diff --git a/src/media.cpp b/src/media.cpp index 218cfff9..1839113e 100644 --- a/src/media.cpp +++ b/src/media.cpp @@ -169,11 +169,13 @@ class FeAudioImp : public FeBaseStream #ifdef DO_RESAMPLE ResampleContext *resample_ctx; #endif - sf::Int16 *buffer; - sf::Mutex buffer_mutex; + sf::Int16 *audio_buff; + sf::Mutex audio_buff_mutex; FeAudioImp(); ~FeAudioImp(); + + bool process_frame( AVFrame *frame, sf::SoundStream::Chunk &data, int &offset ); }; // @@ -236,14 +238,7 @@ FeMediaImp::FeMediaImp( FeMedia::Type t ) void FeMediaImp::close() { if (m_format_ctx) - { -#if (LIBAVFORMAT_VERSION_INT >= AV_VERSION_INT( 53, 17, 0 )) avformat_close_input( &m_format_ctx ); -#else - av_close_input_file( m_format_ctx ); -#endif - m_format_ctx=NULL; - } if (m_io_ctx) { @@ -271,10 +266,7 @@ FeBaseStream::FeBaseStream() FeBaseStream::~FeBaseStream() { if ( codec_ctx ) - { - avcodec_close( codec_ctx ); - codec_ctx = NULL; - } + avcodec_free_context( &codec_ctx ); clear_packet_queue(); @@ -323,26 +315,14 @@ void FeBaseStream::push_packet( AVPacket *pkt ) void FeBaseStream::free_packet( AVPacket *pkt ) { -#if (LIBAVCODEC_VERSION_INT >= AV_VERSION_INT( 55, 16, 0 )) av_packet_unref( pkt ); -#else - av_free_packet( pkt ); -#endif av_free( pkt ); } void FeBaseStream::free_frame( AVFrame *frame ) { -#if (LIBAVCODEC_VERSION_INT >= AV_VERSION_INT( 55, 45, 0 )) av_frame_unref( frame ); av_frame_free( &frame ); -#else - #if (LIBAVCODEC_VERSION_INT >= AV_VERSION_INT( 54, 28, 0 )) - avcodec_free_frame( &frame ); - #else - av_free( frame ); - #endif -#endif } FeAudioImp::FeAudioImp() @@ -350,13 +330,13 @@ FeAudioImp::FeAudioImp() #ifdef DO_RESAMPLE resample_ctx( NULL ), #endif - buffer( NULL ) + audio_buff( NULL ) { } FeAudioImp::~FeAudioImp() { - sf::Lock l( buffer_mutex ); + sf::Lock l( audio_buff_mutex ); #ifdef DO_RESAMPLE if ( resample_ctx ) @@ -366,13 +346,125 @@ FeAudioImp::~FeAudioImp() } #endif - if ( buffer ) + if ( audio_buff ) { - av_free( buffer ); - buffer=NULL; + av_free( audio_buff ); + audio_buff=NULL; } } +// This function frees the frame +bool FeAudioImp::process_frame( AVFrame *frame, sf::SoundStream::Chunk &data, int &offset ) +{ + int data_size = av_samples_get_buffer_size( + NULL, + codec_ctx->channels, + frame->nb_samples, + codec_ctx->sample_fmt, 1); + +#ifdef DO_RESAMPLE + if ( codec_ctx->sample_fmt == AV_SAMPLE_FMT_S16 ) +#endif + { + sf::Lock l( audio_buff_mutex ); + + memcpy( (audio_buff + offset), frame->data[0], data_size ); + offset += data_size / sizeof( sf::Int16 ); + data.sampleCount += data_size / sizeof(sf::Int16); + data.samples = audio_buff; + } +#ifdef DO_RESAMPLE + else + { + sf::Lock l( audio_buff_mutex ); + + if ( !resample_ctx ) + { + resample_ctx = resample_alloc(); + if ( !resample_ctx ) + { + FeLog() << "Error allocating audio format converter." << std::endl; + free_frame( frame ); + return false; + } + + int64_t channel_layout = frame->channel_layout; + if ( !channel_layout ) + { + channel_layout = av_get_default_channel_layout( + codec_ctx->channels ); + } + + av_opt_set_int( resample_ctx, "in_channel_layout", channel_layout, 0 ); + av_opt_set_int( resample_ctx, "in_sample_fmt", frame->format, 0 ); + av_opt_set_int( resample_ctx, "in_sample_rate", frame->sample_rate, 0 ); + av_opt_set_int( resample_ctx, "out_channel_layout", channel_layout, 0 ); + av_opt_set_int( resample_ctx, "out_sample_fmt", AV_SAMPLE_FMT_S16, 0 ); + av_opt_set_int( resample_ctx, "out_sample_rate", frame->sample_rate, 0 ); + + FeDebug() << "Initializing resampler: in_sample_fmt=" + << av_get_sample_fmt_name( (AVSampleFormat)frame->format ) + << ", in_sample_rate=" << frame->sample_rate + << ", out_sample_fmt=" << av_get_sample_fmt_name( AV_SAMPLE_FMT_S16 ) + << ", out_sample_rate=" << frame->sample_rate << std::endl; + + if ( resample_init( resample_ctx ) < 0 ) + { + FeLog() << "Error initializing audio format converter, input format=" + << av_get_sample_fmt_name( (AVSampleFormat)frame->format ) + << ", input sample rate=" << frame->sample_rate << std::endl; + free_frame( frame ); + resample_free( &resample_ctx ); + resample_ctx = NULL; + return false; + } + } + if ( resample_ctx ) + { + int out_linesize; + av_samples_get_buffer_size( + &out_linesize, + codec_ctx->channels, + frame->nb_samples, + AV_SAMPLE_FMT_S16, 0 ); + + uint8_t *tmp_ptr = (uint8_t *)(audio_buff + offset); + +#ifdef USE_SWRESAMPLE + int out_samples = swr_convert( + resample_ctx, + &tmp_ptr, + frame->nb_samples, + (const uint8_t **)frame->data, + frame->nb_samples ); +#else // USE_AVRESAMPLE + int out_samples = avresample_convert( + resample_ctx, + &tmp_ptr, + out_linesize, + frame->nb_samples, + frame->data, + frame->linesize[0], + frame->nb_samples ); +#endif + if ( out_samples < 0 ) + { + FeLog() << "Error performing audio conversion." << std::endl; + free_frame( frame ); + return false; + } + offset += out_samples * codec_ctx->channels; + data.sampleCount += out_samples * codec_ctx->channels; + data.samples = audio_buff; + } + } +#endif + + free_frame( frame ); + return true; +} + + FeVideoImp::FeVideoImp( FeMedia *p ) : FeBaseStream(), m_video_thread( &FeVideoImp::video_thread, this ), @@ -523,6 +615,7 @@ void FeVideoImp::video_thread() AVFrame *detached_frame = NULL; bool degrading = false; + bool do_flush = false; int64_t prev_pts = 0; int64_t prev_duration = 0; @@ -634,6 +727,13 @@ void FeVideoImp::video_thread() // } + if ( do_flush ) + { + // flushed last time we did do_process branch below, so this time we + // exit + goto the_end; + } + if ( do_process ) { if ( !detached_frame ) @@ -647,29 +747,40 @@ void FeVideoImp::video_thread() if ( !m_parent->end_of_file() ) m_parent->read_packet(); else - goto the_end; + do_flush = true; // NULL packet will be fed to avcodec_send_packet() } - else + + if (( packet != NULL ) || ( do_flush )) { // // decompress packet and put it in our frame queue // - int got_frame = 0; -#if (LIBAVCODEC_VERSION_INT >= AV_VERSION_INT( 55, 45, 0 )) - AVFrame *raw_frame = av_frame_alloc(); - codec_ctx->refcounted_frames = 1; -#else - AVFrame *raw_frame = avcodec_alloc_frame(); -#endif + int r = avcodec_send_packet( codec_ctx, packet ); + if (( r < 0 ) && ( r != AVERROR(EAGAIN) )) + { + char buff[256]; + av_strerror( r, buff, 256 ); + FeLog() << "Error decoding video (sending packet): " + << buff << std::endl; + } - int len = avcodec_decode_video2( codec_ctx, raw_frame, - &got_frame, packet ); - if ( len < 0 ) - FeLog() << "Error decoding video" << std::endl; + AVFrame *raw_frame = av_frame_alloc(); + r = avcodec_receive_frame( codec_ctx, raw_frame ); - if ( got_frame ) + if ( r != 0 ) { - raw_frame->pts = raw_frame->pkt_pts; + if ( r != AVERROR( EAGAIN )) + { + char buff[256]; + av_strerror( r, buff, 256 ); + FeLog() << "Error decoding video (receiving frame): " + << buff << std::endl; + } + free_frame( raw_frame ); + } + else + { + raw_frame->pts = raw_frame->best_effort_timestamp; if ( raw_frame->pts == AV_NOPTS_VALUE ) raw_frame->pts = packet->dts; @@ -688,10 +799,9 @@ void FeVideoImp::video_thread() detached_frame = raw_frame; } - else - free_frame( raw_frame ); - free_packet( packet ); + if ( packet ) + free_packet( packet ); } } else if ( !degrading ) @@ -729,7 +839,7 @@ void FeVideoImp::video_thread() int average = ( displayed == 0 ) ? qscore_accum : ( qscore_accum / displayed ); - FeDebug() << "End Video Thread - " << m_parent->m_imp->m_format_ctx->filename << std::endl + FeDebug() << "End Video Thread - " << m_parent->m_imp->m_format_ctx->url << std::endl << " - bit_rate=" << codec_ctx->bit_rate << ", width=" << codec_ctx->width << ", height=" << codec_ctx->height << std::endl << " - displayed=" << displayed << std::endl @@ -753,18 +863,6 @@ FeMedia::~FeMedia() delete m_imp; } -void FeMedia::init_av() -{ - static bool do_init=true; - - if ( do_init ) - { - avcodec_register_all(); - av_register_all(); - do_init=false; - } -} - sf::Time FeMedia::get_video_time() { // @@ -903,7 +1001,6 @@ bool FeMedia::open( const std::string &archive, const std::string &name, sf::Texture *outt ) { close(); - init_av(); sf::InputStream *s = NULL; @@ -957,7 +1054,7 @@ bool FeMedia::open( const std::string &archive, if ( avformat_find_stream_info( m_imp->m_format_ctx, NULL ) < 0 ) { FeLog() << "Error finding stream information in input file: " - << m_imp->m_format_ctx->filename << std::endl; + << m_imp->m_format_ctx->url << std::endl; return false; } @@ -966,17 +1063,23 @@ bool FeMedia::open( const std::string &archive, int stream_id( -1 ); AVCodec *dec; stream_id = av_find_best_stream( m_imp->m_format_ctx, AVMEDIA_TYPE_AUDIO, - -1, -1, &dec, 0 ); + -1, -1, &dec, 0 ); if ( stream_id >= 0 ) { - AVCodecContext *codec_ctx = m_imp->m_format_ctx->streams[stream_id]->codec; + + AVCodecContext *codec_ctx; + codec_ctx = avcodec_alloc_context3( NULL ); + + avcodec_parameters_to_context( codec_ctx, m_imp->m_format_ctx->streams[stream_id]->codecpar ); + codec_ctx->request_sample_fmt = AV_SAMPLE_FMT_S16; if ( avcodec_open2( codec_ctx, dec, NULL ) < 0 ) { FeLog() << "Could not open audio decoder for file: " - << m_imp->m_format_ctx->filename << std::endl; + << m_imp->m_format_ctx->url << std::endl; + avcodec_free_context( &codec_ctx ); } else { @@ -989,7 +1092,7 @@ bool FeMedia::open( const std::string &archive, // TODO: Fix buffer sizing, we allocate way // more than we use // - m_audio->buffer = (sf::Int16 *)av_malloc( + m_audio->audio_buff = (sf::Int16 *)av_malloc( MAX_AUDIO_FRAME_SIZE + AV_INPUT_BUFFER_PADDING_SIZE + codec_ctx->sample_rate ); @@ -1004,7 +1107,7 @@ bool FeMedia::open( const std::string &archive, if ( codec_ctx->sample_fmt != AV_SAMPLE_FMT_S16 ) { FeLog() << "Warning: Attract-Mode was compiled without an audio resampler (libswresample or libavresample)." << std::endl - << "The audio format in " << m_imp->m_format_ctx->filename << " appears to need resampling. It will likely sound like garbage." << std::endl; + << "The audio format in " << m_imp->m_format_ctx->url << " appears to need resampling. It will likely sound like garbage." << std::endl; } #endif } @@ -1024,11 +1127,16 @@ bool FeMedia::open( const std::string &archive, if ( stream_id < 0 ) { FeLog() << "No video stream found, file: " - << m_imp->m_format_ctx->filename << std::endl; + << m_imp->m_format_ctx->url << std::endl; } else { - AVCodecContext *codec_ctx = m_imp->m_format_ctx->streams[stream_id]->codec; + + AVCodecContext *codec_ctx; + codec_ctx = avcodec_alloc_context3( NULL ); + + avcodec_parameters_to_context( codec_ctx, m_imp->m_format_ctx->streams[stream_id]->codecpar ); + codec_ctx->workaround_bugs = FF_BUG_AUTODETECT; // Note also: http://trac.ffmpeg.org/ticket/4404 @@ -1054,7 +1162,7 @@ bool FeMedia::open( const std::string &archive, case AV_CODEC_ID_MPEG4: FeLog() << "mmal video decoding (" << dec->name << ") not supported for file (trying software): " - << m_imp->m_format_ctx->filename << std::endl; + << m_imp->m_format_ctx->url << std::endl; dec = avcodec_find_decoder_by_name(prev_dec_name.c_str()); @@ -1069,7 +1177,8 @@ bool FeMedia::open( const std::string &archive, if ( av_result < 0 ) { FeLog() << "Could not open video decoder for file: " - << m_imp->m_format_ctx->filename << std::endl; + << m_imp->m_format_ctx->url << std::endl; + avcodec_free_context( &codec_ctx ); } } @@ -1191,161 +1300,42 @@ bool FeMedia::onGetData( Chunk &data ) return false; } -#if (LIBAVCODEC_VERSION_INT < AV_VERSION_INT( 53, 25, 0 )) - { - sf::Lock l( m_audio->buffer_mutex ); - - int bsize = MAX_AUDIO_FRAME_SIZE; - if ( avcodec_decode_audio3( - m_audio->codec_ctx, - (m_audio->buffer + offset), - &bsize, packet) < 0 ) - { - FeLog() << "Error decoding audio." << std::endl; - FeBaseStream::free_packet( packet ); - return false; - } - else - { - offset += bsize / sizeof( sf::Int16 ); - data.sampleCount += bsize / sizeof(sf::Int16); - data.samples = m_audio->buffer; - } - } -#else - #if (LIBAVCODEC_VERSION_INT >= AV_VERSION_INT( 55, 45, 0 )) - AVFrame *frame = av_frame_alloc(); - m_audio->codec_ctx->refcounted_frames = 1; - #else - AVFrame *frame = avcodec_alloc_frame(); - #endif - // - // TODO: avcodec_decode_audio4() can return multiple frames per packet depending on the codec. - // We don't deal with this appropriately... - // - int got_frame( 0 ); - int len = avcodec_decode_audio4( m_audio->codec_ctx, frame, &got_frame, packet ); - if ( len < 0 ) + int r = avcodec_send_packet( m_audio->codec_ctx, packet ); + if (( r < 0 ) && ( r != AVERROR(EAGAIN) )) { char buff[256]; - av_strerror( len, buff, 256 ); - FeDebug() << "Error decoding audio: " << buff << std::endl; + av_strerror( r, buff, 256 ); + FeLog() << "Error decoding audio (sending packet): " << buff << std::endl; } - if ( got_frame ) + FeBaseStream::free_packet( packet ); + + r = AVERROR(EAGAIN); + + // + // Note that avcodec_receive_frame() may need to return multiple frames per packet + // depending on the audio codec. + // + do { - int data_size = av_samples_get_buffer_size( - NULL, - m_audio->codec_ctx->channels, - frame->nb_samples, - m_audio->codec_ctx->sample_fmt, 1); + AVFrame *frame = av_frame_alloc(); + r = avcodec_receive_frame( m_audio->codec_ctx, frame ); -#ifdef DO_RESAMPLE - if ( m_audio->codec_ctx->sample_fmt == AV_SAMPLE_FMT_S16 ) -#endif + if ( r == 0 ) { - sf::Lock l( m_audio->buffer_mutex ); - - memcpy( (m_audio->buffer + offset), frame->data[0], data_size ); - offset += data_size / sizeof( sf::Int16 ); - data.sampleCount += data_size / sizeof(sf::Int16); - data.samples = m_audio->buffer; + if ( !m_audio->process_frame( frame, data, offset ) ) + return false; } -#ifdef DO_RESAMPLE else { - sf::Lock l( m_audio->buffer_mutex ); - - if ( !m_audio->resample_ctx ) - { - m_audio->resample_ctx = resample_alloc(); - if ( !m_audio->resample_ctx ) - { - FeLog() << "Error allocating audio format converter." << std::endl; - FeBaseStream::free_packet( packet ); - FeBaseStream::free_frame( frame ); - return false; - } - - int64_t channel_layout = frame->channel_layout; - if ( !channel_layout ) - { - channel_layout = av_get_default_channel_layout( - m_audio->codec_ctx->channels ); - } - - av_opt_set_int( m_audio->resample_ctx, "in_channel_layout", channel_layout, 0 ); - av_opt_set_int( m_audio->resample_ctx, "in_sample_fmt", frame->format, 0 ); - av_opt_set_int( m_audio->resample_ctx, "in_sample_rate", frame->sample_rate, 0 ); - av_opt_set_int( m_audio->resample_ctx, "out_channel_layout", channel_layout, 0 ); - av_opt_set_int( m_audio->resample_ctx, "out_sample_fmt", AV_SAMPLE_FMT_S16, 0 ); - av_opt_set_int( m_audio->resample_ctx, "out_sample_rate", frame->sample_rate, 0 ); - - FeDebug() << "Initializing resampler: in_sample_fmt=" - << av_get_sample_fmt_name( (AVSampleFormat)frame->format ) - << ", in_sample_rate=" << frame->sample_rate - << ", out_sample_fmt=" << av_get_sample_fmt_name( AV_SAMPLE_FMT_S16 ) - << ", out_sample_rate=" << frame->sample_rate << std::endl; - - if ( resample_init( m_audio->resample_ctx ) < 0 ) - { - FeLog() << "Error initializing audio format converter, input format=" - << av_get_sample_fmt_name( (AVSampleFormat)frame->format ) - << ", input sample rate=" << frame->sample_rate << std::endl; - FeBaseStream::free_packet( packet ); - FeBaseStream::free_frame( frame ); - resample_free( &m_audio->resample_ctx ); - m_audio->resample_ctx = NULL; - return false; - } - } - if ( m_audio->resample_ctx ) + if ( r != AVERROR(EAGAIN) ) { - int out_linesize; - av_samples_get_buffer_size( - &out_linesize, - m_audio->codec_ctx->channels, - frame->nb_samples, - AV_SAMPLE_FMT_S16, 0 ); - - uint8_t *tmp_ptr = (uint8_t *)(m_audio->buffer + offset); - -#ifdef USE_SWRESAMPLE - int out_samples = swr_convert( - m_audio->resample_ctx, - &tmp_ptr, - frame->nb_samples, - (const uint8_t **)frame->data, - frame->nb_samples ); -#else // USE_AVRESAMPLE - int out_samples = avresample_convert( - m_audio->resample_ctx, - &tmp_ptr, - out_linesize, - frame->nb_samples, - frame->data, - frame->linesize[0], - frame->nb_samples ); -#endif - if ( out_samples < 0 ) - { - FeLog() << "Error performing audio conversion." << std::endl; - FeBaseStream::free_packet( packet ); - FeBaseStream::free_frame( frame ); - break; - } - offset += out_samples * m_audio->codec_ctx->channels; - data.sampleCount += out_samples * m_audio->codec_ctx->channels; - data.samples = m_audio->buffer; + char buff[256]; + av_strerror( r, buff, 256 ); + FeLog() << "Error decoding audio (receiving frame): " << buff << std::endl; } } -#endif - } - FeBaseStream::free_frame( frame ); - -#endif - - FeBaseStream::free_packet( packet ); + } while ( r != AVERROR(EAGAIN) ); } return true; @@ -1358,8 +1348,6 @@ void FeMedia::onSeek( sf::Time timeOffset ) bool FeMedia::is_supported_media_file( const std::string &filename ) { - init_av(); - // Work around for FFmpeg not recognizing certain file extensions // that it supports (xmv reported as of Dec 2015) // diff --git a/src/media.hpp b/src/media.hpp index c5bf5e6f..b031e12a 100644 --- a/src/media.hpp +++ b/src/media.hpp @@ -120,7 +120,6 @@ friend class FeVideoImp; FeMedia( const FeMedia & ); FeMedia &operator=( const FeMedia & ); - static void init_av(); float m_aspect_ratio; };