Telematic's video module turns a text material into a finished video: script β scenes β frames β voiceover β subtitles β render. This article covers the first and most important stage: the script and the structure.
Where the script comes from
The script is always built from the material's full text (not the short platform post β otherwise the video would come out truncated). Available modes:
- From text β the source text is carefully split into scenes and voiced as is. Predictable: what's written is what you'll hear.
- Hybrid β the AI expands the material into a full narration text (adding transitions, an intro, a conclusion), then splits it into scenes. Good when the source is bullet points or a dry news item.
- Teaser β a short vertical trailer (9:16) based on the full video: the system waits for the main video's finished script and turns it into an announcement. The perfect "long video + Short" pair.
Scenes
A video is made of scenes: each has its own voiceover text and its own frame (an image or a video). You can work on scenes surgically:
- edit any scene's text and regenerate only that scene;
- split a scene β the AI finds a logical boundary itself, and both halves are rebuilt automatically;
- regenerate a scene's frame without touching the others.
Narration pacing
The pacing setting controls the video's "breathing":
- Standard β a neutral breakdown, one scene β one paragraph;
- Dynamic β short scenes and fast cuts (great for Shorts);
- Calm β long scenes, an unhurried rhythm.
You can also set the number of scenes explicitly β the system will rebuild the breakdown to match.
Interstitials
Between logical blocks you can insert interstitials β short branded inserts (a jingle, a bumper card, a subscribe prompt). They can be reused from the library, and the system plans their placement itself based on the script's structure. Toggled on in the studio.
Presets
A preset is a saved bundle of video settings: voice, speech rate, subtitles, music, visual style, volume levels. Presets are created separately for horizontal (16:9) and vertical (9:16) videos, and in a project you can assign default presets for each orientation β then both manual generation and Autopilot will use them automatically.
Drafts and progress
The video editor's state is saved on the server for each material: you can close the tab and pick up later right where you left off. During generation you see a step-by-step log: which stage is running (script, frames, voiceover, render) and how much is done.
What's next
- Voiceover, subtitles and sound β "Video: Voiceover and Subtitles".
- Frames, styles and publishing to YouTube β "Video: Visuals, Formats and YouTube".
- All of the above in one click β "Autopilot".