The Technical Challenge
From MODwiki
Navigation Shortcuts
The challenge facing producers and distributors of digital content is to minimize the effort required to prepare video, audio, and multimedia files for distribution across numerous and varied Consumption Profiles (from broadcast to personal computers, wireless connections, PDAs, and cell phones). At the same time, content providers must enhance the end-users Quality of Experience and the Quality of Service that deliver digital content to the eyeballs and ears of consumers.
It is important for the MOD Committee to acknowledge the New Media Ecosystem as recommendations are built for the encoding and transport of video, audio, and multimedia content within and outside of the University for the benefit of students, faculty, staff, and administration. The recommendations must rely on technologies and solutions that can satisfy the demands of the New Media Ecosystem while also supporting the Office of Information Technology’s strategy to embrace centralized coordination with distributed control (that is affordable and feasible to implement for any and all departments and entities within the University).
The members of the Working Group for Digital Video/Audio Architectures & Compression, as well as the entire Committee for Media On Demand, are unanimous in their desire to avoid producing and distributing redundant digital content (especially video and audio) in multiple architectures, file formats, and dedicated players, across a range of bandwidths and Internet connection data rates. Some sort of “universal, cross-platform” solution is desired that obviates the creation of multiple bandwidth files for proprietary WindowsMedia, Real Media and QuickTime architectures. We seek to satisfy the principle of COPE: Create Once, Peruse (or Play) Everywhere.
Contents |
[edit] A Technical Context
[edit] ...from Joan M. Van Tassel
Joan M. Van Tassel, in her book Digital TV Over Broadband: Harvesting Bandwidth (Focal Press 2001), provides a concise context for the technical challenges associated with delivering video, audio and other media over networks and the Internet. Excerpted here are sections from pages 146-150 of her work.
- "Back in 1994 and 1995, when people first wanted to send out audio and video over the Internet, not only did telephone lines seem too narrow to send AV material, but also the speed of 1200 baud and 14.4 kbps modems was too slow to bring the data into the computer. Until 1996, the only way to get AV was to get it with the "download and play" method. This technique means the entire audio or video file is transported across the Internet to the client computer and stored on the hard drive."
- "When the whole file is transferred, the user plays it from the hard drive. This method allows the delivery of reasonably high-quality audio and video, even over a low bandwidth network. Unfortunately, at 1996 speeds, the receiver had to wait for a long time (sometimes hours) for the download to occur, plus it required what at that time was a huge amount of disk storage space. (In 1996, a 450-MB hard drive was state-of-the-art in the consumer market.) Moreover, the entire process raised serious copyright issues, since an actual digital copy existed on users' hard drives."
- "These problems provided incentives for online wizards to develop new techniques to deliver AV material. It was clear that the video either had to be digital originally [born digital], or analog video had to be digitized in order to take advantage of extreme compression. The extreme compression techniques came from work in fractal and wavelet compression, as well as the H.320 family of standards, which advanced to become H.323 and H.324."
- "The technique for sending DV across the Internet or other computer networks is called "streaming." ...Streamed media is played out from a server. On the user side, enough material is buffered to allow smooth playing until more arrives. This lets video and synchronized audio begin playing on the user's machine before the whole file is received. Due to variability in Net congestion, buffering may or may not work, and there can be delays while more material is buffered. Many owners of intellectual property like this kind of streaming because it plays the content almost immediately, and a complete copy may never reside on the receiver's hard drive at any one time. Finally, the streamed video and audio are integrated with text, graphics, and photos, and it all plays inside the Web browser."
- "Smooth-playing video requires that the streamed data arrives continuously, in the proper sequence without interruption. The term used to describe this kind of time-sensitve data is isochronous. "Iso" means equal and "chronous" means related to time. So isochronous means that the time increments must be equal; that is, smoothly delivered."
- "It is no easy task to deliver continuous data over a computer network. This type of network is packet-switched, which means that a message is broken down into packets and sent every which way. They take different paths, arrive at different times and in different sequences than the order in which the original material was arranged. Sometimes packets may lose out to the transport of other, more high priority packets. The way engineers solved this problem was to create a "playout" buffer. The technique involves measuring the amount of average packet delay and then storing just enough material in a buffer to insure that by the time the buffered video has all played, new packets will have arrived that can play."
- "In order to display streamed video, the user needs a "player." This is software that accepts the incoming stream and decodes it. It also includes an interface that lets the user customize the way the stream will be displayed--volume, size of video window, and stop, start, pause, fast forward, and rewind controls, as well as a slider bar to access different parts of the stream. The three most popular players are the RealPlayer from Real Networks, Windows Media Player from Microsoft, and QuickTime from Apple Computers."
- "The biggest [problem] is the inefficient way video reaches users over the Internet. For the most part, every stream of video that a user requests goes out as an individual stream from the server to the user: one request, one user, one stream. If 100,000 users want to see the video, 100,000 streams must be sent out. This method is called "unicast," or "video on demand" (VOD)."
