From: Thomas D. <tho...@bb...> - 2007-12-12 11:18:10
|
Actually, the specification is broken. Picture data used to be unsigned, and so when no dc value was available for prediction the mid-point was a sensible option. When the picture data became signed, zero became the sensible option. I just forgot to update the spec and the code. Sorry! regards Thomas On 11/12/2007, David Schleef <ds...@sc...> wrote: > > On Tue, Dec 11, 2007 at 01:10:00PM +0100, Wladimir van der Laan wrote: > > This small patch makes that dirac (svn head) can play back streams > created > > by Dirac reference implementation (cvs). It seems schroedinger and dirac > > have a slightly different concept of predicting the motion compensation > DC > > component. It might be useful for people that want to test different > > encoders. > > Good catch. If I'm not mistaken, they're both broken :) > > Fixed now. > > > > dave... > > > ------------------------------------------------------------------------- > SF.Net email is sponsored by: > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Schrodinger-devel mailing list > Sch...@li... > https://lists.sourceforge.net/lists/listinfo/schrodinger-devel > |