This chapter is from the book
This chapter is from the book
This chapter is from the book
Unicast versus Multicast
As described in Chapter 10, streaming technology broadcasts media via either a Unicast or Multicast topology. Recall that in a Unicast mode each user makes a request to the media server to see video and that when they pull the stream they can watch the video file from start; in a Multicast mode many users connect to a single broadcast signal and join the broadcast timeline when they connect.
When broadcasting to mobile devices, you face the same concerns and problems regarding network performance and the user's ability to see a webcast. As described in detail in Chapters 2 and 10, network infrastructures and many compatibility issues (with client players, network protocols, firewalls, and media servers) define our ability to produce a successful webcast. In addition, when we include mobile users in our potential target audience for a webcast, we must address in advance any potential problems they may have, including poor training in how to operate the device for watching streaming media, wrong expectations (thinking it will be similar to a desktop experience), and the inability to see more than text and small images. Even with all these limitations, webcasting a live signal to mobile devices can be a productive experience.
A last important fact to remember when planning to invite mobile users to your webcast is the environment these users may be in during the webcast. By nature, people who use portable devices operate them in transit, while away from their base of operations, where the desktop lives, where information is stored, and where people perform work that requires concentration. We use portable devices the same way we use radios. Radios are used to receive information when we are in transit and to entertain us when we have nothing better to do; we often use them to play music or news in the background. Mobile devices have similar uses. They provide much more than a radio because we can perform many desktop-PC tasks on them, but they have limitations. To successfully attract users to listen to or watch a webcast on handheld devices, the webcast must deliver the message in a proper way and get the full attention of the users to encourage them to participate in such future webcasts.
Scheduled webcasts, regardless of whether they are delivered in Unicast or Multicast mode, can be very costly, especially if you outsource production to vendors, and can be ineffective. It is hard enough to ask people to turn on their desktop PCs for a traditional webcast; it is 10 times more difficult to ask people to turn on their portable devices and join a live webcast. Even though portable devices are much more accessible to users than their desktop computers, you must think about the user's environment. Let us use a cellular phone as an example. We carry our cellular phones wherever we go; because the device is designed to serve us on the move we treat it as a 100% portable product. When we receive an incoming call, we might be driving, walking in the street, shopping, eating dinner, or sleeping; at some of these times it is inconvenient to answer the phone. Handheld PCs are considered to have the same characteristics. I carry my iPAQ PocketPC in my pocket all the time and I can set up my calendar to alert me 15 minutes before a webcast. But what are the chances that I will stop whatever I am doing and listen to a live webcast? If I am at home, the chances are good. But if I am driving, shopping, or dining, the chances are poor.
Media on-demand are the perfect solution for streaming video or audio to mobile devices. Users maintain full control of their schedule and log on to webcasts stored on the media server at their convenience. By giving users the power to decide when and from where to watch a webcast, we guarantee higher attendance, better reception, a better experience, and potential future repeat visits. Whether you transmit the webcast using Unicast or Multicast methods is up to you and the nature of your streaming infrastructure.