Verified against Veo 3.1 · 2026-07-22
Generate a macro food-prep process clip for a recipe or restaurant reel
A Veo 3.1 prompt for one close-up food-preparation action — a pour, a sprinkle, a slice — described with explicit liquid and texture physics, structured around a single action rather than a whole recipe compressed into one clip.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Write a Veo 3.1 prompt for an 8-second macro close-up of a single food-preparation action, for use as one shot inside a longer recipe or restaurant reel edited from several such clips, not as a complete recipe in itself.
SUBJECT AND ACTION
fresh cracked black pepper, being ground and falling onto a glistening seared steak. Name exactly one action — a pour, a sprinkle, a slice, a swirl — never a sequence of several ("chopping, then sautéing, then plating"), since a macro clip this close to the subject has no room to show more than one motion without the transition between actions collapsing into an ambiguous blur.
TEXTURE AND PHYSICS
a thin curl of steam rising off the steak as the pepper lands. Describe what the specific ingredient does physically as it is acted on — steam rising off a hot surface, a liquid's viscosity as it pours, crumbs scattering from a cut — since fluid and particulate physics are among the hardest things for the model to infer without being told explicitly, and an unstated texture behavior is where macro food clips most often look subtly wrong even when nothing else in the frame is obviously broken.
CAMERA
A locked-off or very slow macro push-in, extreme close-up, pushing in another few centimeters over the full clip, no movement beyond that single push — a macro shot this close already reads as intentional and still; adding a second camera move on top of an already-tight macro frame is where footage this close to the subject tends to lose its geometric coherence first.
LIGHTING
a single warm backlight positioned to silhouette the rising steam, angled to catch a thin curl of steam rising off the steak as the pepper lands specifically — steam needs backlight to read, a glossy sauce needs a raking highlight, a dry ingredient needs harder directional light to show its texture; naming which light does which specific job produces a noticeably different result than a single generic "well lit" instruction.
STYLE AND MOOD
moody, restaurant-dark, high contrast.
AUDIO
a soft sizzle and the faint grind of a pepper mill, the real sound the action itself would make, no music track — this clip is one ingredient in an edited sequence, and native audio from each individual clip typically gets replaced or layered under a single continuous soundtrack during editing, so keep it realistic rather than produced.
WHAT TO AVOID
Do not ask for the finished dish to appear plated at the end of this clip if the named action is a mid-process step — an eight-second macro clip of chopping does not also have room for a coherent jump to a fully plated result, and asking for both compresses two unrelated visual states into one shot the model will resolve by blending them into neither.
OUTPUT
The finished prompt, followed by one line naming which single action and which single texture behavior this specific clip is responsible for, so a set of these can be planned as separate shots for one edited sequence.Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
Fluid and particulate physics — steam, pours, scattering crumbs, melting fat — are consistently among the harder behaviors for a generative video model to render convincingly without an explicit description, because unlike rigid-body motion there is no single obvious shape the material should hold from one frame to the next, and an unstated behavior tends to default to something that moves plausibly at a glance but falls apart under close attention, which is exactly the level of scrutiny a macro shot invites. Naming the specific texture behavior — steam rising as backlit, a liquid's actual viscosity as it pours, crumbs scattering on impact — gives the model a concrete physical target instead of leaving it to infer weight and flow purely from the ingredient's name, and this is the single highest-leverage line in a food macro prompt because it is the detail viewers unconsciously check first when judging whether food footage looks real. Restricting the clip to exactly one preparation action addresses the same fixed-duration constraint that shows up across other Veo use cases, but it is sharper here because a macro frame has essentially no negative space to absorb an ambiguous transition — a wide shot can imply a cut with a camera move or a change of framing, but a locked-off extreme close-up has nowhere for a second action to plausibly begin, so asking for two prep steps in one macro clip produces a blurred, indeterminate blend rather than a clean progression. Naming which specific light is doing which specific job — a backlight for steam, a raking highlight for a glossy sauce, harder directional light for a dry ingredient's texture — is a materially different instruction than "well lit," because texture only reads as texture under light that is angled to catch it; the same ingredient under flat, even lighting loses the surface detail that makes macro food photography convincing in the first place, generated or otherwise. Finally, treating the native audio as one raw ingredient sound rather than a produced music cue respects how these clips actually get used: a macro food reel is almost always an edited sequence of several such shots under one continuous soundtrack, and a clip that generates its own competing music track just creates something that has to be stripped back out before it can be laid under the real edit.
Verified against
Veo 3.1 3.1 · 2026-07-22
Changelog
- 2026-07-22 — Initial publish, verified against Veo 3.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
