First batch with Dirac encoded clips

[Fabricio Zuardi]

I have just finished committing some changes in the code and uploaded new clips with 3 different video formats for each size: Dirac/Ogg, Theora/Ogg and H.264/MP4

Dirac encoded clips are experimental, so please let me know if there is something wrong with the downloads or the playback (does any browser even support Dirac on a HTML5 video tag yet?).

The bitrate I've used for the smaller sizes(480x272) was 1000k, and 1700k for the medium sizes(640x360), I've used ffmpeg2dirac and the dirac-research library, I did not used the Schröedinger implementation because even though it was faster, lots of the converted clips were presenting nasty artifacts in comparison with the other, so I assumed that Schrödinger was not quite ready for production (if you think you can help me configuring schro better please let me know).

Here are the new clips to test:

http://alpha.publicvideos.org/set/ace_200812_01

http://alpha.publicvideos.org/set/ace_200812_02

In the process I've also found a bug on Chromium and Google Chrome that forced me to use the dirac version <source> tag as the second option. More details in this ticket I have filled.