Prompting Guide

How to direct the MotionLab video model: what to put in the image, what to put in the prompt, and what each control on the Create page is for.

Before You Generate

Six rules that fix most weak renders

Read these in a minute. Each one opens the chapter that explains it.

  1. Your image does half the workMotionLab animates the Start image. Composition, subject and style are already decided there.
  2. Prompt the motion, not the pictureThe model already sees the frame. Spend every sentence on what changes over time.
  3. One camera move, chosen in the Camera controlPick the move in the control, use the prompt for its speed and what it reveals.
  4. One light source, atmosphere in motionFog, rain, dust and shifting reflections are the cheapest way to make a frame feel alive.
  5. Match the prompt to the clip lengthTwo or three sentences for a short clip, a sequence of beats for a long one.
  6. Stuck? Press Create PromptThe Creative Director writes a first draft in this format from your image. Edit from there.
01

Made for MotionLab

Read This First

This guide is written for the MotionLab video model. MotionLab starts from an image and turns it into a moving clip of 3–20 seconds — so your prompt's job is to direct what happens over time, not to repaint the picture. Renders carry a soundtrack generated from what happens on screen, so keep every sentence about what the camera can see. Other models in the picker follow their own rules; the MotionLab controls described here appear only when a MotionLab model is selected.

•
Start from an image: MotionLab animates a starting frame. The composition, subject, and style are already locked in by your image — choose it carefully, it does half the work.
•
The prompt directs motion: Describe what changes over time: the subject's action, the camera's movement, and how light and atmosphere evolve.
•
Write for the camera, not the microphone: The soundtrack is generated from the finished visuals — there is no separate audio prompt. Instead of "the sound of rain," describe what rain looks like: droplets streaking past a streetlight, ripples spreading in puddles. Strong visual events give the audio something to follow.

Pro Tip: If you re-describe the image in detail, the model spends its effort confirming the frame instead of animating it. One short anchor sentence about the scene is enough — then move straight to motion.

02

Prompt the Motion, Not the Picture

The Golden Rule

The model already sees your image. Focus every sentence on what happens next: how the subject moves, how the camera travels, and what shifts in the light and atmosphere around them.

Prompt Formula

[Scene anchor] + [Subject action] + [Camera movement] + [Environment in motion]

Weak Prompt

"A red sports car parked on a wet city street at night, neon signs, reflections, cinematic, 4K."

Strong Prompt

"The camera slowly orbits the parked sports car as neon reflections slide across its wet paint. Steam drifts up from a grate behind it, and light rain begins to streak through the glow of the signs."

03

Structure & Length

Writing the Prompt

Write one flowing paragraph in present tense — not keyword tags, not bullet points. Use connectors like "as," "then," and "while" so actions flow into each other naturally from beginning to end.

One paragraph, present tense

"The dancer spins, her dress flaring outward" — not "dancer, spinning, dramatic, HD." The model understands natural language and the relationships between elements.

Match length to duration

2–3 sentences of motion suit a 3–6 second clip; a 12–20 second clip needs 5–8 sentences describing a sequence of beats. A short prompt on a long clip makes the model rush or drift.

Sequence longer clips in time

Give longer clips an order of events: "The camera holds steady at first. Then it begins a slow push-in as the fog thickens and the figure turns toward the light."

Show emotion physically

Write "her shoulders slump and she stares at the empty cup" instead of "she is sad." The model renders what a camera can see, not internal states.

There is a ceiling

The model reads a limited amount of text, so past a point extra sentences stop adding anything. Fewer, stronger beats beat a wall of words — and if you use Enhance Prompt, it keeps the result within that limit for you.

04

The Creative Director

Create Prompt & Enhance Prompt

You do not have to write the prompt alone. The button beside the prompt box is the MotionLab Creative Director: with an empty box and a Start image it reads Create Prompt and writes a motion prompt from what it sees in the image; once you have typed an idea it reads Enhance Prompt and rewrites your idea in the structure this guide teaches. Either way the result lands in the box for you to edit before you spend any credits.

