golomb rice codes

use gradients instead of prediction errors as context model
store independant quantization tables for each point
merge contexts with opposit sign

Originally committed as revision 1957 to svn://svn.ffmpeg.org/ffmpeg/trunk

1 files changed, 18 insertions, 2 deletions

diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index d649ea32ef..c2017706b7 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -15,8 +15,8 @@ extern "C" {
 
 #define LIBAVCODEC_VERSION_INT 0x000406
 #define LIBAVCODEC_VERSION     "0.4.6"
-#define LIBAVCODEC_BUILD       4668
-#define LIBAVCODEC_BUILD_STR   "4668"
+#define LIBAVCODEC_BUILD       4669
+#define LIBAVCODEC_BUILD_STR   "4669"
 
 #define LIBAVCODEC_IDENT	"FFmpeg" LIBAVCODEC_VERSION "b" LIBAVCODEC_BUILD_STR
 
@@ -1127,6 +1127,22 @@ typedef struct AVCodecContext {
      * - decoding: unused
      */
     int global_quality;
+    
+#define FF_CODER_TYPE_VLC   0
+#define FF_CODER_TYPE_AC    1
+    /**
+     * coder type
+     * - encoding: set by user.
+     * - decoding: unused
+     */
+    int coder_type;
+
+    /**
+     * context model
+     * - encoding: set by user.
+     * - decoding: unused
+     */
+    int context_model;
 } AVCodecContext;