i'd be truly amazed if the quality wasn't shit. ratdvd is reducing to roughly a quarter of its original size. you're talking 1GB vs 50 MB. as an example of ratio (and i know it's not comparable), you'd be compressing a 32 MB wav into 400k mp3.
the applications of this seem extremely limited:
EuclidVision testing has focused on what the firm calls “streaming commentator” applications. This is video that shows the head and shoulders of the subject, which makes the current generation of the technology optimized for video conferencing applications, or simple newscasts.
i can only imagine that the above is all it'll be useful for.
We believe that because it's object based, the longer the video . . . the better we'll do,' he said. That's because the compression system can remember objects that appear frequently in the video, such as an actor's face, and can store such images in memory after reading them from the disk just once. Thus, many objects need to be recorded just once in the digital file, instead of every time they appear in the film.
i believe that in a 90 minute movie, an actor's face will seldom be shown from the same angle, in the same lighting conditions, in the same pose. that pretty much goes for everything. i mean, static shots of people talking at the table eating breakfast, it might help a bit. i honestly doubt it's worth it; we already have the technology.
