[theora-dev] Theora performance

[xiphmont at xiph.org]

On Mon, Mar 24, 2008 at 11:11 AM, zhongshan zhang <zhangzs at gmail.com> wrote:
> Hi list,
>
> I just came across this site:
> http://web.mit.edu/xiphmont/Public/theora/demo.html
>
> I was wondering are these issues being addressed or will be addressed before
> the 1.0 release?

The Thusnelda encoder, which is correcting the encoder deficiencies,
is an extensive set of changes and we're likely to release 1.0 without
Thusnelda, then immedaitely switch mainline development to the new
encoder in Thusnelda.

The primary motivation is stability; although the old encoder is not
high-performance, it is very rugged and relaible.  The new Thusnelda
encoder needs more testing to be considered trustworthy.

See the update at:

http://web.mit.edu/xiphmont/Public/theora/demo2.html

Monty