theora dev - Aug 2005 - encoding processes uses float or integer ops?

Is the encoding process primarily floating or integer based?

Is there documentation on the encoding process?

I find no mention of the encoding process in the 'Theora I 
Specification'.

Any help, pointers, or suggestions are appreciated. Thanks!

David

On Wed, Aug 10, 2005 at 10:21:23AM -0700, David Oury wrote:
> Is the encoding process primarily floating or integer based?
Integer.
> Is there documentation on the encoding process?
Not really, outside the source code.
> I find no mention of the encoding process in the 'Theora I 
> Specification'.
Yes. As is often the case, the spec is written from the decoder's
point of view to leave plenty of flexibility for encoder designs.

 -r