Conclusions

Our 2007 live video streaming study shows that Videolan.org's VLC is a big winner in the live video streaming technology competition. It is cross-platform both for streaming media producers and for audience. It is very easy to set up for streaming. The image quality is outstanding. There is little rebuffering. The latency still has room for improvement, but is within the tolerable range for broadcasting. It is free. If VLC can be embedded in a Web browser with ease, it will be much more user-friendly. Right now, embedding VLC is possible, but asking thousands of regular Web users to manually install plug-ins to view VLC live streaming is not a realistic solution. VLC is the closest to being an ideal live video streaming technology.

Both our testers and we the authors are highly impressed by what DVTS, the next generation of live video streaming technology, can offer. DVTS is by far the best technology for live video streaming on the Internet 2 speed. The image is of TV broadcast quality and second to none. Its close-to-zero-second latency is appropriate even for real-time communication, let alone broadcasting. It is one of the easiest technologies to set up. Since it takes advantage of the multicast feature on the Internet 2, it saves bandwidth for the signal emitter, thus making live streaming cost effective. Yet, DVTS is still waiting for the multicast feature to be populated to the general public ISP realm to make itself truly relevant to the general public. Like VLC, this amazing technology is free. Making DVTS Web-browser friendly does not seem to be part of DVTS developers' current effort, but, again, to make it relevant to the general public in the future, such effort must be made. Its lame Mac application needs a total revamp.

Flash is a very promising live video streaming technology though it has some room for improvement. All other technologies need to be much more improved to push themselves to the ideal level.

This study has the following limitations:

1. All the cameras point to a projected HD TV signal – not real-life scenes and sound. Though all captured the same scene, not all were from exactly the same angle and distance.
2. Audio was not part of the test due to the nature of the setup.
3. The coordination between audio and video was not tested.
4. Not all the technologies used the same resolution because DVTS and VLC provided only 720x480px resolution while all the rest used 320x240px resolution.
5. No massive simultaneous viewing was tested.

The authors will take into consideration these factors when we repeat this study.