From Wikivideos

About the WikiVideos Extension

Read Edit View history
  1. 00:00 Welcome to Wikivideos, the wiki for collaborative educational videos.
  2. 00:05 Like other wikis, such as Wikipedia, you can edit almost any page by clicking on the Edit buttons near the top.
  3. 00:13 However, unlike other wikis, the main element of every page is one or more wikivideo tags.
  4. 00:20 The syntax is similar to gallery tags, with images followed by text.
  5. 00:25 But when you save the changes, the result will be a video, with each image shown while an automated voice reads the text aloud.
  6. 00:33 This will allow us to collaborate building videos, much like we do with Wikipedia articles, using familiar tools and concepts such as diffs and watchlists.

The WikiVideos Extension

IntroductionCollaborative VideoWikiVideos Extension

This video was generated using the WikiVideos extension itself — wikitext syntax producing a video where images are shown while an automated voice reads the text aloud.

Any logged-in editor can create or improve videos like this directly in the wiki, without needing specialised software or powerful hardware.

View on MediaWiki.org →

How does it work?

Wikivideos supports two modes of video content:

  1. Embedded video — pages embed public domain films and educational media directly from Internet Archive and Wikimedia Commons, with curated chapter notes, Wikipedia links, and references.
  2. Generated video via the WikiVideos extension — using the WikiVideos extension, any editor can collaboratively build new educational videos directly in wikitext. The extension translates sequences of images and text into videos, with audio generated using Google's Text-to-Speech service.

For generated videos, the syntax is similar to gallery tags — images followed by text. When the page is saved, the extension produces a video showing each image while an automated voice reads the associated text aloud.

Example

Here is a basic example based on a famous haiku by Matsuo Bashō:

WikiVideos extension — basic example

Wikitext input:

<wikivideo width="300">
File:Old_pond.jpg | Old pond
File:Frog_leap.jpg | frog leaps in
File:Water_sound.jpg | water's sound
</wikivideo>

Each line pairs an image file with the text spoken aloud for that frame. The extension generates the video automatically when the page is saved.

Rendered output:

  1. 00:00 Old pond
  2. 00:02 frog leaps in
  3. 00:04 water's sound

Current features

WikiVideos extension

  • Audio generated using Google Text-to-Speech — any of the supported voices and languages may be used.
  • Chapter timestamps displayed below the video by default for easy navigation. Disable with chapters=no in the <wikivideo> tag.
  • Optional captions automatically generated (click the three-dot menu to enable). Enable by default with captions=yes.
  • Supports JPG, PNG, and GIF image files.
  • Images may have no associated text (shown for one second before advancing).
  • Text may have no associated image (black background shown while text is read).
  • Control display size with width and height attributes in the <wikivideo> tag.
  • Images may be reused directly from Wikimedia Commons.

Embedded video (Archive.org & Commons)

  • Any Internet Archive item can be embedded using its identifier: archive.org/embed/IDENTIFIER.
  • Wikimedia Commons video files are embedded via their direct media URL.
  • Thumbnails are loaded automatically from Archive.org's image service and Wikimedia Commons.
  • No video is hosted on Wikivideos servers — bandwidth is handled entirely by Archive.org and Commons.
  • Chapter notes, Wikipedia cross-links (class="extiw"), copyright status, and academic references are added on the Wikivideos page.
  • Source attribution is shown below every embedded player.
  • All embedded content is either public domain or freely licensed.

Future features

Planned for the WikiVideos extension

  • Support for more file types (SVG, WebM, etc.)
  • Support for limited wikitext in chapter text, including links and references.
  • Per-line attributes to override global settings, e.g. switching voices to simulate dialogues.
  • Automatic captions in multiple languages via Google Translate.
  • Automatic voice synthesis in multiple languages using Google Translate + Text-to-Speech.
  • Support for additional TTS providers (e.g. Amazon Polly).

Planned for embedded video

  • Category pages that browse all videos in a subject area, with thumbnail grids.
  • Series and playlist pages grouping related films, lectures, or episodes.
  • Full-text search across chapter notes and article text.
  • Integration with the Internet Archive search API to discover new content to add.
  • Structured data for each page (director, year, runtime, subject tags) for filtering and sorting.
  • Mobile-optimised layout for the video browse grid.

Why Wikivideos?

  • Videos are an effective and engaging way to share knowledge, and can reach a growing audience of people who learn better through visual media.
  • Videos can reach audiences who are not well served by text-heavy encyclopaedias, including those with lower literacy levels.
  • The WikiVideos extension allows anyone with a browser to create and improve educational videos, without needing specialised software or powerful hardware.
  • Wikimedia projects notoriously lack video content. Wikivideos aims to fill this gap.
  • Unlike commercial video platforms, Wikivideos carries no advertising, runs no recommendation algorithm, and does not track users.
  • All content is available under a Creative Commons Attribution-ShareAlike licence unless otherwise noted. Embedded public domain films remain in the public domain.

More examples

The following pages demonstrate both embedded public domain video and WikiVideos extension-generated content: