XML in Higher Education
SMIL: Multimedia Rides the XML Wave
SMIL (pronounced "smile") is
an acronym for Synchronized Multimedia Integration Language, an XML-based dialect
for describing the layout and synchronization of multimedia applications.
For educators, SMIL opens the door to sophisticated multimedia development.
With minimal effort, SMIL makes it possible for authors to:
- Add audio commentary to images and text
- Animate slide presentations that dynamically change as different elements
become the focus of attention
- Add on-screen controls that allow users to stop and start a presentation
- Create courseware that integrates audio, video, animation, and text
As illustrated in the figure below, individual multimedia components can be
stored either on a user's PC or delivered from a Web server. SMIL presentations
may play in a browser with a SMIL plug-in or in a standalone player such as
RealOne or QuickTime that reside on consumer devices and are independent of
browsers. Because SMIL documents are text files, SMIL files can be customized
on a server manually with a text editor or by using a script, such as AppleScript
or PERL, or through the use of XML transformation tools such as XSLT. What's
exciting for the aspiring multimedia author is that anything that can generate
text can create a SMIL document.
A New Way of Authoring
SMIL's ability to dynamically assemble multimedia represents a departure from
the standard "top-down" approach to authoring adopted by many multimedia presentation
packages. Because SMIL describes how individual multimedia "chunks" are assembled,
SMIL opens the door to collaborative project development, encouraging an exploratory
approach to content creation. With SMIL, authors can update presentation content
without the overhead of expensive and elaborate post-production processing.
If an error is found in any multimedia component, only that one component needs
to be corrected, rather than the entire presentation.
The SMIL Standard
As a W3C standard, SMIL follows a standards process where competing ideas are
hammered out in Working Groups and after agreement, are submitted to the W3C
for approval. Proposed standards, when approved by the W3C, become Recommendations—the
highest approval stamp a specification can receive. SMIL 1.0 was first standardized
as a Recommendation in June 1998. After use and feedback by the community, SMIL
2.0 was proposed, adding new capabilities for timing control, layout animation,
and transition effects. In September 2001, SMIL was officially established as
a W3C Recommendation, allowing companies to safely begin development of compatible
SMIL in a Nutshell
To understand the simplicity and power of SMIL, let's look at the structure
of a SMIL document, which in many ways resembles an HTML document:
Like HTML, a SMIL document has a head and a body, each serving a specific function.
The head is where the layout is specified through the declaration of
screen regions for the display of different media. The body element is
where timing is specified. There are two aspects to timing—things that happen
in sequence and things that happen in parallel. Sequences are surrounded by
the and tags. Media elements defined within
a sequence are presented one after the other—with each element starting up after
the previous one ends. For example, the following simple sequence plays three
audio tracks one after the other:
For media elements with no inherent duration, such as images or text, explicit
durations may be assigned, as in:
Here, the first image ends after being displayed for five seconds, after which
the second image appears and is displayed for seven seconds.
In addition to the various SMIL players there are also a number of SMIL
development environments that help an author assemble a final multimedia
product. These include:
More complex multimedia displays are made possible through the use of the parallel
tag. Elements enclosed in a
tag are started at the same time
and can run to completion or be terminated after a specific time interval. For
example, to play an MP3 audio file while simultaneously displaying a JPEG image
and some text, simply place the three media references in a
the following example, the image and the text are displayed for 30 seconds,
while the audio element ends when the MP3 finishes playing.
Complex multimedia presentations can be built up by combining parallel elements
within sequences. Each parallel group is treated as a single element in its
enclosing sequence. All elements of a parallel group begin together. When the
last element in the parallel group ends, the sequence continues. For example:
In the above example, "leadin.mov" plays first. The narration and the slides
start together as soon as "leadin.mov" ends. When both the narration and the
slides have ended, the credits are displayed.
