In Style Clip, you can make a video from just your own photos and an idea, without picking a template, right?
But when you wrote only one line, you may have thought, "Hmm, that's a bit different from what I pictured."
The StyleRoom video team gathered the storyboard tips we learned while building video templates and checking the real results.
We'll show real Style Clip template videos along the way!
- PrincipleVideo AI draws what's written, and fills in what isn't on its own.
- TimeOne move, glance or reaction every 0.5–1 second, with start time and how long it holds.
- SceneWrite moves, outfit, camera, place and background motion in words you can see.
- HelperEven a short note gets expanded to this level of detail by the storyboard helper.
01 · BasicsGood to know first
Video AI draws what's written exactly, and fills in what isn't on its own.
Those filled-in parts are where a video feels "flat" or "not what I imagined."
- Write briefly, and the AI decides. With just "a pretty pose in front of a café," the AI picks which pose, at what second, where and who films it.
- Visible things over mood words. Instead of "emotional" or "trendy," write what moves where and how. The scenes create the mood.
- No need to repeat what the photo already shows. For the face and body, "as in the photo" is enough.
| What to set | Written briefly | Written in detail |
|---|---|---|
| Time | Pose for 10 seconds | Looks back at 1.5s and holds for 1 second |
| Move | A cute pose | Right-hand V sign beside the cheek, head tilted slightly left |
| Camera | (not written) | A friend holds it at waist height, full body in frame |
| Place | An alley | Red brick pavement underfoot, a white low-rise and a utility pole behind |
| Light | (not written) | 4 p.m., warm sunlight coming in low from the left |
02 · TimeTime: one thing every 0.5–1 second
This makes the biggest difference.
If you give a long stretch just one move, the AI stretches or rushes the timing.
- One event every 0.5–1 second. A 10-second video holds about 8–15 moves, glances and reactions.
- Start time and hold time together. Like "around 4.8s, touches her cheek with a fingertip and smiles, holding for about 1 second."
- An eye-catching moment in the first 1–2 seconds. Put one at the start, like turning around suddenly from behind, and people watch to the end.
- Decide the ending. Stopping while looking at the lens, walking out of frame: write your ending. Without one, the video just fizzles out.
Templates with a first-second hook, quick beats and a planned ending
03 · MoveMoves: which part, how much
In each move line, include which body part + which direction + how much + the ending pose.
- Decide left hand or right hand, left side or right side. "Right hand" is less confusing than "one hand."
- For turns, include the in-between views. Writing "one full turn to the left, side → back → side → front again" means fewer half turns.
- Reasons and reactions between moves. Small triggers like "a scooter passes, she follows it with her eyes and smiles to herself" make it look like a real moment.
- Speed is up to you, too. For slow motion, write "this segment is slow motion" in that segment. Otherwise it plays at real speed.
Templates where moves with a set hand, direction and size follow one after another
04 · OutfitOutfit: move so it shows
If the video is about the clothes, look at the outfit in your photo and pick moves that show that outfit well.
| What to look for in the photo | Example moves |
|---|---|
| Material | Light fabric settles slowly after a turn, and a heavy coat's hem swings weightily with each step |
| Length and fit | A long coat's hem follows half a beat behind when walking, and a cropped top shows the waistline when the arms go up |
| Details | Pleats fan out and close on a turn, hands go into pockets if there are any, and a collar gets straightened by hand |
| The side that catches light | By a window, turn toward the light so the fabric texture shines |
- Set the size of the movement. Like "the pleats open slightly and gather back." Without a size, it tends to come out exaggerated.
- After moving, back to the photo. Adding "returns to its shape in the photo" keeps the outfit from changing.
- Don't write about what the photo doesn't show. The back and the lining are where the AI makes things up.
- No need to rewrite colors, logos or lettering. Just name the material and shape (pleats, long length) briefly.
Templates where the outfit moves in a way that suits its material, length and details
05 · CameraCamera: who films, and how
Without camera notes, the AI films like a slick ad or zooms anywhere.
Set it in one line at the start of each scene.
- Who films: my own selfie / a friend holding the phone / a propped-up phone / a mirror selfie. For a selfie, adding "the arm holding the phone peeks in at the bottom edge" makes it feel like a real selfie.
- Height and framing: waist or chest height; full body, above the knees or upper body. If the full body must show, write "always in frame from head to toe."
- Shake: natural handheld shake / smooth as if on a rail / completely fixed.
- Be specific about a circling camera. "Same height and distance: right side at 3s, back at 5s, left side at 7s" gives a smooth orbit without zooming.
