Mercurial > libavcodec.hg
graph
-
optimize the first vertical lifting step, this also prevents another libavcodecSun, 26 Aug 2007 11:16:23 +0000, by michael
-
get rid of totally senseless "m" + read in register we have enough libavcodecSun, 26 Aug 2007 06:51:51 +0000, by michael
-
avoid overflow in the 3rd lifting step, this now needs mmx2 at minimum libavcodecSun, 26 Aug 2007 01:11:02 +0000, by michael
-
avoid an overflow in the 1 horizontal lifting step libavcodecSat, 25 Aug 2007 19:04:33 +0000, by michael
-
prevent one overflow in the first vertical lifting step libavcodecSat, 25 Aug 2007 16:28:45 +0000, by michael
-
cleanup (remove some old experimentation related code) libavcodecSat, 25 Aug 2007 03:19:39 +0000, by michael
-
we only have 2 wavelets, the 3rd was just for experimentation ... libavcodecSat, 25 Aug 2007 03:04:13 +0000, by michael
-
use 16bit IDWT (a SIMD implementation of it should be >2x faster then with libavcodecSat, 25 Aug 2007 03:00:51 +0000, by michael