Cybercasting

In the Beginning
Is This Cybercasting?
Cybercasting Proper
The Players
Two Way Communication
The Bottom Line
Cybercasting is still an ambiguous term. It is most commonly associated with the process of streaming visual or audio information through cyberspace (another undefined term). The suffix "casting" insinuates that it is related to the more traditional form of widespread information distribution, broadcasting. In reality, cybercasting is fundamentally different from broadcasting because it not passive. The client must request the information rather than just tune in the broadcast information.

The phrase that best describes this is broadcatching. It was coined by a professor at MIT to explain the server/client communication necessary on the web. The forms of cybercasting that are becoming popular for transmission on the web are beginning be received in a more traditional real time playback format but, they must still be initiated by the client. The fact that the playback and interfaces are getting to be more traditional is good for the user - we like things to look like what we know.

I'm getting ahead of myself here. How did we get this far? Let's look at where cybercasting started and how we got to where we are today.

In the Beginning...

When the first browsers were introduced in 1992 the created a revolution on the web. The web been textual domain in which only the "techno geeks" could be comfortable. Browsers that utilized a friendlier (relatively) graphic interface and made the transfer and display of images possible created a broader audience. The natural progression was to try to improve upon this basic display of images by including sound. And while we're at it why not include the moving image? We can transmit plenty of great visual and audio information through coaxial cable.

The idea seems simple enough to the non-techno nerd but, the differences in transmission methods, media, and protocols makes it much more difficult. The first transmission of images and sound was simply the FTP of a file that, once received, could be processed by some application on the clients machine.

In 1992 a format for the identification of file type was proposed by Nathaniel S. Borenstein at the ULPAA '92 Conference in Vancouver. Borenstein thought that it would be a great idea for people to be able to do "multimedia email." He proposed the creation of MIME (Multipurpose Internet Mail Extensions) extensions. These extensions allow browsers to know what kind of file to interpret. As browsers became more powerful they could process more of the information within the program. Netscape now includes a player for .au and .aif audio files. When combined with the ability to display .gif and .jpg images in a document, this makes for a better, but still very basic, multimedia experience.

The addition of the programming language Java has further increased the capability of web browsing. It enables the browsers to update images, trigger sounds, react to user input, and more. And now there is Shockwave from Macromedia. This new plugin for browsers enables clients to view Director movies. All these improvements add versatility for the presentation of materials on the web but the key question is:

Is this cybercasting?
No. Not really.

There is one fundamental aspect that separates these forms of viewing and listening from cybercasting. Whether it is FTP or a browser that gets the files, they must be received and stored, at least temporarily, by the client. Then they can be processed into something useful to the client. Even Java must load all the information it will need to carry out its duties as an applet (a miniature program).

This is a major shortcoming with Shock wave as well. The client must wait for the information to be received, temporarily stored, and then processed by the plugin program. In fact, at present (Dec. 8, 1995), Shockwave doesn't support linked files. This makes it unable to include Quicktime movies because it treats them as linked files in all Director movies.

Cybercasting Proper

That which separates cybercasting from the other forms of viewing and listening on the web is the streaming of data. Cybercasting servers serve out signals to client machines that request it. These streams are processed in real time. This means that the data is processed as it is received. The information never has to be stored. The players usually utilize a caching system in order to allow more seamless playback but it is not necessary.

The Client

On the client end, the user, there needs to be a player. This player can be designed to decode video, audio, or both. The players are usually free to the user and can be downloaded and installed with relative ease. The companies that support cybercasting ventures are giving the players away. Their revenues will come from the server end through sales and support of their systems.

The Server

The servers stream the data to the client upon request. The data is a digitized and compressed video or audio signal that may be a stored file or a real time translation. The compression format used is one that each specific player is designed to interpret. The most popular currently in use is a form of MPEG.

A likely reason for stored files to be served is in the case of archived material. The Internet Multicasting Service is good example of this. They use a player called Real Audio to serve a weekly interview program. They now support Xing Technologies Streamworks player as well.

The serving of real time audio or video is much more complex process. It requires the encoding and compression of the information instantaneously. This requires expensive hardware in addition to the serving hardware. Video encoding requires a great deal of information in comparison to audio so it suffers more degradation with the current schemes. Real time audio encoding is getting very good with little loss of quality and greatly reduced bandwidth requirements.

The Players

Audio

There are several players available and the numbers are growing. Those companies that currently have players and servers are constantly updating them every time they find a more efficient solution. The updates can definitely show marked improvement.

The Streamworks system made great improvements with their last update. The realtime audio from radio stations like WXYC come through with relatively few dropouts and at very low bandwidth. If bandwidth isn't a consideration they can serve great stereo music sampled at 44Khz (that's CD sampling rate), again with little dropout.

