How to turn a video into Markdown notes for Obsidian, pictures included

If your notes live in Obsidian, a PDF of a video is a dead end. You can open it, but you cannot link to a line in it, search it with the rest of your notes, or add your own thoughts under the part that mattered. What you actually want is a normal note: plain text you can edit, with the pictures of the screen sitting in it.

Markdown is that format, and it is just a text file with a few marks in it, like a # in front of a heading, plus a short line that points to a picture file, and Obsidian is built on it. The good news is that a video note can be ordinary Markdown with ordinary picture links, and Obsidian will show the pictures as long as the files sit where the links say they are.

Is there a free way to get a video into Obsidian?

Yes, for one video at a time, and it is worth doing once by hand so you know what a good note looks like. For a YouTube video it goes like this.

1. Open the video on a computer, open the description and click Show transcript.
2. Copy the transcript lines and paste them into a new Obsidian note.
3. Pause on each slide or screen you care about, take a screenshot and paste it into the note. Obsidian saves each pasted picture as a file in your vault and links it for you.
4. Put a heading above each part so you can find it later.

This works, and for a ten minute video it is fine. It stops working on a long lecture or a coding stream, because you end up pausing a hundred times, and the pictures land in the note in whatever order you remembered to grab them. A video with no captions also has no Show transcript button at all, so step 2 is gone.

What does VideoDoc actually put in the folder?

I built VideoDoc to make this note for me. You paste a public video link or open a file on your computer, choose A document for AI, and it writes everything into one folder for that video. Here is what is inside, straight from my own output folder:

• doc.md, the Markdown note: the title, a short list with the source link, channel, duration and where the transcript came from, then the words in timeline order with the pictures placed between them.
• images/, the pictures of the screen, named by the second they came from, like frame_000015s.jpg.
• doc.pdf, the same thing as a PDF, for sharing or printing.
• transcript.txt, the plain words with timestamps, and meta.json, the details of the job.

Each paragraph of words starts with its time, like 12:40, and each picture is a normal Markdown picture link that points to images/frame_000015s.jpg, with the time written under it. The link is relative, which means it says "look in the images folder next to me" instead of pointing at a full path on my disk. That is the whole reason it works in a vault.

To give you a real size, I ran a three hour coding tutorial from JavaScript Mastery through it. The note came out with 210 timestamped paragraphs and 123 pictures, and the images folder was 4.8 MB. It took about 16 minutes on my PC, because that video had captions and VideoDoc uses them first. A video without captions takes longer, since the app has to listen to the audio itself.

How do I put it in my Obsidian vault so the pictures show?

Copy the whole video folder into your vault, not just doc.md. The note only knows that its pictures are in a folder called images right beside it. If you move the note on its own in File Explorer or Finder, every picture link breaks and you see empty boxes.

1. Open your VideoDoc output folder and find the folder named after the video.
2. Copy that whole folder into your vault, for example into a folder called Videos.
3. In Obsidian, rename doc.md to the video's name. Every VideoDoc note is called doc.md, so after three videos you would have three notes all called doc.
4. Open the note, and the pictures appear in place, between the words.

Renaming or moving things inside Obsidian is safe, because Obsidian can update the links for you when a file is renamed, as its help page on internal links explains. Doing the same in File Explorer is not, because Obsidian never sees the change. The pictures are JPG files, which is on Obsidian's list of supported file formats, so nothing needs converting.

Why is Markdown better than a PDF for notes?

A PDF is a finished page. Markdown is a note you keep working on. For anyone whose notes live in Obsidian, that difference decides how much of the video you will ever use again.

So I keep both: the PDF goes to whoever I share with, and the Markdown note goes into my vault where I actually study from it.

Does it work with Notion or Logseq too?

Logseq and other apps that read Markdown files from a folder work the same way: keep the note and its images folder together. Notion is different, because it imports a copy of your files instead of reading them from your disk. Notion's import help page takes Markdown and zip files, and pictures only come along when the image files are part of the import. So zip the whole video folder and import the zip, then check the first few pictures before you trust the rest.

Quick questions

Pick one lecture you watched this week, make its note, and copy the whole folder into your vault tonight.