Opencast began as a UC Berkeley-led project to build a global partnership towards the development of an open source lecture capture system. Duke was involved in early discussions in 2007. Opencast received financial backing of the Hewlett and Mellon Foundations, and assembled thirteen partner institutions led by UC Berkeley and ETH Zurich and built Opencast Matterhorn, a suite of open source software to produce, manage and distribute academic audio and video content, especially lecture recordings. In 2015, the “Matterhorn” portion of the name was officially dropped.
Opencast provides hardware and software-based specifications for building your own Linux-based “capture agents,” which are the devices that would be installed in rooms and record connected classroom A/V source feeds. When you set up a recording, you assign it a specific set of processing instructions (output resolution, frame rate, etc.). Users can create their own Capture Agent using Galicaster or Pyca software, or use community-developed ones, such as a Windows Recording software tool (for manual, not scheduled captures) called “TheRec.” TheRec allows recording of as many simultaneous video capture devices as your PC can handle and uses FFMPEG as an encoder, supporting H.264 and free codecs in this version. Users have the option to specify encoding “flavors” (i.e., sets of compression options like resolution and frame rate) and can modify FFMPEG encoding parameters via the UI. TheRec also offers a VU-Meter to show audio levels. A video introducting TheRec can be found here: https://www.youtube.com/watch?v=e1sjFzkwr3M. Additionally, commercial vendors are now beginning to build Capture Agents for Opencast, such as NCast, Teltek, and Extron. The Extron SMP 351 hardware appliance is reviewed in this survey.
In 2015 a major new version was released (version 2.0 on July 17). Since our last survey, Opencast solidified its relationship with the Apereo foundation, which sponsors Sakai among other open source initiatives. The Apereo foundation was formed in 2012 with the merger of Jasig and the Sakai foundations and is dedicated to open technologies with a non-exclusive focus on academia. Duke attended the Open Apereo conference in June, 2015, where the partnership with Opencast was featured.
Core Distinguishing Feature Set
- Scheduling/ automation (single classes, or groups of classes)
- Hold for administrative review option (i.e., recordings are uploaded, queued, and processed but are not made available to viewers until they are cleared via manual review)
- Support for captions
- Administrative tools for monitoring and management of recordings
- Edit/ trim/ delete
- Administrative tabs show recordings in various stages of processing (All, Capturing, Processing, Finished, On Hold, Failed)
- Media gallery for searching and viewing recordings
- Upload of media files to Opencast via a software tool called MHRI (Matterhorn Remote Inbox). A video introducing MHRI can be found here: https://www.youtube.com/watch?v=UGlBYjdZVPI
Recent Updates (from Opencast 2.0: Release Notes)
- Cleaner, re-written administrative user interface
- New HTML5 Engage player – Opencast 2.0 offers a new, accessible video player. It supports keyboard-shortcuts, ARIA profiles support screen-readers and captions. Supports HLS with RTMP for Flash fallback.
- Updated Media Module adapts to different screen sizes, from mobile devices to regular desktop resolutions.
- New FFmpeg-based video editor backend – no more GStreamer dependency.
- New video segmenter based on the FFmpeg select filter. Faster than before and allows detection of scene changes in presenter videos.
- New silence detector – As with the video segmenter, the silence detector has been replaced with an FFmpeg-based implementation.
- New documentation – Until recently, the Opencast documentation was confusing because it was split-up into several wikis and people never knew where to look for a topic. All official documentation can now be found athttp://docs.opencast.org/. The documentation is also included in the source code, so that it is connected with the current state of development. Apart from the official documentation, two wikis still exist. These are the Opencast Adopters Wiki (meant for users to share their guides) and the Opencast Development Wiki (meant for storing working drafts).
Opencast and Duke
Opencast has made great strides and seems to have a solid core of committed community members. However, the main reason it isn’t currently a good fit for Duke is the same reason that kept us from moving forward with the project after initial discussions in 2007–it requires a team of programmers to design, build, and maintain a custom deployment. It is also encouraged that users of Opencast contribute back to the community by developing or helping with the development of parts of the platform. Unlike Duke, many of the groups pursuing this approach have large teams of coders to draw from for such an initiative.