•
It looks at your image: The draft is grounded in the Start image — what is actually in the frame, where there is room to move, which parts are static and need a job. Two or three words of intent are enough to steer it: "slow, moody" or "fast product reveal".
•
It respects your controls: Camera, Loop and Audio choices are read from the controls, not invented in the text, so the draft will not fight the move you picked or write a one-time event into a clip you marked as a loop.
•
Treat it as a first draft: Read it back against the golden rule: is every sentence about something that changes? Cut what re-describes the picture, sharpen the one beat you care about, then generate.

Pro Tip: Use the Creative Director as a teacher as well as a tool. Compare what it wrote with what you typed and you will see the structure — scene anchor, subject action, camera, environment — in practice on your own image.

05

Camera Language

Cinematography

The model understands standard cinematography vocabulary, and MotionLab also gives you a Camera control under the prompt. Commit to one clear camera behavior per clip — a single clean move always looks better than two competing ones.

Choose the move in the Camera control

Static, Dolly In, Dolly Out, Pan Left, Pan Right, Jib Up, Jib Down, and 360° orbits in four directions. The choice is passed to the render as a dedicated camera instruction, which is more reliable than hoping the prompt is read the way you meant it. The control stays available in Reference mode and in Loop mode.

Use the prompt for intensity and reveal

Once the move is set, the prompt decides how it feels: "subtle," "gentle," "steady," "gradual" for calm shots — "sweeping," "rapid" for energy — and what the move uncovers as it goes. Do not write a second, conflicting move in the text.

Scale & framing

Close-up, medium shot, wide establishing shot, overhead view. Close-ups need more fine detail in the prompt; wide shots need more environment.

Film effects

Film grain, lens flares, shallow depth of field, motion blur, slow motion, time-lapse — name them explicitly if you want them.

06

Light, Atmosphere & Style

Setting the Mood

Light and atmosphere are the strongest mood levers — and the best source of motion in otherwise static parts of the frame.

One lighting logic

Pick a single light source and give it behavior: "low golden-hour sun casts long shadows," "a neon sign flickers, tinting the fog pink." Conflicting light directions confuse the model.

Animate the atmosphere

Fog rolls, dust drifts through light beams, rain streaks, reflections shift, embers rise. Atmosphere in motion is what makes a clip feel alive.

Give static regions a job

If part of your image is empty — sky, water, a wall — assign it motion: "clouds drift slowly across the sky," "ripples spread across the still water."

Name the medium for stylized looks

"Watercolor," "hand-drawn animation," "claymation," "analog film" — committing to a medium keeps the style stable across every frame.

07

Start, End & Middle Frames

Steering With Images

The Start image is required; it is the first frame of your clip. You can also hand MotionLab a destination and waypoints: an End image the clip must arrive at, and up to two Middle keyframes it should pass through on the way. The prompt then describes the journey between those frames rather than inventing the whole shot.

End image

Drop a second image into the End slot and the render finishes on it. Use it for a reveal, a transformation, or a camera move that has to land on a specific framing. Loop and End cannot be combined — a loop returns to the Start image instead.

Middle keyframes

Each keyframe has a position (where in the clip it should appear, as a percentage) and a strength (how closely the render must match it). Full strength pins the frame; lower strength treats it as a suggestion the motion can flow around.

Keep the frames related

Frames that share subject, style, lens and camera scale produce smooth journeys. A wide shot into an extreme close-up, or a photo into a painting, forces a hard transition the model has to invent.

Write the path, not the pictures

The frames already say what the start and end look like. The prompt should say how one becomes the other: "the camera drifts right and the shutters swing open as daylight floods the room."

References take over

Adding Reference images switches MotionLab into Reference mode, which uses the Start image as the scene and turns off End, keyframes, Loop and Motion Reference automatically.

08

Reference Images in MotionLab

Direct the Cast

Use the Start image as the scene anchor, then add 1–4 Reference images for the people, objects, clothing, or locations you want MotionLab to bring into that scene. The model plans the images together, so the prompt should give every reference a clear purpose, role, action, and position instead of simply asking it to copy the uploads.

Pro Tip: Describe relationships, actions, and placement rather than re-describing every pixel. Clear direction gives MotionLab a better plan for combining the scene and subjects.

