From b30b19125098e907ee60e823317a5c6f01138c82 Mon Sep 17 00:00:00 2001 From: nicodvb Date: Mon, 27 Jun 2005 19:10:11 +0000 Subject: -don't encode more audio than needed; -edl_skip is int, not short; -don't read audio_data to skip in mux_a->buffer; -edl_seek works on input streams, not output; -one-frame accuracy fix ; patch by Oded Shimon git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@15832 b3059339-0415-0410-9bf9-f77b7e298cf2 --- mencoder.c | 30 +++++++++++++++++++----------- 1 file changed, 19 insertions(+), 11 deletions(-) (limited to 'mencoder.c') diff --git a/mencoder.c b/mencoder.c index 1e8251a400..f252ba2bbc 100644 --- a/mencoder.c +++ b/mencoder.c @@ -239,7 +239,7 @@ static edl_record_ptr next_edl_record = NULL; ///< only for traversing edl_recor static short edl_muted; ///< Stores whether EDL is currently in muted mode. static short edl_seeking; ///< When non-zero, stream is seekable. static short edl_seek_type; ///< When non-zero, frames are discarded instead of seeking. -static short edl_skip; ///< -1 OR the value of in_size of an already read frame. +static int edl_skip; ///< -1 OR the value of in_size of an already read frame. /** \brief Seeks for EDL \return 1 for success, 0 for failure, 2 for EOF. */ @@ -1045,7 +1045,7 @@ goto_redo_edl: result = edl_seek(next_edl_record, demuxer, d_audio, mux_a, &frame_time, &start, mux_v->codec==VCODEC_COPY); - if (result == 2) break; // EOF + if (result == 2) { at_eof=1; break; } // EOF else if (result == 0) edl_seeking = 0; // no seeking else { // sucess edl_muted = 0; @@ -1091,6 +1091,8 @@ if(sh_audio){ } else tottime = 1./audio_density; + // let's not output more audio than necessary + if (tottime <= 0) break; if(aencoder) { @@ -1169,8 +1171,6 @@ if(sh_audio){ audiosamples++; audiorate+= (GetTimerMS() - ptimer_start); - // let's not output more audio than necessary - if (stop_time(demuxer, mux_v) != -1 && stop_time(demuxer, mux_v) <= mux_a->timer) break; } } @@ -1592,6 +1592,8 @@ int edl_seek(edl_record_ptr next_edl_record, demuxer_t* demuxer, demux_stream_t sh_audio_t * sh_audio = d_audio->sh; sh_video_t * sh_video = demuxer->video ? demuxer->video->sh : NULL; vf_instance_t * vfilter = sh_video ? sh_video->vfilter : NULL; + int done = 0; + int samplesize, avg; if (!sh_video) return 0; if (sh_video->pts >= next_edl_record->stop_sec) return 1; // nothing to do... @@ -1610,6 +1612,10 @@ int edl_seek(edl_record_ptr next_edl_record, demuxer_t* demuxer, demux_stream_t // slow seek, read every frame. + if (sh_audio->audio.dwScale) samplesize = sh_audio->audio.dwSampleSize; + else samplesize = (sh_audio->wf ? sh_audio->wf->nBlockAlign : 1); + avg = (sh_audio->wf ? sh_audio->wf->nAvgBytesPerSec : sh_audio->i_bps); + while (!interrupted) { float a_pts = 0.; int in_size; @@ -1620,25 +1626,27 @@ int edl_seek(edl_record_ptr next_edl_record, demuxer_t* demuxer, demux_stream_t if (sh_audio) { a_pts = d_audio->pts + (ds_tell_pts(d_audio) - sh_audio->a_in_buffer_len)/(float)sh_audio->i_bps; - while (sh_video->pts > a_pts) { + while (sh_video->pts - *frame_time > a_pts) { int len; - if (mux_a->h.dwSampleSize) { - len = mux_a->wf->nAvgBytesPerSec * (sh_video->pts - a_pts); - len/= mux_a->h.dwSampleSize; if(len<1) len=1; - len*= mux_a->h.dwSampleSize; + if (samplesize) { + len = avg * (sh_video->pts - a_pts - *frame_time); + len/= samplesize; if(len<1) len=1; + len*= samplesize; len = demux_read_data(sh_audio->ds,mux_a->buffer,len); } else { - len = ds_get_packet(sh_audio->ds,(unsigned char**) &mux_a->buffer); + unsigned char * crap; + len = ds_get_packet(sh_audio->ds, &crap); } if (len <= 0) break; // EOF of audio. a_pts = d_audio->pts + (ds_tell_pts(d_audio)-sh_audio->a_in_buffer_len)/(float)sh_audio->i_bps; } } - if (sh_video->pts >= next_edl_record->stop_sec) { + if (done) { edl_skip = in_size; if (!framecopy || (sh_video->ds->flags & 1)) return 1; } + if (sh_video->pts >= next_edl_record->stop_sec) done = 1; if (vfilter) { int softskip = (vfilter->control(vfilter, VFCTRL_SKIP_NEXT_FRAME, 0) == CONTROL_TRUE); -- cgit v1.2.3