Language-level Support For Universal Usability Of Multimedia Presentation

Vivian Zhang

Department of Computer Science

University of Maryland, College Park, MD 20742 USA

suje@wam.umd.edu

April 21, 2000

Abstract

With the high-speed Internet access from cable modems and digital subscriber lines (DSL), the next step of communication is synchronized multimedia. However universal usability remains an important issue of this computing research field. The document focuses on Synchronized Multimedia Integration Language (SMIL), which is a language-level support to achieve the goal of universal usability of multimedia presentation.

Introduction

Text documents containing images have long been commonplace on the web infrastructure. It provides most of the information. Multimedia documents on the web do not provide new information. However, the integration of multimedia components into a united and adapted presentation can greatly enhance our perception of the information. It also enables adaptation of how information is conveyed to users with different abilities on different systems. But, as we already see, Multimedia documents are more complex and cover more issues. The Synchronized Multimedia Integration Language (SMIL) was designed to schedule multimedia presentations where audio, video, text and graphics are combined in real-time. In June 1998 SMIL was released and becomes a W3C recommendation [6].

Different from Java, with which the benefits of multimedia have long been achievable, SMIL is encoded in a standardized XML-defined format, which is HTML-syntax like. It places multimedia within the developing web framework at a higher level than Java. And rather than having this done with an application, each SMIL element or attribute can be sewn using HTML-like coding, which can be done in a text edit program such as Vi, Notepad. Because it is simple and descriptive, it makes the creation of web multimedia possible for more people than ever before.

Recently much attention has been paid toward making the web widely usable. Users with different abilities in different situations browse the same web. How can the information on the web be encoded once and used simultaneously by sight-impaired users, hearing-impaired users and common users? Multimedia is an answer and SMIL provides facilities for using multimedia to do this.

Analysis

To achieve universal usability, we should deal with diversity of technologies and users [2]. Web content Accessibility Guidelines 1.0 discusses the principles about how to develop diverse abilities, tools and software for all Web users, including people with visual, auditory, physical or cognitive disabilities [3].

1. Web Content Accessibility Guideline 1 states: provide equivalent alternatives to auditory and visual content [3]. In SMIL, there are two types of equivalent alternatives: discrete equivalent and continuous equivalent.

• Discrete equivalents include alt, longdesc, title, author and abstract. Alt specifies a short text equivalent that conveys the same function as the media object. Longdesc designates a more complete description of media content supplementing description provided by alt attribute. Authors can use longdesc to provide description of complex content, such as charts and graphs. Title provides advisory information about the nature of the element.

With these discrete equivalents, text displayed visually benefits users who are deaf as well as most web users.

• The more interesting part of SMIL’s equivalent alternatives which promote accessibility are the continuous equivalents, which have intrinsic duration and contain references to time. SMIL includes two major attributes: caption and auditory description.

A caption is a text transcript of spoken words that provides the same information as a presentation’s audio stream and is synchronized with the video track of the presentation. Captions benefit people who are deaf, or hearing-impaired. They also benefit anyone in a situation where audio track would cause disturbance, or where noise prevents them from hearing the audio track, or when they have difficulty in understanding the spoken language.

On the contrary, auditory description is a synthesized voice that describes the presentation including information about actions, body language and graphics. Auditory descriptions benefit people with blindness, or low vision. They also benefit anyone in an eyes-busy situation, for example, when driving a car.

2. Web Content Accessibility Guideline 4 states: clarify natural language usage [3]. In SMIL, there is a very useful attribute: system-language. The value of this attribute is a comma-separated list of language name as defined in RFC1766 [1]. Additionally, another very important attribute in SMIL, switch, can define two or more alternatives, each with different language. During playback, the browser can choose one based on which is the user’s preferred language. Actually in SMIL almost all the universal usability are achieved with the help of the switch element.

3. In SMIL, there are some important and handy test attributes, which greatly enhance the usability. With those test attributes, the author can adapt the content according to different user and system setting.

• The attribute system-bitrate specifies the approximate bandwidth [6]. This attribute can be used to make a choice based on the user connection to the network. For example, for a slow connection, the author may suggest the user to download the text instead the audio or video.

• The system-screen-size and the system-screen-depth specify the screen height, screen width and the screen color palette in bits required for displaying the element [6]. So the SMIL playback can display the images or videos with the given color depth.

