How to make a sketchbook animation with stick figures
Tell Vizard Agent the story and name everyone in it. Vizard Agent checks how fast it can produce frames before committing to the style, generates narration and cartoon sound effects, finds a real handwriting typeface for the names, and draws the whole thing as if on a sketchbook page.
What is the short version?
Stick figures on a sketchbook page are the friendliest way to tell a story about real people, because nobody has to be drawn accurately. Names written above the figures do the identifying that a likeness would otherwise have to.
- Go to Vizard Agent and tell the story in plain words.
- Name each character so their name can sit above them.
- Say sketchbook style with stick figures.
What do you need before you start?
The story and the names of the people in it. Vizard Agent draws, narrates and scores the whole thing, so nothing at all needs supplying, and the story can be told as casually as you would tell it to a friend across a table.
- The story. What happened, in order.
- The names. Yours and everyone else's.
- The style. Sketchbook, whiteboard, notebook.
- The length. Thirty seconds to a couple of minutes.
- The tone. Funny, fond, dramatic.
What do you type into Vizard Agent?
Write it the way you would tell it. Vizard Agent turns a plainly told story into scenes, and naming who is in it — "me, Rosalee, a guy called Athren" — is enough for it to keep the figures consistent and label them.
Prompt
Variants worth knowing:
- Handwritten labels. Names in a real handwriting typeface.
- Cartoon sound effects. Landing on the moments they belong to.
- Your own narration. Rather than a generated voice.
What does Vizard Agent actually do?
Here is the order Vizard Agent worked in on a real sketchbook animation about a group of friends. The step nobody would think to take is the third one: it benchmarked its own frame-generation speed before deciding how to build the piece at all.
- Checks the motion video playbook and the available asset creation tools.
- Checks the installed environment tools and the motion-graphics capabilities.
- Benchmarks how fast it can generate frames, which decides the approach.
- Checks the voiceover, music and sound-effect options.
- Generates the narration, the background music and the sound effects together.
- Analyses the narration's timing and transcribes it for exact word timestamps.
- Searches for the best narrating voice and regenerates the narration in it.
- Extracts and inspects the exact word timestamps so the drawing can follow the words.
- Checks the installed handwritten and sketch typefaces, locates the specific handwriting font file, and generates additional cartoon sound effects.
Step 3 is unusual and sensible. A frame-by-frame drawn style is only affordable if frames come quickly, so measuring that first is what decides whether the piece gets built this way at all.
What does the result look like?
A short animation that looks like somebody drew it on a sketchbook page while telling you the story: stick figures with names written above them in handwriting, moving through the events as they are narrated, with cartoon sound effects landing on the moments they belong to.
Nobody is drawn accurately, which is the appeal. A stick figure labelled with your name is funnier and more forgiving than any attempt at a likeness, and it means everyone in the story can appear without anyone being caricatured.
When does this not work well?
The style is deliberately crude, which suits some stories beautifully and quietly undercuts others entirely. These are the cases where Vizard Agent builds exactly what was asked for and the drawing itself ends up working against the material.
- It reads as light. A serious story fights the drawing.
- Frame-drawn styles cost time. Long pieces multiply the frames.
- Detail cannot be shown. Stick figures cannot demonstrate a technique.
- Names must be legible. Handwriting typefaces are harder to read at small sizes.
- Real people are still named. Ask before putting someone's name in a public video.
How do you fix a result that came back wrong?
Say which scene or which character is wrong and how. Vizard Agent keeps the narration timings, all the generated sound effects, the typeface set and the scene drawings, so one part can be redone without rebuilding the animation.
- "That is not what happened next." That scene is redrawn to the corrected story.
- "Her name is spelled wrong." Re-rendered in the same handwriting.
- "The sound effect lands late." Re-timed to the word timestamps.
How does Vizard Agent compare to doing it yourself?
By hand this means drawing every frame, which is charming and slow, or using a whiteboard-animation tool that produces something that looks like every other whiteboard animation. The narration timing is the part that decides whether the jokes land.
| By hand | Vizard Agent | |
|---|---|---|
| Drawing | Frame by frame, by hand | Generated, after a speed benchmark |
| The voice | Record it yourself | Several voices searched, one chosen |
| Timing | Nudge to the narration | Cut to exact word timestamps |
| Typefaces | Whatever looks handwritten | A real handwriting font located |
Common questions
Do I need to draw anything? No. Vizard Agent draws the figures and the scene entirely from your description.
Can it use my own voice? Yes, recorded or cloned. Otherwise Vizard Agent searches for a suitable narrator.
Will the names look handwritten? Yes. Vizard Agent locates an actual handwriting typeface rather than an approximation.
How long can it be? Thirty seconds to a couple of minutes. Frame-drawn styles get expensive beyond that.
Can it include lots of characters? Yes. Vizard Agent labels the figures, which is how the audience keeps track of everyone.
Why benchmark the frame speed first? Because a drawn style is built frame by frame, and knowing how fast that runs decides whether the piece is a thirty-second sketch or a two-minute one before any of it is made.
Does it add sound effects? Yes. Vizard Agent generates cartoon effects and lands them on the word timings.
Is it suitable for a serious story? Rarely. The style reads as light, and Vizard Agent will say when a subject fights it.
Can I put real people in it? Yes, and ask them first. A named stick figure is still a named person.
Can the style carry a series? Yes, and it holds up well. Vizard Agent keeps the typeface, the figures and the drawing style, so a run of stories about the same people accumulates into something recognisable.
What length suits the format? Under a minute for a single story. Vizard Agent benchmarks the frame speed first precisely because that measurement decides whether a longer piece is affordable at all.
Can it animate a story about something that never happened? Yes, and the style suits invention well. Vizard Agent draws whatever you describe, and stick figures carry an imagined scene as comfortably as a real one.
Does the sketchbook have to look like paper? Not necessarily. Vizard Agent builds a whiteboard or a notebook page just as readily, and the choice mostly changes whether the piece feels like a doodle or an explanation.
Does Vizard Agent check the result? Yes. It reviews the scenes and the sound timing before delivery.