Prompt Formula

[Scene from the Start image] + [Reference 1 role and action] + [Reference 2 role and action] + [Camera and environment motion]

Weak Prompt

"Use these references in the scene."

Strong Prompt

"On the subway platform from the Start image, Reference 1 is the woman in the red coat walking toward the camera. Reference 2 is the silver case in her right hand, and Reference 3 supplies her black jacket. The camera tracks backward as train lights sweep across the platform."

Start image = scene anchor

Treat the Start image as the composition, background, and spatial logic of the shot. References are the cast and ingredients placed inside it; they should not compete with or replace the scene.

Name every reference

Use the exact slot names — "Reference 1," "Reference 2," and so on — and state what each one is doing and where it belongs. The numbering follows the slots from left to right.

Choose the type when it matters

Each slot starts on Automatic, which identifies the reference from the image and your prompt. Tag it as Person, Object, Clothing, or Location when you want to be explicit — the type helps MotionLab frame and crop that source correctly. Several images tagged Person can show the same character from different angles.

Prepare clean source images

Use one clear subject per image. A transparent PNG is ideal; a plain, high-contrast background is the next best option. Ordinary photos can work, but extra scenery may be interpreted as part of the reference.

Camera remains available

You can still choose a Camera move in MotionLab reference mode. Other incompatible controls are disabled automatically while Reference images are present.

Identity is guidance, not a guarantee

References strongly guide appearance, palette, and silhouette, but exact identity, fine details, readable text, and logos are not guaranteed. Use fewer, cleaner references with non-conflicting roles for the most consistent result.

09

Motion Reference Video

Bring Your Own Movement

Drop an MP4 into the Motion slot next to your image and the render follows that clip's movement instead of inventing its own. It transfers motion — the camera path, the timing, the rhythm — not the look: the subject, palette, and style still come from your start image and prompt.

Use it when timing matters

A move you already storyboarded, a beat you have to hit, a camera path from an existing shot, or a gesture you want repeated across several clips. A phone clip is enough — it is read for movement, not quality.

Clip requirements

MP4, up to 50 MB and 60 seconds, and at least as long as the clip you are rendering. MotionLab re-encodes it in your browser to the fps you selected and trims it to your duration before it is sent.

The prompt still directs the scene

Keep writing the shot as usual. The reference decides how things move; the prompt decides what is moving, how it is lit, and what the atmosphere does.

Costs a small add-on

A reference video adds a fixed amount of credits to the render, shown in the price next to the generate button before you commit.

10

Loop & Cyclic Motion

VJ & Motion Graphics

Turn on Loop for a supported MotionLab image-to-video model to use the opening image as the return frame. Loop needs a Start image and cannot be combined with an End image. The Loop label identifies the requested mode; it is not a quality verdict. Watch the downloaded clip across several repeats and check the join, subject stability and any audio. Simplify the motion, regenerate or finish the clip if the join remains visible.

Prompt motion that wants to cycle

The switch closes the seam; the motion decides whether it reads. Direction like "cyclic motion," "steady rhythm," or "continuous repetitive movement" — wind, waves, smoke, rotation, breathing light — loops convincingly. A one-time event inside the clip still reads as a restart on the second pass.

Loop and audio

The soundtrack is joined smoothly at the seam. If you upload your own audio, it only loops cleanly when its length matches the clip — so you can either keep the smooth join or mute the audio and let the video loop on its own.

Subject Isolation

Keep the background description minimal or solid-colored if you plan to extract the subject later for transparent overlays.

11

Sound & Music

Audio

Every MotionLab render can carry sound. The Audio switch on the video page is on by default, and the soundtrack is generated to match what happens on screen. You can also hand the render a track of your own, or write music from scratch in the separate Music mode.

Pro Tip: For lyrics, use tags such as [Verse], [Chorus], and [Bridge]; for a timed arrangement, write lines like "[0:00 - 0:10] soft piano intro". Choosing Instrumental explicitly is more reliable than hoping a prompt stays wordless.

Audio switch

On by default — the finished MP4 carries a generated soundtrack. Turn it off when you need a silent deliverable: LED walls, loops running under a live set, or footage going into an edit with its own mix.