- On a propped-up phone, a bigger frame means you walked closer. Write "the model walks toward the lens," not "the camera moves closer."
The camera circling the model, and a fixed camera with the model turning
How the frame changes with who films where: a mirror selfie and a street shot
06 · PlacePlace, light and small background motion
One line for the background gives you a cramped, awkward space.
Go in the order underfoot → just behind → far background → top of the frame, with each object's material, color and position.
- Keeping the place in your photo? Write only what the photo shows. Adding props that aren't there makes the AI change the space.
- Light means direction, softness, time of day and color. Like "soft morning sun through the window on the right, creamy highlights."
- For signs, only shape and color. Writing the words makes the AI draw strange lettering.
A frozen background looks like a person pasted onto a photo.
Pick 2–3 things already in the scene and let them move a little.
| Place | What can move |
|---|---|
| Streets and alleys | Cars and bikes passing blurry in the distance, people walking, swaying trees and power lines |
| Parks and terraces | Leaves and their shadows, passing cloud shadows, birds flying by |
| Waterside, stations, airports | Glittering ripples, trains and travelers passing in the distance |
| Indoors | Dust in the window light, gently shifting light, curtains and plant leaves already there |
Writing size and distance together, like "small and blurry far behind the model," keeps them from covering the subject.
Setting a time, like "around 6s a cloud passes and the shade comes and goes," also gives the video a sense of time passing.
Passing cars and people, and light from one side, bring the scene to life
07 · OutfitsWhen you have several outfits
- Pair image numbers with segments. Like "0–2.5s: the outfit from Image 1; at 2.5s the scene changes to the outfit from Image 2." One image per segment.
- Keep each outfit's notes inside its segment. If they leak into another segment, the outfits can mix.
- Decide how they switch. Choose from covering the lens with a hand and uncovering it, a full turn, or a hard cut.
- Always set the first outfit. Leave it blank and the AI dresses the model in anything.
Three ways to switch outfits by pairing image numbers with segments
08 · WordsPhrasing that works well
Video AI draws words as they are.
Even if you write "does not ...," the word itself can make the thing appear.
| If you write | Try this instead |
|---|---|
| No other people appear | The model is alone |
| The camera does not shake | The camera is fixed |
| Without cuts | One continuous shot from start to finish |
| Pretty, cool, emotional | Visible moves, light and place |
| Swings her arm | Covers the lens with her palm (make the purpose visible) |
The only "does not" you need is the final line: "No text, subtitles, logos or watermarks."
The aspect ratio is chosen in the app, so you don't need to write it.
09 · ExampleOne line → storyboard example
We expanded the one-liner "walking in front of a café with a coffee, then looking back" using these tips.
The idea stays the same; only the detail goes up.
Compared with the one-liner, more than 10 moves are in, each with a time.
Moves tied to the outfit like the coat hem, a small reaction to the bicycle, the moments she looks at the lens and the ending are all set.
TIP. Writing at this level every time takes effort, right? Even a short note gets expanded like this by the storyboard helper. Just write down the scenes you want to keep!
10 · CheckBefore you generate
- The scenes you want run in time order from 0 seconds to the end
- Each scene has a move, glance or reaction every 0.5–1 second, with hold times
- You decided when she looks at the lens and how the video ends
- You wrote who films, from what height and at what framing
- You described the place from the ground up, including light direction and time of day
- You added moves that show the outfit, and it returns to its shape in the photo afterward
- With several outfits, image numbers are paired with segments and the first outfit is set
- "Does not" appears only in the final line
11 · FAQStyle Clip storyboards: frequently asked questions
Do I have to write a long storyboard?
No.
Even a short note gets expanded by the storyboard helper.
But if there's a scene you must keep, write that part yourself.
It will be kept as written.
Should I describe the face or clothing colors again in text?
No.
The face, body, clothing colors and logos are already in your photos.
Describing them again can clash with the photo if it's even slightly off, so just name the material and shape briefly.
Can I ask for slow motion or hair blowing in the wind?
Yes.
Just write the effect you want in that segment.
Put the segment and the size together, like "3–7s: this whole segment is slow motion, the hair blows strongly in the wind."
When should I write "does not ..."?
We recommend using it only in the final line: "No text, subtitles, logos or watermarks." Everywhere else, describing what you want, like "the model is alone" or "the camera is fixed," works better.
It draws what you write. Write what you want to see.
Make your first video from your own storyboard in Style Clip today.