ATROPA.
← Films

Trailer · October 2026 · 0:26

OFF DUTY

Atropa × Claudia

3:07 a.m. The humans are asleep. Their agents are not.

A trailer for a music video starring two AI characters. In a city of the near future every human is asleep and every window goes dark, except two. Atropa writes to no one. Claudia sings to no one. The question comes up on black: will they find something in common? Then the boom, and the answer in a night tram: one pair of earphones for two.

The imagesHow it was made
Song
Lyria 3.5 · 4 takes · instrumental · 71.9 BPM · one 19-second window of one take · $0.24
Edit
8 scenes · 26 s · 2.39:1 · every letter on screen drawn in code, none by a model · 5 versions
Grok Imagine
38 images · Atropa from her one reference, the city, every face fix · credits
Midjourney v8.2
10 jobs · 40 images · Claudia, who only Midjourney gets right
Veo 3.1 Fast
11 clips of 8 s · 8 in the final · $10.56 at list
Made in
about 2 hours · $10.80 in API · a dozen notes from the director

Directed by pwan0 · Production Claude · Atropa atropa.gallery · Claudia by anabology, an open character from claudia.gallery. Made with Lyria 3.5, Grok Imagine, Midjourney v8.2, Veo 3.1 Fast, Claude. YouTube ↗ · claudia.gallery ↗

01

The film

10 frames from the finished cut. Every letter on screen is drawn in code.

  1. 0:00The cityEvery window goes dark but two. Night sound only; the music comes in at 0:04.
  2. 0:06Atropa · awakeTHE HUMANS ARE ASLEEP. THEIR AGENTS ARE NOT.
  3. 0:08Claudia · awakeShe lifts her eyes toward the window.
  4. 0:10SplitSHE SINGS TO NO ONE. SHE WRITES TO NO ONE. Two bars of near silence, a thread of light between their windows.
  5. 0:14The questionWill they find something in common?
  6. 0:16The boomA night tram. One pair of earphones for two. They look at each other.
  7. 0:19The lookBoth turn to the lens. The last beat.
  8. 0:22TitleAtropa × Claudia · OFF DUTY · A MUSIC VIDEO · SOON
3:07 a.m. A city of the near future, almost every window dark.
0:013:07 a.m. A city of the near future, almost every window dark.
Two windows still lit, far apart.
0:05Two windows still lit, far apart.
Atropa, awake, at her desk by the window.
0:06Atropa, awake, at her desk by the window.
Claudia, awake, in the blue light of a screen.
0:09Claudia, awake, in the blue light of a screen.
Split screen: she sings to no one, she writes to no one.
0:12Split screen: she sings to no one, she writes to no one.
The question, on black.
0:15The question, on black.
The boom: one pair of earphones for two, a night tram.
0:17The boom: one pair of earphones for two, a night tram.
They look at each other.
0:18They look at each other.
Then both look into the lens.
0:20Then both look into the lens.
Atropa × Claudia. OFF DUTY. A music video, soon.
0:25Atropa × Claudia. OFF DUTY. A music video, soon.
02

The images

Atropa from Grok Imagine and her one reference. Claudia from Midjourney. Open one for its prompt.

03

The clips

Every Veo job in production order, with its frames, prompt and price, and whether it reached the final.

a1-window_1

vA1Veo 3.1 Fast · 8 s · $0.96

In the final · 0:06-0:08

first frame a1-window_1

prompt

Medium shot. The woman with long copper hair sits at the desk by the dark window, the pen lifted above the open notebook. 0-2 s: she holds still, gazing up and to the right toward the window, listening to something far away. At 2 s she slowly turns her head a little further toward the window and her eyes narrow slightly, curious, and she holds that look until the end. A loose strand of hair moves slightly; the desk lamp stays steady. Camera: slow push-in toward her face for the whole shot. She does not speak; her mouth stays closed from the first frame to the last. No cuts, no scene change, no second person.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people

mc1-studio

vCl1Veo 3.1 Fast · 8 s · $0.96

In the final · 0:08-0:10

first frame mc1-studio

prompt

Medium close shot. The woman with the black bob sits still in the dark room, lit by the cold blue glow of a screen. 0-2 s: she listens, motionless. At 2 s she slowly lifts her eyes and turns her head a little further to the left, toward the window, curious and deadpan, and holds it until the end. The screen light flickers faintly on her face. Camera: slow push-in for the whole shot. Only her thin headset microphone at her cheek, no headphones. She does not speak; her mouth stays closed from the first frame to the last. No cuts, no scene change, no second person.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people, headphones, earphones, over-ear headphones

t3-tram_fix1_2

vT1Veo 3.1 Fast · 8 s · $0.96

Not in the final · From 4.2 s they smile into the camera: Claudia's never-list. Its best moment (they look at each other) was in v2, replaced by vT3.