Dynamic Content Control
What sets SMIL apart from other multimedia presentation schemes is SMIL's support
for dynamic content. Appropriate SMIL media can be selected during a presentation
based on user preferences or hardware capabilities. For example, a presentation
can be easily adapted for international audiences with the use of SMIL's switch
tag, which can trigger different audio tracks based on the setting of a system
In the above example, a SMIL player will select the first item in the list
that matches the user's system language attribute. Similarly, SMIL can select
an item based on connection speed. As the following example shows, an appropriate
audio file will be selected based on the bandwidth capability of the target
SMIL in Practice
When considering SMIL for multimedia applications it's important to keep in
mind that, as with all specifications, the proof is in the implementation. Currently,
Internet Explorer 5.5 and 6.0 support many of the SMIL 2.0 features. SMIL is
also supported by Real Networks' RealPlayer, Apple's QuickTime player, Ovatrix's
GRiNS for SMIL 2.0, and InterObject's SMIL Player.
However, as with any specification implementation, there are compatibility
differences, similar to the problem of rendering HTML code in browsers. While
this makes it difficult for an author to always be able to predict what a SMIL
presentation will do across all implementations, there is considerable momentum
behind SMIL and differences are minimal. Sticking to the basic features of the
SMIL specification is always a safe bet.
The Future of SMIL
The recent finalization of the SMIL 2.0 specification, coupled with significant
industry support, has made SMIL an attractive option for educators. As authors
gain experience using SMIL, expect new ideas to emerge that leverage SMIL's
capacity for delivering dynamic content based on the assembly of individual
multimedia components. Currently, SMIL is stewarded by the SYMM Working Group,
a mix of experts from a wide range of industries including CD-ROM manufacturers,
Interactive TV, mobile communications, and audio/video streaming—all interested
in bringing synchronized multimedia to the Web. A recent initiative includes
bringing SMIL content into the hands of the mobile user via PDAs and even cell
phones. For example, the SMIL 2.0 Recommendation includes a simplified version
of SMIL targeted for mobile devices. Known as the SMIL 2.0 basic profile, it
includes features that map to the limitations of hand-held devices. This is
an exciting development for educators, because it opens the door to the reuse
of those same multimedia chunks used in the creation of desktop multimedia presentations.
XML vs. HTML
While SMIL is often compared to HTML, there are significant differences
between the two tag-based languages. As illustrated in the figure below,
SMIL is based on XML while HTML is based on SGML (Standard Generalized
Markup Language), an XML precursor with a long-standing history in the
document community. Although SGML has been widely used by organizations
seeking to structure their documents and documentation (e.g. the General
Motors Parts Catalog), its pre-Web complexity has been the main obstacle
to widespread use and acceptance by the Web community. That's where XML
The XML proposal, begun in 1995 and approved by the W3C in 1998, represented
an effort to simplify SGML, the ISO standard for defining data vocabularies.
Technically, XML is a subset of SGML designed to simplify the exchange
of structured documents over the Internet through the definition of tags
that add semantic meaning to text. While XML's rules are simple, much
of XML's strength derives more from what it d'es not address.
There are three key design elements that by omission contribute to XML's
- No display is assumed. XML makes no assumptions about how it will
be rendered in a browser or other display device.
- There is no built-in data typing. Ancillary XML technologies such
as DTDs and XML Schema provide support for defining the structure and
data types associated with an XML document.
- No transport is assumed. The XML specification makes no assumption
about how XML is to be transported across the Internet. This has opened
the door to creative ideas about delivering XML over HTTP, FTP or SMTP.
Interestingly, these omissions are what have allowed XML to flourish.
Since XML's approval as a W3C Recommendation, hundreds of XML vocabularies
have been used to standardize information exchanges across a wide range
of industries. Microsoft Corp. has even rebuilt its entire software infrastructure
around XML in the form of its .NET initiative.
Insight into the breadth and scope of XML's reach can be obtained by
www.xml.com and www.oasis-open.org.