LESSON 2 · THE CORE PRINCIPLE

Describe the motion, not the frame

In image-to-video mode, WAN 2.7 Spicy already has your starting frame [Image 1] in front of its eyes. The model doesn't need you to tell it what's in the image — it needs you to tell it what should come into motion starting from that frame. This is the heart of the entire course: a prompt isn't a caption for a picture, it's a director's command for motion.

One principle from which everything else grows
The frame is already a given. The model sees the hair color, the coffee shop, the light from the window. Every word spent retelling the static image is a word not spent on action, camera, and pacing. A WAN 2.7 Spicy prompt describes what is not yet in the frame: how it will come alive.

Why retelling the frame is wasted words

Picture a starting frame: a woman with chestnut hair sits by the window of a coffee shop, a cup on the table, soft light beyond the glass. If you write the prompt "a woman with brown hair sitting in a cafe, warm light, a cup on the table," you're telling the model exactly what it already sees. The result is almost always the same — a limp, barely-moving picture: the model got not a single cue for motion, so it just gently ripples the pixels, often adding artifacts and image "breathing."

Compare that with a prompt that has not a word about appearance or setting, but does have an action verb, a camera move, and pacing. The model will fill in the appearance from the frame itself — and you hand it the very thing generation is launched for.

Bad: the prompt retells the still life of the frame

A woman with brown hair sitting by the window in a cozy cafe. Warm afternoon light, a white cup of coffee on the wooden table, a peaceful and calm atmosphere, soft background blur.

There's zero motion here. No action verb, no camera, no pacing — just an inventory of what's already in the frame. WAN 2.7 Spicy has nothing to animate.

Good: the prompt describes motion from the frame

She slowly lifts the cup, takes a sip, then turns her head toward the camera and smiles softly. Steam drifts upward from the coffee. The camera performs a gentle slow push-in on her face. Calm, steady pacing throughout the shot; the motion settles as she lowers the cup.

Not a word about "brown hair" or "cafe" — that's in the frame. Instead there's action (lifts, sips, turns, smiles), environmental physics (steam drifts up), camera (slow push-in), and a pacing arc (settles).

Quick check before you launch
Read your prompt and honestly answer three questions. If even one answer is "no," the prompt isn't about motion yet.
  • Is there an action verb? (turns, walks, rises, pours, sweeps — not "a peaceful scene")
  • Is a camera move specified? (static / push-in / pan / orbit — or a deliberate "static")
  • Is the pacing set? (slowly / steady / builds up / rapid)

Frame vs. prompt: who's responsible for what

It helps to divide the zones of responsibility once and for all. The frame is the "what," the prompt is the "how it moves."

Give to the starting frame [Image 1]Give to the prompt
The character's appearance, clothing, faceWhat the character does: a specific action verb
Location, objects, compositionHow the camera moves and where
The original light and colorHow light and the environment change over time
The picture's style and moodThe pacing and motion arc: start → peak → settle
The #1 beginner mistake
Over-describing the static frame with adjectives ("beautiful," "peaceful," "detailed") in the hope of "improving" the result. It doesn't improve it — it dilutes the motion command and lowers coherence. Describe the action; the frame already carries the beauty on its own.

The five layers of motion — in brief

Every good motion prompt is assembled from five layers. Here's just the essence with verb examples; we break each layer down in detail in the lesson on prompt anatomy.

01
Subject action
A specific verb: turns, walks, reaches, pours, exhales, spins
02
Camera
push-in, pull-out, pan left, tilt up, 360 orbit, tracking, handheld
03
Pacing
slowly, steady, builds up, rapid — the rhythm of motion over time
04
Light and environment
light flares, rain intensifies, fog disperses, steam rises
05
Continuity / arc
how the motion starts and how it settles: calm → peak → settle, "throughout the shot"

Action verbs instead of static. Replace "a woman standing" with what she's doing: turns toward the camera, raises her hand, steps forward. The verb is the heart of layer 1.

Camera to match the action. A calm scene — a smooth push-in; running or an impact — tracking or handheld. A static camera under frenzied action (or a wild one under a quiet scene) looks fake.

Pacing sets the energy. The same action with slowly and with rapidly makes two different clips. Always tell the model the rhythm.

The environment lives over time. Smoke rises, rain intensifies, fog disperses — these changes make the frame truly "video" rather than an animated photo.

The lesson formula
Action + Camera + Pacing is the minimum viable motion prompt. Add environmental changes and a motion arc and you get an expressive shot. Retell the static and you waste words for nothing.
motion, not description action verb camera to match action pacing bad → good

Try it yourself in ZenCreator

Generate your first clip — everything this course teaches is available in the ZenCreator video tools.

Start Creating
PreviousNext lesson