Introduction to SMIL
Real Networks was the driving force behind SMIL 1.0, which is now a World Wide Web Consortium (W3C) standard. Real introduced its implementation of SMIL with the introduction of its G2 platform, and the most current SMIL applications were developed using G2 technology. However, the latest version of Apple's Quicktime (4.1) rolled out late last year with SMIL 1.0 support, and Quicktime-based SMIL applications are beginning to take hold. Unfortunately, the Real and Apple Quicktime implementations include proprietary extensions that preclude interoperability.
Microsoft --the other big player in the streaming media space-- has a very different perspective on media integration, and it declined to implement SMIL 1.0 in its Windows Media Technology. We'll come back to Microsoft and the future of SMIL after we review some SMIL basics.
SMIL Basics
SMIL is a text markup language that can establish a spatial framework for displaying media and coordinate a media timeline with multiple sequential and parallel tracks. Here is a simple example:
The <layout> section describes two display regions: one region for an advertisement and another for a video stream. The <body> section specifies that an advertisement --consisting of a GIF image and an audio track that are streamed in parallel-- is played first. The ad is then followed by a streaming video
SMIL 1.0 is a powerful and fairly straightforward language for specifying synchronized multimedia. SMIL also lets you link media elements to other media or to external Web pages. If you're comfortable with HTML, you can easily learn how to "program" using SMIL. It is a simple language that novices and media specialists alike can use to create compelling synchronized rich media presentations.
The fact that SMIL is a simple text specification means that SMIL files can be generated automatically and dynamically with standard Web backend programming and database tools. Earlier efforts to integrate media meant adding event triggers to the contents of a streaming media file (e.g., using rmevents for Real Media files and ASF Index for Microsoft Windows Media Technology ASF files). This approach limited the flexibility of on-demand integrated streaming media presentations. The creative possibilities of dynamically generated SMIL include audio playlists, targeted ad insertion, user customized sports highlight clips, and custom generated instructional content.
SMIL is also useful for overcoming streaming media challenges posed by the Internet's variable bandwidth. You can calculate the overall bandwidth requirements for media-integrated stream presentations, and you can adjust your SMIL files so they are targeted for one or more bandwidths, ensuring smooth playback on users' desktops. SMIL also includes a "switch" statement that can be used to selectively execute alternate streams based on user bandwidth (the "switch" statement is also useful for introducing closed captioning for hearing impaired users and for language switching, for access by non-English speaking users).
Hand coding SMIL, like writing HTML code, is too tedious and time consuming for many. Thus, a variety of GUI SMIL editors have emerged that enable users to integrate and edit multimedia streaming content using a timeline. These tools vary from the limited function RealSlideshow, for quickly and easily creating SMIL-based narrated slide shows, to the much more capable T.A.G., from Extend Media, which enables sophisticated editing and incorporates logic for automatically calculating bandwidth utilization and generating SMIL code that will ensure smooth playback of SMIL synchronized media streams.
Stay tuned for an upcoming streamingmedia.com article that profiles some of the SMIL-based software tools you can use today to integrate streaming media content.
Now, Back to Microsoft and the Future of SMIL
Real Networks spearheaded SMIL 1.0 development. Microsoft was involved during development, but in the end chose not to support SMIL 1.0 because it viewed the spec as a limited, proprietary response to the challenge of synchronizing time-based Web content and interactions. Microsoft seeks a browser-centric synchronization language, and it threw its energies into alternative browser-based synchronization specification, HTML+TIME. Microsoft submitted HTML+TIME to the W3C, but it never made it to recommendation status (the final standardized version of a W3C specification).
Now, Microsoft has joined forces with Real Networks and others to work on a successor to SMIL 1.0. This new version, called SMIL Boston, will add multimedia synchronization features and functions (e.g., improved transitions between data types), and it will extend the SMIL specification to enable the browser-based synchronization of Web elements --such as Javascript and other dynamic HTML components-- that Microsoft had sought with HTML+TIME.
SMIL Boston is currently under development as a working draft, and Microsoft has already implemented a set of the preliminary SMIL specifications in the beta version of Internet Explorer 5.5. A Windows Media ASF file is just one of a number of media and Web elements that can be synchronized within IE 5.5 using SMIL Boston.
Real Networks is also anxious to incorporate SMIL Boston improvements in its streaming products. And while Real Networks has indicated that it is strictly focused on streaming multimedia, many in the industry believe that Real will eventually incorporate Web browser functionality in the Real Player to support increased interactivity and e-commerce functionality -- a notion that fits nicely with the expanded scope of SMIL Boston. Real recently created a custom version of its player for Global Media Corp. that includes HTML support based on the Mozilla open source code. However, Real Networks spokespersons deny any plans for general HTML support in the Real Player.
SMIL has already proven itself as a powerful tool for integrating streaming media content, and it clearly has a future that will keep streaming media enthusiasts smiling for a long time.