How to turn a timestamped script into kinetic typography
Paste your script with its timestamps and tell Vizard Agent you want bold kinetic typography. Vizard Agent generates the voiceover, pulls timings for every spoken word, renders the type scenes in batches, reviews the frames of each one, and assembles a piece where the words land as they are said.
What is the short version?
Kinetic typography lives or dies on which words get emphasised and when. A script with timestamps already contains that decision, so the job is to honour it — every emphasised word arriving on the frame where it is spoken, not near it.
- Go to Vizard Agent and paste the script with its timestamps.
- Say you want bold kinetic type and give the frame shape.
- Point at your previous videos so the style matches.
What do you need before you start?
The script and a style reference. Vizard Agent generates the voice and builds the type itself, so what it needs from you is the words, the emphasis, and something to look at so the piece belongs with the rest of your output.
- The script. With your own timestamps.
- The emphasis. Which words carry weight.
- The frame. Vertical for social, wide for a channel.
- Past work. A previous video or a brand image.
- The voice. Coaching, authoritative, conversational.
What do you type into Vizard Agent?
Paste the script as you wrote it, timestamps included. Vizard Agent reads those as the intended structure rather than as formatting to strip out, so the scene breaks in the finished piece follow the ones you already decided on.
Prompt
Variants worth knowing:
- Emphasis words called out. Set apart in size or colour.
- A house style carried over. Fonts and captions read from past work.
- A music bed underneath. Rather than a full score.
What does Vizard Agent actually do?
Here is the order Vizard Agent worked in on a real kinetic type piece from a timestamped script. The habit worth copying comes late: it rendered the scenes in batches and then looked at the frames of every scene individually before assembling anything.
- Loads the ledger, the playbook and the available tools, then the motion-graphics documentation.
- Finds the motion playbook files and views the client's uploaded brand image.
- Samples the past style, typefaces and captions from previous deliveries.
- Looks at a recently delivered frame and at the existing advertisement style.
- Searches for coaching voices and music beds, then generates the voiceover.
- Measures the voice and music loudness and reads the word-level timings.
- Prints every spoken word with its timing so each scene can be placed exactly.
- Renders the first six kinetic type scenes and extracts review frames from them.
- Reviews every scene's frames individually — the hook, each emphasis word, each turn — then renders the rest.
Step 9 is what stops a kinetic piece falling over. A scene that looks fine in a description can read badly on the frame: the word wraps, the emphasis lands on the wrong line, the colour fights the background. Only looking at it catches that.
What does the result look like?
From the run this page is written from, probed on the delivered file: 1080x1920, H.264, 30fps, 56 seconds, AAC audio. Vertical, just under a minute, a dozen kinetic type scenes cut to word-level timings with the emphasis words set apart, in the same typefaces and colours as the client's previous videos.
Fifty-six seconds of typography is a lot of individual scenes. That is why they were reviewed in batches rather than at the end: one bad scene in twelve is easier to fix before the other eleven have been built around it.
When does this not work well?
Kinetic typography suits some material and actively fights other material, and no amount of care from Vizard Agent makes a long script full of subordinate clauses work in this format. These are the limits that come from the writing rather than from the build.
- Long sentences do not fit. Kinetic type wants short, punchy lines.
- Every word emphasised is none emphasised. The contrast has to be earned.
- Dense scripts run fast. A minute of text is not a minute of readable frames.
- No style reference means a default. With nothing to match, it picks for you.
- Your claims are yours. What the script asserts is your responsibility.
How do you fix a result that came back wrong?
Name the scene that is wrong and say what is wrong with it. Vizard Agent keeps the word timings, every rendered scene, all the review frames and the sampled house style, so one scene can be re-rendered without disturbing the eleven around it.
- "Scene four wraps badly." Re-rendered with the line break moved.
- "Wrong word emphasised." Re-set against the same timing.
- "It does not match our other videos." Restyled from the sampled past work.
How does Vizard Agent compare to doing it yourself?
By hand this is a scene-by-scene build in a motion tool: type each line, choose the emphasis, keyframe it, then nudge everything when the voiceover turns out to run a second longer than planned. A minute of it is most of a day.
| By hand | Vizard Agent | |
|---|---|---|
| Timing | Nudge keyframes to the voice | Cut to printed word timings |
| Building scenes | One at a time, manually | Rendered in batches |
| Checking | Watch at the end | Every scene's frames reviewed |
| Style | Rebuild from memory | Sampled from your past deliveries |
Common questions
Do I need to supply timestamps? Not necessarily, but Vizard Agent follows them exactly when you include them.
Can Vizard Agent generate the voice? Yes. It searched for a suitable voice and generated the narration on this run.
Will it match my existing videos? Yes, if you point at them. Vizard Agent samples the typefaces, colours and caption style.
How long can a kinetic piece be? A minute works well. Beyond that Vizard Agent is fighting a format that tires the reader.
Can I mark which words to emphasise? Yes. Vizard Agent sets those words apart and lands them on the exact word timings.
Why review every scene individually? Because a scene can be technically correct and still read badly — the line wraps awkwardly, the emphasis falls on the wrong word, the colour disappears into the background. Those only show on the frame.
Does it work in widescreen? Yes, though Vizard Agent gets better results vertically, because the lines stack naturally.
Can I use my own voice? Yes, cloned or uploaded. Vizard Agent times the scenes to it either way.
What if my script is too long? Vizard Agent will say so. A minute of speech is more text than a minute of readable type.
Can I reuse the style on the next script? Yes. Vizard Agent samples the delivered piece the same way it sampled your earlier work, so a run of kinetic videos settles into one look instead of drifting.
What makes a script work in this format? Short declarative lines with one idea each. Vizard Agent can set almost anything, but the pieces that land are the ones where each scene holds a phrase a reader can take in at a glance.
Does Vizard Agent check the finished piece? Yes. It reviews the assembled scenes and the audio balance before delivery.