Revision: 32144
          
http://projects.blender.org/plugins/scmsvn/viewcvs.php?view=rev&root=bf-blender&revision=32144
Author:   schlaile
Date:     2010-09-27 09:37:36 +0200 (Mon, 27 Sep 2010)

Log Message:
-----------
Fix: [#24006] writeffmpeg doesn't flush delayed frames - fix attached
and [#20843] FFmpeg H264 preset gives "Couldn't initialize codec"

Thanks to Leo Sutic for the patch!

Modified Paths:
--------------
    trunk/blender/source/blender/blenkernel/intern/writeffmpeg.c

Modified: trunk/blender/source/blender/blenkernel/intern/writeffmpeg.c
===================================================================
--- trunk/blender/source/blender/blenkernel/intern/writeffmpeg.c        
2010-09-27 07:01:08 UTC (rev 32143)
+++ trunk/blender/source/blender/blenkernel/intern/writeffmpeg.c        
2010-09-27 07:37:36 UTC (rev 32144)
@@ -160,7 +160,7 @@
        pkt.stream_index = audio_stream->index;
        pkt.flags |= PKT_FLAG_KEY;
        if (av_interleaved_write_frame(outfile, &pkt) != 0) {
-               // XXX error("Error writing audio packet");
+               fprintf(stderr, "Error writing audio packet!\n");
                return -1;
        }
        return 0;
@@ -290,7 +290,9 @@
                packet.data = video_buffer;
                packet.size = outsize;
                ret = av_interleaved_write_frame(outfile, &packet);
-       } else ret = 0;
+       } else {
+               ret = 0;
+       }
 
        if (ret != 0) {
                success= 0;
@@ -778,6 +780,69 @@
        return 1;
 }
 
+/**
+ * Writes any delayed frames in the encoder. This function is called before 
+ * closing the encoder.
+ *
+ * <p>
+ * Since an encoder may use both past and future frames to predict 
+ * inter-frames (H.264 B-frames, for example), it can output the frames 
+ * in a different order from the one it was given.
+ * For example, when sending frames 1, 2, 3, 4 to the encoder, it may write
+ * them in the order 1, 4, 2, 3 - first the two frames used for predition, 
+ * and then the bidirectionally-predicted frames. What this means in practice 
+ * is that the encoder may not immediately produce one output frame for each 
+ * input frame. These delayed frames must be flushed before we close the 
+ * stream. We do this by calling avcodec_encode_video with NULL for the last 
+ * parameter.
+ * </p>
+ */
+void flush_ffmpeg(void)
+{
+       int outsize = 0;
+       int ret = 0;
+       
+       AVCodecContext* c = get_codec_from_stream(video_stream);
+       /* get the delayed frames */
+       while (1) {
+               AVPacket packet;
+               av_init_packet(&packet);
+               
+               outsize = avcodec_encode_video(c, video_buffer, 
video_buffersize, NULL);
+               if (outsize < 0) {
+                       fprintf(stderr, "Error encoding delayed frame %d\n", 
outsize);
+                       break;
+               }
+               if (outsize == 0) {
+                       break;
+               }
+               if (c->coded_frame->pts != AV_NOPTS_VALUE) {
+#ifdef FFMPEG_CODEC_TIME_BASE
+                       packet.pts = av_rescale_q(c->coded_frame->pts,
+                                                 c->time_base,
+                                                 video_stream->time_base);
+#else
+                       packet.pts = c->coded_frame->pts;
+#endif
+                       fprintf(stderr, "Video Frame PTS: %d\n", 
(int)packet.pts);
+               } else {
+                       fprintf(stderr, "Video Frame PTS: not set\n");
+               }
+               if (c->coded_frame->key_frame) {
+                       packet.flags |= PKT_FLAG_KEY;
+               }
+               packet.stream_index = video_stream->index;
+               packet.data = video_buffer;
+               packet.size = outsize;
+               ret = av_interleaved_write_frame(outfile, &packet);
+               if (ret != 0) {
+                       fprintf(stderr, "Error writing delayed frame %d\n", 
ret);
+                       break;
+               }
+       }
+       avcodec_flush_buffers(get_codec_from_stream(video_stream));
+}
+
 /* **********************************************************************
    * public interface
    ********************************************************************** */
@@ -888,7 +953,6 @@
        return success;
 }
 
