At present, the slickest is unquestionably the $149 box from Vudu The Vudu box is actually a rather interesting case. From the start, their marketing team wan

At present, the slickest is unquestionably the $149 box from Vudu The Vudu box is actually a rather interesting case. From the start, their marketing team wanted to be able to offer relatively high-quality 1080p movies, with audio, at 3 megabits per second--enough to play in realtime on an ordinary broadband connection without prebuffering. In almost any other situation, this would have been a total disaster: 3 megabits for 1080p is practically impossible with current technology. However, a few things saved them: 1. They realized that you don't actually need to use constant bitrate for streaming. If you have a local hard disk for buffering, you can let the bitrate variance increase over time by slowly buffering ahead of the main movie, so that once you're a few minutes in, you have a very large buffer within which to allow variable bitrate. This is the extra benefit of having local storage: a normal VOD service forces you to use constant bitrate for the entire stream. 2. They used H.264 video encoded with x264. This is the best case I've ever seen of the fortunes of a company practically resting on a single open source program. They tried all sorts of solutions, and x264 was literally the only option that gave reasonably viewable output at the rates their marketers insisted on. (Disclaimer: I have worked for Vudu and I develop x264 ;) ) IMO, this is going to be one of the make-it-or-break-it factors in the video on demand business: can you provide good enough video quality to satisfy people? Nobody cares much about quality when watching Youtube, but in order to convince someone to pay for your content, the bar is vastly higher; you can no longer get away with your streams looking awful. Many VOD services have failed because of this: if someone is going to pay the price of a DVD rental to watch a movie, they want it to look as good as a DVD. This is exacerbated by the fact that the initial adopters of such services are going to be more technical users who are willing to spend significant amounts of money on their home entertainment system: the same people who will be turned off by Youtube-like video quality. This is also the challenge of competing with television services: while companies like Comcast prove again and again that having tons of bandwidth doesn't mean you can't completely cock up your video quality, it's vastly easier to do a good job when you have a gigantic pipe than when you're running over someone's low-end DSL.