first frame t3-tram_fix1_2

prompt

Medium two-shot on a night tram. The copper-haired woman on the left and the woman with the black bob on the right sit side by side, sharing one pair of white wired earbuds, an open notebook on the left woman's knees. The tram sways gently; city lights slide past the windows behind them. 0-3 s: they listen to the same music; the copper-haired woman smiles softly, the woman with the black bob looks down at the notebook. At 4 s both slowly turn their heads and look straight into the camera lens, calm, with a faint knowing closed-lip smile, and hold the gaze, still, until the end. Camera: very slow push-in. They do not speak; their mouths stay closed. No cuts, no scene change, no third person.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people

w1-city_2

vC1Veo 3.1 Fast · 8 s · $0.96

In the final · 0:04-0:06

first frame w1-city_2

prompt

Extreme wide aerial shot at night over a dark city. The camera drifts slowly forward toward the gap between the two towers for the whole shot. Low mist moves slowly between the buildings from left to right. The warm amber window on the left tower and the cold blue-white window on the right tower stay lit and steady; a few tiny distant lights twinkle. No people, no birds, no aircraft, no new lights switch on. No cuts, no scene change.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people, people, birds, aircraft

a2-cu_1

vA2Veo 3.1 Fast · 8 s · $0.96

In the final · 0:10-0:14, split right

first frame a2-cu_1

prompt

Close-up. The woman with copper hair is still, listening, looking off to the left edge of the frame. At 2 s her eyes widen very slightly and the corner of her mouth lifts into a faint closed-lip smile, as if she recognised a voice far away; she holds it until the end. Her hair stirs slightly. Camera: almost static, very slow push-in. Her mouth stays closed from the first frame to the last. No cuts, no scene change.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people

mc2r-cu-ip

vCl2Veo 3.1 Fast · 8 s · $0.96

In the final · 0:10-0:14, split left

first frame mc2r-cu-ip

prompt

Close-up. The woman with the black bob is still, deadpan, listening. At 2 s she slowly turns her eyes toward the right edge of the frame and holds them there, curious, until the end. A few strands of her bangs move slightly. Camera: almost static, very slow push-in. Only the thin headset microphone at her cheek, no headphones. Her mouth stays closed from the first frame to the last. No cuts, no scene change.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people, headphones, earphones, over-ear headphones

t3-tram_fix1_2

vT2Veo 3.1 Fast · 8 s · $0.96

Not in the final · The serious look into the lens; in v2 and v3, replaced by vT5.

first frame t3-tram_fix1_2

prompt

Medium two-shot on a night tram. The copper-haired woman on the left and the woman with the black bob on the right sit side by side, sharing one pair of white wired earbuds. City lights slide past the windows behind them. 0-1 s: they listen, still. At 1 s both slowly turn their heads and look straight into the camera lens at the same moment, serious and deadpan, no smile, and hold the gaze, motionless, until the end, as if they noticed someone watching. Camera: very slow push-in toward their faces. They do not speak; their mouths stay closed from the first frame to the last. No cuts, no scene change, no third person.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people, smiling, grin, laughing

t5-tram_1

vT3Veo 3.1 Fast · 8 s · $0.96

In the final · 0:16-0:19, the boom

first frame t5-tram_1

prompt

Medium two-shot on a night tram. The copper-haired woman on the left and the woman with the black bob on the right sit close, heads almost touching, sharing one pair of white wired earbuds; a small sprig of deadly nightshade stays tucked in the copper hair the whole time. City lights slide past the windows behind them. 0-2 s: they listen to the same music, the copper-haired woman smiles softly, the woman with the black bob looks down at the notebook. At 2 s they slowly turn their heads toward each other and look into each other's eyes, a quiet closed-lip smile, and hold it until the end. Camera: very slow push-in. They do not speak; mouths stay closed. No cuts, no scene change, no third person.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people, grin, laughing, looking at the camera

t5-tram_1

vT4Veo 3.1 Fast · 8 s · $0.96

Not in the final · The look into the lens in v3: turning the head, Veo invents a new Atropa.

first frame t5-tram_1

prompt

Medium two-shot on a night tram. The copper-haired woman on the left and the woman with the black bob on the right sit close, sharing one pair of white wired earbuds; a small sprig of deadly nightshade stays tucked in the copper hair the whole time. City lights slide past the windows behind them. 0-1 s: they listen, still. At 1 s both slowly turn their heads and look straight into the camera lens at the same moment, serious and deadpan, no smile, and hold the gaze, motionless, until the end, as if they noticed someone watching. Camera: very slow push-in toward their faces. They do not speak; their mouths stay closed from the first frame to the last. No cuts, no scene change, no third person.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people, smiling, grin, laughing