-
 void end_ffmpeg(void)
 {
        int i;
@@ -905,6 +969,11 @@
                audio_mixdown_device = 0;
        }
        
+       if (video_stream && get_codec_from_stream(video_stream)) {
+               fprintf(stderr, "Flushing delayed frames...\n");
+               flush_ffmpeg ();                
+       }
+       
        if (outfile) {
                av_write_trailer(outfile);
        }
@@ -913,8 +982,8 @@
 
        if (video_stream && get_codec_from_stream(video_stream)) {
                avcodec_close(get_codec_from_stream(video_stream));
+               printf("zero video stream %p\n", video_stream);
                video_stream = 0;
-               printf("zero video stream %p\n", video_stream);
        }
 
        
@@ -1195,20 +1264,47 @@
                rd->ffcodecdata.mux_packet_size = 2048;
                rd->ffcodecdata.mux_rate = 10080000;
 
+               /*
+                * All options here are for x264, but must be set via ffmpeg.
+                * The names are therefore different - Search for "x264 to 
FFmpeg option mapping"
+                * to get a list.
+                */
+               
+               /*
+                * Use CABAC coder. Using "coder:1", which should be equivalent,
+                * crashes Blender for some reason. Either way - this is no big 
deal.
+                */
                ffmpeg_property_add_string(rd, "video", "coder:vlc");
+               
+               /* 
+                * The other options were taken from the libx264-default.preset
+                * included in the ffmpeg distribution.
+                */
                ffmpeg_property_add_string(rd, "video", "flags:loop");
                ffmpeg_property_add_string(rd, "video", "cmp:chroma");
                ffmpeg_property_add_string(rd, "video", "partitions:parti4x4");
                ffmpeg_property_add_string(rd, "video", "partitions:partp8x8");
                ffmpeg_property_add_string(rd, "video", "partitions:partb8x8");
                ffmpeg_property_add_string(rd, "video", "me:hex");
-               ffmpeg_property_add_string(rd, "video", "subq:5");
+               ffmpeg_property_add_string(rd, "video", "subq:6");
                ffmpeg_property_add_string(rd, "video", "me_range:16");
+               ffmpeg_property_add_string(rd, "video", "qdiff:4");
                ffmpeg_property_add_string(rd, "video", "keyint_min:25");
                ffmpeg_property_add_string(rd, "video", "sc_threshold:40");
                ffmpeg_property_add_string(rd, "video", "i_qfactor:0.71");
                ffmpeg_property_add_string(rd, "video", "b_strategy:1");
-
+               ffmpeg_property_add_string(rd, "video", "bf:3");
+               ffmpeg_property_add_string(rd, "video", "refs:2");
+               ffmpeg_property_add_string(rd, "video", "qcomp:0.6");
+               ffmpeg_property_add_string(rd, "video", "directpred:3");
+               ffmpeg_property_add_string(rd, "video", "trellis:0");
+               ffmpeg_property_add_string(rd, "video", "flags2:wpred");
+               ffmpeg_property_add_string(rd, "video", "flags2:dct8x8");
+               ffmpeg_property_add_string(rd, "video", "flags2:fastpskip");
+               ffmpeg_property_add_string(rd, "video", "wpredp:2");
+               
+               // This makes x264 output lossless. Will be a separate option 
later.
+               //ffmpeg_property_add_string(rd, "video", "cqp:0");
                break;
 
        case FFMPEG_PRESET_THEORA:


_______________________________________________
Bf-blender-cvs mailing list
Bf-blender-cvs@blender.org
http://lists.blender.org/mailman/listinfo/bf-blender-cvs

Reply via email to