, still walking, grinning.
Cut as she comes out of the turn.
Speech: SHE SAYS NOTHING AT ALL IN THIS SHOT. Her lips stay closed the whole way through it,
apart from a wordless laugh.
Camera: FRIEND. ONE move only — the friend walks backwards ahead of her at her pace, holding
a full-body framing, with the natural bounce of someone walking while filming.
Light: hard sun through leaves, a moving dapple across the pavement and across her. Cumulus
clouds stacked over the rooftops behind her.
Audio: her laugh, footsteps, a friend laughing off-screen, cicadas, the street. Sound effects
only. No music.
22.0 - 25.0s 【close — her hand touches the lens / SET DOWN / 3.0s】
Character: the same woman, SAME LOOSE WAVY HAIR WORN DOWN, same white tank top and beige linen
skirt, same gold chain, black elastic on her right wrist, an evening warmth on her skin.
Action: She sits on park grass with her knees drawn up, low orange evening sun behind her
lighting only the tips of the grass. SHE IS ALREADY TURNING HER FACE TO THE LENS IN THE
FIRST FRAME. She speaks, smiles, and THEN REACHES HER RIGHT HAND STRAIGHT OUT TOWARD THE
CAMERA — the unhurried reach of someone going to stop a recording. Her open hand grows in
the frame as it comes, fingers leading, and her fingertips make contact with the lens. On
contact the whole frame jolts very slightly, the way a phone does when it is touched. THE
FILM ENDS ON THAT CONTACT. Nothing is revealed after her hand lands.
Dialogue (Japanese, warm and content, at about 22.6s):「またね」
Speech: that is the last thing said in the film. FROM 23.0s TO THE FINAL FRAME HER MOUTH IS
CLOSED AND STILL. She adds no goodbye, no thanks to the viewer and no muttering while she
reaches for the phone. The reach happens in silence.
Camera: SET DOWN. She has propped the phone in the grass in front of her, so the frame is
completely locked and never pushes, pans or drifts. It simply waits as her hand comes
toward it. Her hand starts reaching at about 23.6s and her fingertips reach the lens by
24.8s.
Light: low orange evening sun behind her, the sky cooling above.
Audio: an evening park, children far away, cicadas thinning out, her voice, and a soft contact
sound as her fingertips meet the phone. Sound effects only. No music.