Lesson to knowledge-point clips
Cette page n’est pas encore traduite : voici la version anglaise.
For teachers, tutors and coaches who record the same kind of session every week. You hand over a 60-minute lesson and get one short clip per teaching point instead of clips cut by length: “Today’s phrase: break the ice”, “Common mistake: I am agree → I agree”, “Word of the day: reluctant”. Each clip keeps the teacher’s explanation and examples. Language lessons can use title cards, key-term cards and bilingual captions. Screen-recorded lessons prioritize readable code and slides. A recap clip strings every point together, and the lesson’s study notes come as Markdown and PDF.
Before you start
Section titled “Before you start”- The recording. A screen share, a camera recording, or both as two files. With two files Reelfold lines them up by their sound, so they do not have to start at the same moment.
- Your students. Names and faces of students should not end up in a public clip. Reelfold lists lines that name someone before keeping them; hide faces with a mask (below) or crop them out of the recording first.
- A translation model if you want bilingual captions (the default). The translation goes through your AI provider routing (task
translate); without one, choose captions in the lesson’s language only.
In the Mac app
Section titled “In the Mac app”- On Home, pick Lesson → knowledge-point clips (or describe it: “把这节课切成知识点,中英双语字幕,再出一份学习笔记”) and drop the recording, plus the camera file if you have one.
- The plan names the recipe, the layout (with a camera: picture-in-picture), the caption mode and the platforms.
- Approve the points. Reelfold transcribes once and lists the teaching points it found: kind, start and end, title, meaning, key terms. Edit, drop or keep them.
- Rendering makes one clip per point for every platform shape (9:16 and 16:9 render separately, so nothing is cropped later), the recap, and the study notes.
- Review and publish. Every file was checked before you see it (sound, loudness, first frame, captions, cover); a red item blocks the default approval.
With the Claude Code skill
Section titled “With the Claude Code skill”The skill follows workflows/lesson-clips:
python -m vstudio.lesson plan --source lesson.mp4 --out work/points.json --count 6python -m vstudio.lesson sync --source lesson.mp4 --camera cam.mov --out work/sync.json # two files onlypython -m vstudio.lesson render work/points.json --source lesson.mp4 --camera cam.mov --sync work/sync.json \ --layout pip --platforms xiaohongshu:full,douyin --subtitles bilingual --recap --out outpython -m vstudio.lesson notes work/points.json --out out/notes.md --pdf out/notes.pdf --title "Small talk at work"What a clip looks like
Section titled “What a clip looks like”| Part | What it shows |
|---|---|
| Title card (2.2 s) | The kind (“Today’s phrase · 今日短语”), the phrase or point, its meaning in the other language. Also the cover. |
| The explanation | The teacher’s own segment at the original speed, under a header that keeps the point on screen. |
| Key-term card | The term and its meaning, shown for a few seconds when it is first said. |
| Captions | The spoken line with the target phrase highlighted, the translation as a smaller second line. |
Points come in four kinds: phrase, vocab, concept (a rule or a difference) and correction (a common mistake and its fix). Examples attach to the point they illustrate.
Layouts with a camera file
Section titled “Layouts with a camera file”| Layout | What it does |
|---|---|
screen |
The screen share only. |
pip |
Screen share with the camera as a small inset (default when a camera is given). |
band |
Vertical: screen band with the camera band under it. Horizontal: screen plus a camera column. |
camera |
The camera only. |
If the two recordings cannot be lined up by sound, the run stops and asks for the offset instead of guessing.
Options worth knowing
Section titled “Options worth knowing”| Option | Default | What it does |
|---|---|---|
| Captions | bilingual | bilingual, mono (the lesson’s language only) or translated (the translation only). |
| Translate to | auto | The other of Chinese and English; or pick one. |
| Points | 6 | How many teaching points to cut, each 20 to 90 seconds. |
| Speed | 1.0 | Language lessons stay at the original speed. |
| Hide faces | off | sticker or blur on the whole picture, the camera tile or the screen. A face found on fewer than half the frames is flagged red. |
| Recap | on | One clip with a mini card and the teacher’s own line per point. |
Example prompts
Section titled “Example prompts”- “把这节课切成知识点,每条一个短语,中英双语字幕,再出一份学习笔记”
- “Cut this English lesson into today’s-phrase clips with Chinese subtitles only.”
- “Screen share and camera are two files: picture-in-picture, mask my face.”
Related
Section titled “Related”- Bilingual subtitles
- Interview to Q&A clips
- Course slicing (a course master and episodes instead of points)
- lesson-clips workflow reference
Screen recordings, code and slides
Section titled “Screen recordings, code and slides”For screen lessons, Reelfold preserves the source picture and puts burned-in captions in a separate rail, clear of the screen and camera. Added cards that would obstruct the lesson are deferred or omitted. Mono captions and added labels follow the source language. Where the platform allows it, the source orientation is retained unless you ask for another shape. Review technical terms and small-screen readability before publishing.