Course Workshop & Avatars
Explanatory graphics and images
Graphics are created from the narration and reveal themselves in time with the voice-over — as text instead of pixels, hence translatable; images show a scene. Both only as a suggestion, never unasked.
目的#
A graphic that shows exactly what the speaker is talking about explains better than a full slide. And nothing should land on a slide that the author has not chosen.
使い方#
クリック手順CourseContentSlideGenerate graphic
- 1Write the narration first — the graphic follows its structure.
- 2Describe under “What should the graphic or image show?” (e.g. “Three profiles side by side, each with an icon and two keywords”) — or choose “Ideas for a graphic” or “Ideas for an image”. Without a description nothing is generated (図 1).
図 1 “What should the graphic or image show?” with the buttons for ideas, “Generate graphic” and “Generate image”. - 3“Generate graphic” (explanatory graphic from the narration) or “Generate image” (photo or image scene according to the description).
- 4The suggestion appears as a version: “Version n of m”, browse with ‹ and ›, “Accept” or “Discard”. Nothing changes on the slide until “Accept”.
- 5Set reveal markers in the narration at which the next element appears.
- 6In the new slide editor a graphic can be broken down to every label and edited freely.
設定とスイッチ#
機能カタログに独自のスイッチはありません — 周囲の機能に含まれます。
- Part of the module “Course Studio”.
Revealing in time with the voice#
An explanatory graphic consists of stages. Every stage appears at a marker ⟨r⟩ in the narration — without matching markers evenly over the speaking time. In the slide editor this can be dragged and reordered under “Build-up in time with the voice”.
Changing only markers leaves a finished video valid; whoever changes the spoken text invalidates the slide's video — it has to be created again, and that never happens by itself.
Translating#
When the slide is translated, the labels of the graphic are translated too — in one pass with the slide text.
Who may do what#
Every author of the course. A guard in the code ensures that no AI output gets onto the slide past the version bar.
ヒント#
- The graphic explains and reveals along with the narration; it is translatable and stays sharp. The image shows — it is not translated along with it.
- Even after accepting you can browse back to an earlier version with ‹; the versions stay in the browser, a reload does not eat them.
- If a slide comes from the platform's kit and no longer shows the original graphic, the workshop offers “Restore original”.
関連する機能#
- Phase 1 · Content: outline and slidesDefine modules and slides, then work out every slide with narration and presentation — the longest and most important step.
- New slide editor (“Stage”)Edit slides directly on the slide: grab, move, rotate, align every element — with history, named versions and the original, to which you can return at any time.
- Phase 2 · Design: look & soundStyle, logo, signet, surface behind the speaker, background music, transitions and further versions of the course.
背景知識
- Writing narration scripts that sound goodHow to write narration for learning videos, narrated slides and AI voices: short sentences, spoken language, rhythm, pronunciation and alignment with visuals.
- Designing good slides: rules for clear presentationsWhat clear slides look like: one message per slide, little text, strong visuals, clear hierarchy, contrast and legibility — with a checklist and examples.
- Accessible presentations: slides for everyoneHow to make presentations and slides accessible: contrast, type, reading order, text alternatives, captions, colour and language — with a WCAG-based checklist.