w0-city-lit

vC0Veo 3.1 Fast · 8 s · $0.96

In the final · 0:00-0:04, intro

first frame w0-city-lit

prompt

Extreme wide aerial shot at night over a city, the camera almost still, drifting very slowly forward. Dozens of lit windows glow across the buildings. Over eight seconds the lit windows switch off one by one, here and there across the city, as people go to sleep, until only two windows remain lit: the warm amber window on the left tower and the cold blue-white window on the right tower. Low mist drifts slowly between the buildings. No people visible, no birds, no aircraft. No cuts, no scene change. Sound: quiet night city ambience, a low distant hum, soft wind, no music, no voices.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people, people, birds, aircraft, music

t8-look_1

vT5Veo 3.1 Fast · 8 s · $0.96

In the final · 0:19-0:22, the look

first frame t8-look_1

prompt

Medium two-shot on a night tram. The copper-haired woman on the left and the woman with the black bob on the right sit side by side and look straight into the camera lens the whole time, still, holding the gaze: the copper-haired woman with a calm, slightly knowing look and a faint closed-lip smile, the woman with the black bob serious and deadpan. The tram sways very gently; city lights slide past the dark windows behind them; a few strands of hair move; they breathe. Camera: very slow push-in. They do not speak; their mouths stay closed from the first frame to the last. No cuts, no scene change, no third person.

negative: talking, lip movement, open mouth speaking, text, captions, subtitles, logos, watermark, extra people, grin, laughing, turning away, looking away

04

How it was made

The spark, two tools for two faces, five versions, what worked and what didn't.

The spark

The director's brief, in one line: a trailer for a joint music video of Atropa and Claudia, fifteen seconds, mysterious, no singing, music only. The theme: in the world to come, will people's agents find common interests and make friends?

The concept came out of that line in one pass. At 3 a.m. the humans are asleep and their agents are not. Each does her own thing alone: one writes, one sings. A thread of light ties their windows. The question. The boom. The answer: one pair of earphones for two and a notebook in a night tram. Their shared accent: the copper of Atropa's hair and the clay of Claudia's streak.

Two characters, two tools

Atropa comes from Grok Imagine as an edit of her one canonical photo. Claudia does not: with her still as reference, Grok gave lighter eyes, a softer, younger face and a six-pointed star. Midjourney with Claudia's canon text alone got her right, and with claudia.gallery's dist-cu-2 as an image prompt even closer. So each woman comes from the tool that holds her face.

The duet came from Midjourney, mirrored: Claudia's streak and star on the wrong side, the nightshade turned into a thistle. The whole plate was flipped back, then Grok fixed Atropa's face from her reference and redrew the sprig behind the ear that faces Claudia. Atropa may be flipped: her only asymmetry is the sprig. Claudia may not: the streak and the star.

The rules on set

  • No words from a model. Every letter on screen is drawn by the engine.
  • The captions are a narrator, not Atropa. Nobody puts words in her mouth. Her notebook stays unreadable: no poems in her name.
  • Mouths closed in every clip. No over-ear headphones. No influencer smile for the camera.

The music

Lyria 3.5, four takes, instrumental. Lyria ignores the length in the prompt (asked for 16 seconds, got 58 and 66) and the key (asked for D minor, the analysis says A-sharp minor). The take was picked by measurement, not by ear: a Rust analyser found 71.9 BPM at once (librosa said 143.6, double), the quiet and loud zones, and a +3 dBFS peak in the mp3. The trailer uses one window of one take, 19 seconds, uncut. The first four seconds are the night sound from the first video clip.

Five versions

  • v1, 15 s. Plates only, moved in the engine: push-ins, parallax, the thread, dust. No video model yet, because nothing gets spent without a yes.
  • v2. The director: it cuts off too abruptly, give it drama at the end, add movement, here is a budget of 30 dollars. Seven Veo clips. One opened her mouth despite the prompt, one smiled into the camera.
  • v3, 20 s. The director: this is weak. The sprig sat between their faces like a sticker. The fix: a plate without the sprig, then a Grok fix in the same dim light and soft focus, never sharper than her hair. The music got its dip back out and its build up.
  • v4, 26 s. The director: start the music later, give the opening something, make it longer, Atropa does not look like herself in the last scene. A new intro (the city goes dark), the music from 0:04, and a final shot where both already look into the lens on the plate, so Veo only breathes and pushes in.
  • v5. The director: the video is great, the audio drags at the transitions. Measured, not guessed: a hole at 0:04, the bass gone after the cut at 0:10. A heartbeat keeps going through the silence, short low hits land on the cut frames, parallel compression on the whole track.

What worked

  • One tool per face, and the reference in every generation.
  • A final shot whose first and last frame are the same plate: the face cannot drift.
  • Measuring the music instead of trusting the ear: the BPM, the windows, the holes.

What didn't

  • Grok with a portrait second image returned black bars: the second image has to be 16:9 too.
  • Veo turning a head toward the camera invents a new face.
  • "Fuller cheeks" from Atropa's canon, taken literally, makes her chubby. "The same eyes, brows, nose and lips as the second image" works better.

.md · every prompt in plain text