Taking notes from YouTube videos in Obsidian

Turning YouTube learning into usable notes inside Obsidian hinges on one practical problem: getting reliable timestamps or highlights from a watched video into a knowledge system without losing your place. The workflow options range from quick-and-simple embeds to purpose-built plugins, a browser extension that pauses playback for note-taking, and a newer approach that sidesteps timestamps entirely by using YouTube captions.

The simplest method requires no plugins: split the screen between Obsidian and YouTube, then manually create notes while watching. A more “integrated” version pastes YouTube’s embed code into Obsidian so the video plays inside the vault. That integration is convenient for keeping focus, but the embedded player is limited—scrubbing works, yet accidentally clicking back into the code can reset playback position, making it less reliable for precise note-taking.

For a more robust in-Obsidian workflow, the media extended plugin adds a dedicated media player experience. It can embed and play the video without relying on an iframe, and it supports timestamping from the player. A hotkey (control T in the walkthrough) inserts timestamps into the note as the video runs, producing a trail of timestamped snippets. However, clicking those timestamps may open the video in an external browser rather than staying within Obsidian, and timestamp accuracy can drift—one example lands back near the beginning instead of the intended 56-minute mark.

Another Obsidian-first option, timestamp notes (via a plugin), also timestamps while watching, but it renders timestamps as clickable buttons inside the sidebar. That makes it easier to jump around within the same Obsidian workspace. The tradeoff is that the notes are still tied to the video player’s behavior, and the workflow shown doesn’t solve the “pause, write, resume” friction.

That gap is addressed by the Chrome extension e-note. As the video plays, typing triggers an automatic pause, giving time to capture notes without losing context. After finishing, e-note provides an “open management” view with screenshots and timestamped notes; exporting as Markdown sends the material into an Obsidian vault. The downside is that navigation back to exact moments happens in the browser, not directly inside Obsidian, and the export is manual rather than one-click to a reading system.

The most recent and arguably most scalable approach is Readwise Reader, now in public beta with a YouTube integration. Instead of timestamping, it uses YouTube’s automatic captions: as the video plays, caption text highlights track the spoken content. Notes and highlights are then synced into Obsidian via an official Readwise plugin, bringing metadata and searchable text. The limitation is clear—highlights are text-based, not tied to exact playback timestamps.

Readwise Reader also adds an AI layer through Ghostreader, which can summarize caption-derived highlights and generate prompts or questions. In practice, it produces useful summaries and thought-provoking questions, though some tasks (like generating Q&A pairs) can fail depending on the prompt and available highlights.

In the end, the choice depends on where the user is watching and what they need most: timestamped, in-sidebar navigation (timestamp notes) on a laptop; “watch now, organize later” with searchable text (Readwise Reader) on mobile; and e-note when pausing for richer notes matters. The broader goal remains consistent: convert YouTube’s fast, often distracting consumption into structured, retrievable knowledge inside Obsidian.