Theora 1.1 Thusnelda Is Released

[SavageX]

Did anybody notice that there possible more bluray out there with VC1 and not H264? Using Theora is nice, but as long as there is no hw accelleration available it is pretty much useless for 1080p - when you want to be able to play it with a slower system with a cheap gfx card.

Actually Theora is decoding 1080p on a normal machine with like 2 GHz in realtime with an ASM-enabled reference-decoder - it's pretty lightweight. Even the Java decoder (nearly) pulls 720p on my machine, and I haven't really searched for hot spots yet ;-)

[SavageX]

Or make a version 2 or call it something different, instead of sitting around waiting. Why let H264 take the high bitrate best codec crown?

Because for high-bitrate high-resolution Dirac has promise and currently development for Theora is best targeted at being good at where it's most useful: Web video. Xiph.org won't be sitting around doing nothing.

[Kano]

Did anybody notice that there possible more bluray out there with VC1 and not H264? Using Theora is nice, but as long as there is no hw accelleration available it is pretty much useless for 1080p - when you want to be able to play it with a slower system with a cheap gfx card.

[Yfrwlf]

It would be a pretty small (but incompatible) change to solve that problem, but currently it seems most wise to keep compatible with the current bitstream (millions of deployments, plays just fine without giving users a headache) and see what e.g. Google will do with On2 (open-source VP6 or VP7?), to see if Sun's OMS video codec will actually be released and to see how Dirac develops.

Or make a version 2 or call it something different, instead of sitting around waiting. Why let H264 take the high bitrate best codec crown?

[SavageX]

Any way, if a new version of Theora could be written that was for higher resolutions, without the barrier it currently has, without rewriting the whole entire thing and basically creating a whole new codec, that would be cool.

It would be a pretty small (but incompatible) change to solve that problem, but currently it seems most wise to keep compatible with the current bitstream (millions of deployments, plays just fine without giving users a headache) and see what e.g. Google will do with On2 (open-source VP6 or VP7?), to see if Sun's OMS video codec will actually be released and to see how Dirac develops.

[Yfrwlf]

Hell they can offer 1080p or god knows what, just as long as the other options are on the table. But on youtubes side, thats alot of bandwidth to be supported by google ads. Not to mention increased disk space and server cpu usage.

You mean server I/O, decompression is client-side. :P

Any way, if a new version of Theora could be written that was for higher resolutions, without the barrier it currently has, without rewriting the whole entire thing and basically creating a whole new codec, that would be cool. Otherwise I guess Dirac or something else will have to be the future for higher resolution movies.

[L33F3R]

Hell they can offer 1080p or god knows what, just as long as the other options are on the table. But on youtubes side, thats alot of bandwidth to be supported by google ads. Not to mention increased disk space and server cpu usage.

[Remco]

The only thing to consider, IMO, is what YouTube does: offering both 480p and 720p versions of the same video. Everyone's happy that way.

Note that 720p is usually still acceptable for Theora. On the Theora mailinglist, it was calculated that panning faster than 465 pixels per second would get you into trouble. When you hit that limit the quality has to be reduced, resulting in lots of blockiness. Luckily, most material is talking heads (that's what Youtube is for, mostly).

[RealNC]

The only thing to consider, IMO, is what YouTube does: offering both 480p and 720p versions of the same video. Everyone's happy that way.

[L33F3R]

I don't agree. 24mbit connections are widespread, with 8mbit being a very common lower-bound.

i would also like to point out, that the world is not only NA and Europe. Connections in south korea are amazing but to south korea from the USA for example are utter crap. Asia as a whole is much slower. Japan isnt too bad off. African internet is an oxymoron. So as you can see there is alot to consider.