Blip hero videos, prompts v3 (two-person)

Date: 22 April 2026 Supersedes for prompt craft: Section 5 of PROJECT-SNAPSHOT-260324-v2.md Session goal: spend remaining Kling subscription credits (2,300) on a coherent two-person reel before the monthly renews. 1080p, 15s, 16:9, no audio, no multishot. Roughly 120 credits per render, so ~19 renders in the pot.

Premise

v6 Family Dinner is the north star. It works because two people are interacting with each other. The camera could plausibly be held by a family member in the room. Every solo-reaction scene in the old reel fails the “who’s holding the phone?” test and reads as staged, however carefully the subject is described.

Every prompt in this doc must pass that test before we render.

The scaffold (pulled from v6)

Authentic, handheld iPhone 16 Pro footage of [PERSON A] and [PERSON B] [ordinary shared action in a universal domestic space].

[PERSON A: age, ethnicity, distinct features, ordinary clothing, natural imperfections.]
[PERSON B: age, ethnicity, distinct features, ordinary clothing, natural imperfections.]

[Setting with 2-3 specific ordinary objects, mismatched plates, a half-drunk mug, a cardigan on a chair, post on the counter. Universal homes, not aspirational lifestyle interiors.]

[THE BEAT AS INTERACTION: A does X, B reacts Y, A responds Z. The emotional payoff is the exchange between them, never a solo reaction.]

The mood is easy, warm, and completely unposed.

Subjects positioned in the right two-thirds of the frame, left third softer and out of focus for text overlay. The camera has a slight organic shake, visible skin textures, stray hair, and natural motion blur.

[Lighting: ordinary domestic, overhead bulb, grey daylight through a window, single lamp. No candles, no dramatic lighting.]

Looks like a real moment someone captured on their phone. Single continuous shot, no camera cuts. 4K, 24fps.

Craft rules

New in v3:

Carried over from v2:

Credit ledger

#ScenePlanned creditsSpentResult
1v6 Family Dinner (re-render, anchor)120
2New-v4 Late-Night Worry120
3New-v5 Curious Learner120
4New-v3 Shop Owner120
5The Remote120
6The Hand-off120
ReserveRe-takes, optional Hillside, new ideas~1,580
Total2,300

Card 1, v6 Family Dinner (re-render, anchor)

Who holds the phone: a third family member at the kitchen table (implied), or phone propped on the counter across from them pointing at the table.

Note: this is the existing approved v6 prompt. Paste it unchanged into Kling as the first render, both to verify paid-tier strips the watermark and to confirm 1080p produces the same look as the original 720p favourite.

Authentic, handheld iPhone 16 Pro footage of a mother and her adult daughter sharing a casual weeknight dinner at a kitchen table. The mother is in her late 50s, East Asian, with short grey hair and reading glasses pushed up on her forehead, wearing a loose cotton shirt. She has natural wrinkles, visible laugh lines, and an expressive round face. The daughter is in her late 20s with her hair tied back. The table has simple home-cooked food on mismatched plates, mugs of tea, a kitchen counter visible behind them. The mother finishes saying something with a casual shrug, and the daughter bursts out laughing in surprise and reaches over to playfully squeeze her mum's arm. The mother grins, pleased with herself. The mood is easy, warm, and completely unposed. Subjects positioned in the right two-thirds of the frame, left third softer and out of focus for text overlay. The camera has a slight organic shake, visible skin textures, stray hair, and natural motion blur. Overhead kitchen light, no candles, no dramatic lighting. Looks like a real moment someone captured on their phone. Single continuous shot, no camera cuts. 4K, 24fps.

Generation notes:


Card 2, New-v4 Late-Night Worry (two-person)

Who holds the phone: an adult child who’s round for the evening, sitting at the far end of the sofa filming casually. Or phone propped on the coffee table angled slightly up.

Draft prompt:

Authentic, handheld iPhone 16 Pro footage of a man in his late 40s and his partner sitting together on a worn sofa on a quiet weeknight evening. The man has light stubble with some grey in it, slightly dishevelled dark hair, reading glasses on, and natural tired-at-the-end-of-the-day features with visible lines around his eyes. He is wearing a plain t-shirt that has gone soft from washing and loose tracksuit bottoms, bare feet. His partner is the same age, greying hair pulled back loosely, no makeup, wearing a worn cardigan over pyjamas with her bare feet tucked up under her, a paperback open on her lap. Mismatched mugs of tea on the coffee table, a throw blanket pooled across their knees, an overflowing bookshelf visible behind them, a floor lamp with a slightly crooked shade. He is reading something on a tablet, frowns slightly, then turns the tablet toward her without saying much. She glances over, tilts her head to read, murmurs something brief and pragmatic, and goes back to her book. He nods to himself, picks up his phone from the arm of the sofa, taps out a short message, sets the phone face down on the throw blanket, and leans back into the cushions. His shoulders settle. The quiet of someone who has just handed a small worry off to someone else. She does not look up again. The moment is completely unremarkable. Subjects positioned in the right two-thirds of the frame, left third softer and out of focus for text overlay with a soft-focus corner of the living room visible. The camera has a slight organic shake, visible skin textures, stray hair, and natural motion blur, as if their adult child is sitting at the far end of the sofa filming casually. Single floor lamp glow, no overheads, no dramatic lighting. Looks like a real moment someone captured on their phone. Single continuous shot, no camera cuts. 4K, 24fps.