- "It makes sense for little-requested material to be unicast, but it is a wildly inefficient use of bandwidth to unicast popular material. "
- "The alternative to unicasting is "multicasting," a technique that falls between broadcasting and unicasting. Its way of distributing video brings the streams to servers that are closer to end-users, usually the Internet Service Provider (ISP). The ISP notifies people when the material is "live" and they "tune in" to the stream. ...Multicast reduces bandwidth requirements through the network because the minimum number of packets are replicated to service a multiplicity of receivers. Multicasting also reduces connection requirements, which alleviates overhead imposed on the network and the server."
[edit] ...Content Delivery Network
A method that increases the efficiency of delivering large media content with high bandwidth requirements is to deliver the content over separate delivery networks, isolated from the regular network or Internet data transfers. Such systems are referred to as Content Delivery Networks. According to an article found in the Wikipedia:
- "Content Delivery Network (CDN) is a term coined in the late 1990s to describe a system of computers networked together across the Internet that cooperate transparently to deliver content (especially large media content) to end users."
- "CDN nodes are deployed in multiple locations, often over multiple backbones. These nodes cooperate with each other to satisfy requests for content by end users, transparently moving content behind the scenes to optimize the delivery process. Optimization can take the form of reducing bandwidth costs, improving end-user performance, or both."
- "The number of nodes and servers making up a CDN varies, depending on the architecture, some reaching thousands of nodes with tens of thousands of servers."
- "Requests for content are intelligently directed to nodes that are optimal in some way. When optimizing for performance, locations that can serve content quickly to the user may be chosen. This may be measured by choosing locations that are the fewest hops or fewest number of network seconds away from the requestor, so as to optimize delivery across local networks. When optimizing for cost, locations that are less expensive to serve from may be chosen instead. Often these two goals tend to align, as servers that are close to the end user sometimes have an advantage in serving costs, perhaps because they are located within the same network as the end user."
[edit] Proprietary vs Standards-based Media Architectures
It is no surprise that industry players actively engaged in the research and development of digital media architectures and file formats aggressively promote their proprietary technologies, encouraging content providers and end-users to embrace their tools and players. Like most private and public sector entities, the University’s various departments each subscribe to different technologies, depending on their needs and their support and maintenance resources.
The threat to the University, if some consistency is not promoted throughout the delivery of digital media through networks and the Internet, will be a proliferation of files and formats which cannot be played, cannot be shared, and (where appropriate) cannot be re-purposed. None of these conditions are acceptable within an academic institution (except where information must be protected for financial, privacy, patent, copyright, unpublished research, security or other similar circumstances).
[edit] ...The Advantage of a Standard
Because we are witnessing the emergence and wide adoption of the MPEG-4 standardized, non-proprietary media architecture across multiple profiles and consumption points of the New Media Ecosystem, there is an opportunity for the University to promote common methods of content preparation and distribution. The tools and best practices for exploiting the MPEG-4 architecture are either already in place or emerging quickly.
Since many existing installations of servers and media players have been updated to accommodate the MPEG-4 media architecture, adoption of a system-wide standard does not necessarily demand the purchase and implementation of new equipment. In excerpts from a description of MPEG-4, as published by Apple Computer, the architecture…
- "…provides an open playing field. As an open, industry standard, anyone can create an MPEG-4 player or encoder that will work with other manufacturer’s devices."
- "Media companies save time and resources by encoding material once for playback everywhere. No longer will content providers need to encode, host, and store media in multiple formats. Instead, a single format can reach a broad audience equipped with playback devices from not one, but a multitude of companies across a wide array of platforms."
- "Finally, content creators have a format that will reach a global audience and will stand the test of time. While other formats and versions come and go, MPEG-4 will safeguard multimedia content for a secure future."
- "Like MPEG-1 and MPEG-2 previously did for CD-ROMs and DVDs, MPEG-4 promises to create interoperability for video delivered over the Internet and other distribution channels. MPEG-4 will play back on many different devices, from satellite television to wireless devices."
- "To ensure that different products that use MPEG-4 each implement the standard in the same way, Apple, together with Cisco, IBM, Kasenna, Philips and Sun Microsystems, formed the Internet Streaming Media Alliance (ISMA). Other participants include AOL Time Warner, Dolby Laboratories, Lucent Technologies, National Semiconductor, Sony, and 25 other companies. The ISMA defines profiles that companies implement to ensure interoperability."
- "In addition to being adopted by many of the Internet’s premiere content providers, the MPEG-4 standard is receiving tremendous support in other industries. For example, the new standards for high-quality multimedia on wireless devices, 3GPP (3rd Generation Partnership Project) and 3GPP2 (3rd Generation Partnership Project 2), are based on the solid foundation of MPEG-4."
For a more thorough overview of the MPEG-4 architecture, navigate to the wiki page on ...MPEG-4 Architecture & Compressions.
Navigation Shortcuts
- MODWiki Table of Contents (Main Page)
- ...go to The University of Utah Office of Information Technology website
- ...go to The University of Utah Media On Demand Committee Minutes & Reports
- A Quickie Wiki Guide to Editing Content