Audio to Video

Drop an MP3 or WAV into the Audio slot and the render follows your track's timing and energy instead of generating its own. The file is trimmed in your browser to the clip length before upload, and the render costs a small add-on.

Audio and Loop together

With Loop on, the audio is joined smoothly at the seam. If you uploaded your own track, it only loops cleanly when its length matches the clip — so you can keep the smooth join, or mute the audio and let the picture loop on its own.

Music mode

Switch the create dock to Music to generate a track: Lyria 3 writes 30 seconds, Lyria 3 Pro up to three minutes, both as MP3. Describe the brief in plain language, then open Advanced for genre, mood, instruments, structure, BPM, key, vocals, language, lyrics, and a timed arrangement.

12

Resolution, Frame Rate & Upscale

Output Settings

MotionLab renders at the resolution tier and frame rate you choose, and the price next to the generate button updates as you change them. Duration, resolution and frame rate all move the price; aspect ratio never does. The tiers on offer can change, so trust what the page shows you.

Resolution tiers

720p, 1080p and 4K when available. Test ideas at the lowest tier, then re-render the keeper at the resolution the screen actually needs — an LED wall pixel map rarely wants 4K, a cinema still often does.

Frame rate

Choices run from 12 to 50 fps. 24 or 25 feels filmic and suits atmospheric shots; 30 matches most screens; 40–50 smooths fast camera moves and rapid motion, at a higher price. A Motion Reference video is re-encoded to the fps you pick, so set fps before you upload one.

Upscale Lab

A finished clip can go to Upscale Lab for a MotionLab generative upscale that can also double the frame rate. It re-renders the clip and invents detail, so faces, text and logos may be recreated — use it for atmosphere and texture, not for exact restoration.

13

What Works Well

Strengths

The MotionLab model excels at these types of content:

Cinematic Compositions

Wide, medium, and close-up shots with thoughtful lighting, shallow depth of field, and natural motion.

Emotive Human Moments

Strong single-subject emotional expressions, subtle gestures, and facial nuance.

Atmosphere & Setting

Fog, mist, golden-hour light, rain, reflections, ambient textures.

Stylized Aesthetics

Painterly, noir, analog film, fashion editorial, pixelated animation, stop-motion, claymation.

14

What to Avoid

Known Limitations

To get the best results, avoid these patterns:

Things to Avoid

Re-describing your input image in detailWriting the soundtrack in words (it follows the visuals)A second camera move in the prompt that fights the Camera controlInternal emotions (use visual cues instead)Readable text or logosComplex physics / chaotic motionToo many characters in one sceneConflicting lighting directionsOverly complex prompts (start simple, add detail gradually)
15

Motion Studio (Beta)

Chat-to-Motion-Graphics

Motion Studio is a different kind of tool: instead of generating footage from a model, you describe a motion-graphics piece in chat — a stage opener, kinetic type, an LED loop, a product reveal — and it is composed as live motion code that previews instantly and exports to MP4 from your browser. It works best for graphic, typographic, and geometric motion rather than photorealistic scenes.

•
Describe the piece, not a picture: Say what the composition is, its pacing, colors, and text content: "a dark stage opener, bold white title sliding in, purple accent lines pulsing on the beat."
•
Iterate in conversation: The current composition stays in context — ask for one change at a time: "slow the intro down," "make the background deep blue," "add a subtle grain."
•
Know the boundaries: Compositions run up to 30 seconds and render in your browser. Sound is part of the composition itself — simple synthesized cues or a built-in soundtrack — so ask for it in the chat if you want it. Exports are saved to your Dashboard like any other creation.

Pro Tip: Motion Studio is in beta: behavior and available models may change, and complex requests can take a retry or two. Precise, incremental direction gets the best results.

Production Guides

Practical workflows for screen-based visuals

Studio-grade notes for live visuals, stage screens, and motion tests before the work moves into playback, compositing, or finishing.

How to Make AI VJ LoopsFollow prompt recipes, calculate loop length from BPM, and troubleshoot seams before live playback.LED Screen VisualsPlan visuals for LED walls, strips, banners, safe zones, audience distance, and non-standard screen formats.Image to Video Camera MotionTurn still frames into useful motion references with camera direction, depth cues, light, and atmosphere.