Generation notes:


Card 3, New-v5 Curious Learner (two-person, role-reversed)

To be drafted.

Concept: Black man in his 60s at the kitchen table, granddaughter (around 11) on the next chair. He is showing her something he figured out on the tablet, small shared grin. Reverses the tech-support cliché: the older person teaches. Camera-holder is another family member across the table.


Card 4, New-v3 Shop Owner (two-person)

To be drafted.

Concept: Woman early 50s unlocking her shop plus a regular customer, sister, or neighbour who has arrived at the same time. Laptop boots, small shared glance, mutual quiet smile. Camera-holder is the friend, phone at chest height.


Card 5, The Remote

To be drafted.

Concept: Older couple on the sofa, TV stuck on a menu screen. One is jabbing at the remote, frustrated. The other leans over, takes it, hits one button, it works. Silent nod, handed back. No dialogue. Camera-holder is an adult child or grandchild in the room.


Card 6, The Hand-off

To be drafted.

Concept: Woman at the kitchen counter. Husband comes in from the hall with a printed email, sets it down beside her, taps it. She reads, takes out her phone, sends a message, looks up at him. Brief nod. He refills the kettle. Scene is the trust, not the problem. Camera-holder position: propped on the fridge or another kitchen surface across the room.


Held in reserve

New-v9 Hillside Serenity (two-person): older couple walking on a ridge, small unspoken “alright?” moment. Caveat: forcing a second person may dilute the original FB2412 absence-of-tech solo freedom shot. Only render if the core 6 land and credits allow.

Dropped entirely:


Card 7, Small Office Team (three-person, small business)

Added late in the session after live preview on localhost: v4 couple-on-sofa was rendered well but the reel gained more from shifting the third slot to a small-business-team scene. Source: an AI reference image in Blip 26/previous blip assets/2025 site images/Three-people-working-in-a-small-office.jpg Christian has loved for a long time. Half of Blip’s clients are small businesses; no competitor hero shows this.

Who holds the phone: propped on a filing cabinet or plant stand across the room, as if a colleague briefly set it down while going to get a coffee.

Compositional note: three faces is near the edge of Kling’s reliable two-subjects rule. The two women share the foreground laugh; the male colleague sits in the soft-focus left third, partially visible at his own desk, so only two faces need crisp rendering. Left third also serves as text overlay space.

Draft prompt:

Authentic, handheld iPhone 16 Pro footage of three colleagues working together in a small creative studio on a warm late afternoon. The seated woman at the central desk is in her late 30s, dark hair pulled back in a loose messy bun with strands escaping, natural smile lines around her eyes, no makeup, wearing a soft white cotton blouse with the sleeves loosely pushed up. She is at a large iMac, hands on the keyboard. Behind her, a woman in her early 40s is standing and leaning in slightly to watch the screen, shoulder-length brown hair, warm open face with natural laugh lines, wearing a loose charcoal grey jumper and jeans. To the left in soft focus, a male colleague in his mid-30s sits at his own desk working on a laptop, light beard, mid-brown hair slightly tousled, pale blue button-down, wearing over-ear headphones as he types, fully absorbed in his own work. The office is lived-in and specific: plants crowding the windowsill and every surface, an anglepoise lamp arched over each desk, a pinboard thick with photos and colour swatches, stacked notebooks, mismatched mugs, sticky notes everywhere, a wooden shelf with ring binders, paper and cables on the desks. Warm late afternoon golden light comes in through a tall window at the back showing a soft-focus urban rooftop outside. The seated woman reads something on her screen and bursts into a real unforced laugh, shaking her head. The standing woman leans in further to see the screen, her eyes widen in recognition and she laughs too, hand landing briefly and lightly on the seated woman’s shoulder. They share the moment, both grinning. The man at the left desk glances up for a second, catches the sound, small wry smile, goes straight back to typing. The standing woman straightens up still smiling, turns toward her own desk, the seated woman shakes her head amused and resumes typing. Nobody speaks. Subjects positioned in the right two-thirds of the frame, the left third is a soft-focus bookshelf and corner of the man’s desk, kept peripheral for text overlay. The camera has a slight organic shake, visible skin textures, stray hair, and natural motion blur, propped on a filing cabinet across the room as if a colleague briefly set their phone down while getting a coffee. Warm golden-hour window light, no dramatic lighting, no studio look. Looks like a real moment someone captured on their phone. Single continuous shot, no camera cuts. 4K, 24fps.

Generation notes:

If Kling struggles with three people: fallback is to drop the male colleague, render just the two women sharing the laugh at the central desk with the office still visible around them. Same prompt minus the paragraph describing the man, and adjust “three colleagues” to “two colleagues”.


Session journal

Filled in live as renders come back.