Excalidraw-Obsidian Feature Walkthrough - Customizing Embeddable Elements

The biggest upgrade is a new way to scale Excalidraw embeddable elements—especially YouTube-style video embeds and Markdown embeds—so they stay readable and usable at extreme zoom levels on an infinite canvas. Instead of shrinking an embed until controls and text become unusable, users can resize selected embeds to match the current canvas zoom, enabling a “fractal” workflow where tiny details remain legible when zoomed in and don’t collapse into clutter when zoomed out.

The walkthrough starts with a practical problem: embedded videos and embedded Markdown documents can be resized manually, but beyond a certain point the result becomes nearly unwatchable—video controls and content scale down relative to the canvas in a way that breaks the viewing experience. The same failure shows up with Markdown embeds: reducing size eventually leaves only the title visible, while the rest of the content becomes effectively inaccessible. The goal is a different behavior—keeping embeds properly proportioned across Excalidraw’s wide zoom range (described as roughly 10% to 3,000%) so users can place detailed notes and media next to specific points in a drawing.

A concrete example shows an overview canvas with a central Excalidraw video embed and surrounding categories linking to other videos and literature notes. When zoomed in, the embedded videos keep correctly sized YouTube controls and play normally; when zoomed out, the Markdown side documents remain editable and readable at their intended relative scale. The key difference is that the embeds are resized to fit the canvas zoom context rather than being forced into a one-size-fits-all scale.

To achieve that, the update adds a command palette action: “scale selected embeddable elements to 100% relative to the current canvas zoom.” The transcript gives an example canvas zoom of 487% and shows that selecting this action resizes the embed so it looks right at that zoom ratio. A parallel action exists for scaling Markdown embeds as well. For users who prefer a UI control, the properties panel includes a zoom-ratio toggle for scaling in and out, plus a one-click option to set the embed to the same “100% relative” sizing.

Beyond scaling, the walkthrough adds richer control over Markdown embed appearance. Embeds can follow the Excalidraw theme automatically (switching between light and dark modes), and users can override defaults to hide the file name, choose whether the embed background follows the canvas background, the element background, or a custom color, and adjust background opacity. Border styling can be tied to the element stroke color or set independently. Field patterns and background transparency are also configurable, enabling fully immersive notes that blend into the canvas.

Finally, the CSS formatting layer expands customization through front matter: canvas candy-style CSS classes can be applied to Markdown embeds to create effects like rotated labels, dashed borders, and shaped note containers (including circular shapes). Settings also allow defaults for new Markdown embeds, and there’s an “immersive embedding” option for Obsidian Canvas support that removes borders. For a truly borderless, transparent embed, the transcript notes a specific requirement: exporting in Markdown view mode with transparent background settings.

Overall, the changes target one core need—keeping embedded media and notes functional and visually coherent across extreme zoom—while adding theme-aware styling and CSS-driven presentation for more expressive visual knowledge management.