• The system-captions attribute allows authors to distinguish between text equivalents of the audio (intended for those who have hearing-impair or those who need the text) and text intended for a wide audience [6].

• The system-overdub-or-caption determines if users prefer overdubbing or captioning [6]. So users can choose to watch subtitles or listen to overdubbing.

4. In SMIL, the hyperlink can be associated with spatial subparts, which is the same manner as in HTML. The difference is that hyperlinks in SMIL can relate to time. in another word, it can have temporal subparts. For example, the duration of a video clip can be split into to subintervals. A different link can be associated with each of these subintervals. Such link is activated only if the user clicks on the video, during that period. This helps blind people who may not grasp the structure of a presentation by visual clues.

However, Web Content Accessibility Guideline 7 states: Ensure user control of time-sensitive content changes [3]. Because some people with cognitive or visual disabilities are unable to read moving objects. SMIL achieve this goal by two ways. One is to allow the user to control the overall timing of the presentation, allow the user to pause or rewind the presentation. Another way is to provide users with time-independent access to those time-dependent links.

Industrial support for SMIL 

Any technology, no matter how prospective it is, if it doesn’t get support from the industry, the technology is just an empty promise. As we mentioned earlier, SMIL is a relatively new technology, released in June 1998. One year after that, it has got much support from the multimedia software vendors. RealNetworks keep the dominant position in the SMIL players market. Its streaming media system, dubbed G2 is the first commercial product implementing SMIL play standard multimedia contents [4]. QuickTime 4.1 is a SMIL player from Apple [8]. Microsoft announced its new browser IE5.5 (beta version) would also support part of SMIL Modules [9]. The Hypermedia Presentation and Authoring System (HPAS) from Compaq Inc. [10], the GRiNS authoring software from Oratrix Inc. [11], and the LpPlayer from Productivity Works Inc. [12] all support SMIL.

Authoring tools assist designers to add alternative content with less cost [3]. Electricleisureland’s Fluition 1.1 is an authoring tool for the multimedia content. It enables designers to work on full multimedia layout and sequencing capabilities in a visual environment without manually writing code [7]. In addition, there are also Sausage Software’s SMIL Composer [13], Veon Company’s authoring tool [14], etc.

SMIL vs. ASF

ASF, Active Streaming Format, is Microsoft’s all-in-one web multimedia solution: NetShow. It organizes the preparing multimedia content with intended sequence and builds them into one package. So all the elements for the presentation, such as texts, images, audios and videos, are saved into one ASF file [15]. SMIL, however, uses URLs to access elements on the remote servers, which can be different. By this way SMIL could reuse the existing raw materials. On the other hand, since the text field of SMIL is not built into a package, it can be indexed. So the text components of SMIL are searchable by Yahoo, Excite, etc.

As we analyzed in the above section, by using the attributes switch, system-language, SMIL can deliver different version of audio track in different languages or deliver video of different size according to the user’s bandwidth. Simply put, SMIL has more universal usability than ASF.

Future direction and recommendation

In August 3, 1999, The W3c released the first working draft of SMIL Boston [16]. This version divides SMIL functionalities into nine modules. Each of them can be associated with Document Object Model (DOM) [16]. Another goal of SMIL Boston working draft is to encourage accessibility to multimedia for the visually or auditorily impaired people. Still, many issues about the universal usability need to be addressed in the near future.

• Alternative or equivalent content (such as text description of an image) must remain in synchronized with the rest of the presentation. This is because “Alternatives to video and audio content must be synchronized with video and audio tracks. Alternatives that are improperly synchronized may be so confusing as to be unusable” [6].

• Provide the users active interaction and flexible control to alternative multimedia contents.

• Be adaptable to determine which service is more accessible and more appropriate for the user.

Summary

Despite of its newness, the simplicity of SMIL, and its ability to integrate other formats and load balance of presentation elements from different servers makes it a powerful language to support universal usability of multimedia presentation.

References

More References:

http://www.w3.org/TR/smil-boston/refs.html

http://www.xml.com/pub/98/10/htmltime.html

http://www.xml.com/pub/Multimedia

http://www.w3.org/WAI/

http://smw.internet.com/smil/docs/