Real Audio is a another player that claims it can play streams back at quality equal to FM radio signals. It is a scalable system that allows the server to adjust for the available bandwidth. The Streamworks server can do this as well.

Truespeechis a more basic player that is meant to be used as an alternative to putting traditional audio files in a web page. The page will link to the Truespeech file and the clients browser spawns the player that decodes the file as it comes in.

Video

Video streaming on the web has a lot of improvement to make. There have been tremendous improvements but, picture quality, size, and frame rate are still inadequate. The amount of information that must be encoded from each frame of video creates a great deal of difficulty. The compression process includes a number of tricks in order to reduce the amount of information that is streamed. The image size is reduced and the number of frames per second are reduced. This makes the video small and jumpy. Combined with the blockiness caused by the compression algorithm, the video has a way to go.

There are two companies that are currently providing video streaming servers and players on the web. The first was mentioned above from Xing Technologies. Streamworks will also stream video but not as nicely as the audio. The second video streaming company is VDOnet with their player called VDOLive. VDOLive looks like the best at the low bandwidth video transmission game. They have the problems listed above but, improved compression, increased computation speeds, and increased bandwidth will all help this situation improve quickly.

Two Way Communications

Cybercasting doesn't need to be a passive experience. This is an issue of enormous importance to mass communication. While the players that were previously mentioned allow the user to watch or listen to a stream of data from an expensive server, there are new devices that allow any computer user with a web connection to send information.

The most common method of upstream cybercasting on the web is CU SeeME. It is a basic teleconferencing package that was created at Cornell University and is available for free. It allows users to send audio and video at about 4 frames per second, depending on bandwidth, to a reflector. The reflector is a server that puts the image and the audio up as if it were a bulletin board. The users that are logged onto the reflector can see and/or hear the others on the reflector. The number of participants is limited by the administrator of the reflector.

This technology was used in a project by the people at SunSITE to cybercast WXYC. This was the first full time cybercasting of a radio station on the web. Users could logon to the reflector and hear WXYC from anywhere in the world. The audio quality was poor compared to the cybercasting capabilities available today but, it focussed attention on the new possibilities and sparked a number of legal questions for communication lawmakers.

CUSeeMe has illustrated the enthusiasm for this type of upstream communication. There is a CUSeeME event guide that list what's happening on the web. There's also a program that can connect two users directly. In the words of its creator,

"Who knows, you may just end up talking to
your soul-mate. (Well, OK, the odds of that
are infinitesimal, but at least you will have
CUSeeMe in common with that other person)."

The new devices that are bound shake up the communications industry are those are specifically made to allow person to person communication over the web. The Internet Phone is an intriguing device that has tremendous implications. Another similar device is the Mac Phone. These devices allow transmission of audio between two parties using no more bandwidth than a medium speed modem. The impact of these systems and those that will surely follow them will not be fully realized for some time but, it could be significant.

The MBONE is a multicasting backbone that was set aside in 1992 for video and audio conferencing. It requires an extremely large bandwidth connection. The MBONE is relatively exclusive as an interactive medium. Those wanting to cybercast their information make a sort of reservation. For those with right machine and connection, monitoring the MBONE is easy. A key benefit is that many machines can monitor the transmission using the same amount of bandwidth as a single machine. It also allows the transmitter to control the dissemination of information by controlling how far the packets of information can travel.

The Bottom Line On Cybercasting

In spite of its current shortcomings, the potential for cybercasting is enormous. It already lets parties communicate orally from anywhere around the world. It allows users to listen to high quality audio from anywhere. And it allows us to see video images - not very good, but understandable - from anywhere.

Cybercasting has overcome the dilemna of storage limitations through real time players. The user no longer has to store the file so the server can retain the quality that is desired. Combined with more effective streaming methods, user can enjoy beautifully clean, stereo audio with little drop out.

The issue of bandwidth is still a limiting factor and is now the primary limiting factor. The bandwidth issue is being attacked in a number of ways that have allowed for impressive progress. The new hardware and software compression/decompression schemes are becoming more efficient, reducing the need for bandwidth. The amount of bandwidth that is physically available is constantly growing and will substantially increase for the general public in the near future.

All of this makes for an exciting future for cybercasting. The future for cybercasting has the potential to be great. Whether it can fulfill that promise will depend on many factors; regulation, bandwidth, dissemination, and the capability for upstream communication. These issues will dictate whether cybercasting is a triumph for world communication and tool that empowers individuals, or just a new way to catch a "Happy Days" rerun.