X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=modules%2Fffmpeg%2Fproducer%2Faudio%2Faudio_decoder.cpp;h=ac7b44e423e91f2913bbd6c20bf4389349d1aab5;hb=c0dc760a3d87b346c9f267cd9d74c67c55d3bdc3;hp=1ac7fdeaa796fe61c3db3c1602f62b11e747ca5a;hpb=2e34146cbc85f7069f3a6d81d95a2270de979681;p=casparcg diff --git a/modules/ffmpeg/producer/audio/audio_decoder.cpp b/modules/ffmpeg/producer/audio/audio_decoder.cpp index 1ac7fdeaa..ac7b44e42 100644 --- a/modules/ffmpeg/producer/audio/audio_decoder.cpp +++ b/modules/ffmpeg/producer/audio/audio_decoder.cpp @@ -1,153 +1,181 @@ -/* -* Copyright (c) 2011 Sveriges Television AB -* -* This file is part of CasparCG (www.casparcg.com). -* -* CasparCG is free software: you can redistribute it and/or modify -* it under the terms of the GNU General Public License as published by -* the Free Software Foundation, either version 3 of the License, or -* (at your option) any later version. -* -* CasparCG is distributed in the hope that it will be useful, -* but WITHOUT ANY WARRANTY; without even the implied warranty of -* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -* GNU General Public License for more details. -* -* You should have received a copy of the GNU General Public License -* along with CasparCG. If not, see . -* -* Author: Robert Nagy, ronag89@gmail.com -*/ - -#include "../../stdafx.h" - -#include "audio_decoder.h" - -#include "audio_resampler.h" - -#include "../util/util.h" -#include "../../ffmpeg_error.h" - -#include - -#include - -#include - -#if defined(_MSC_VER) -#pragma warning (push) -#pragma warning (disable : 4244) -#endif -extern "C" -{ - #include - #include -} -#if defined(_MSC_VER) -#pragma warning (pop) -#endif - -namespace caspar { namespace ffmpeg { - -struct audio_decoder::implementation : boost::noncopyable -{ - int index_; - const safe_ptr codec_context_; - const core::video_format_desc format_desc_; - - audio_resampler resampler_; - - std::vector> buffer1_; - - std::queue> packets_; - - const int64_t nb_frames_; - tbb::atomic file_frame_number_; -public: - explicit implementation(const safe_ptr& context, const core::video_format_desc& format_desc) - : format_desc_(format_desc) - , codec_context_(open_codec(*context, AVMEDIA_TYPE_AUDIO, index_)) - , resampler_(format_desc.audio_channels, codec_context_->channels, - format_desc.audio_sample_rate, codec_context_->sample_rate, - AV_SAMPLE_FMT_S32, codec_context_->sample_fmt) - , buffer1_(AVCODEC_MAX_AUDIO_FRAME_SIZE*2) - , nb_frames_(0)//context->streams[index_]->nb_frames) - { - file_frame_number_ = 0; - CASPAR_LOG(debug) << "[audio_decoder] " << context->streams[index_]->codec->codec->long_name; - } - - void push(const std::shared_ptr& packet) - { - if(!packet) - return; - - if(packet->stream_index == index_ || packet->data == nullptr) - packets_.push(make_safe_ptr(packet)); - } - - std::shared_ptr poll() - { - if(packets_.empty()) - return nullptr; - - auto packet = packets_.front(); - - if(packet->data == nullptr) - { - packets_.pop(); - file_frame_number_ = static_cast(packet->pos); - avcodec_flush_buffers(codec_context_.get()); - return flush_audio(); - } - - auto audio = decode(*packet); - - if(packet->size == 0) - packets_.pop(); - - return audio; - } - - std::shared_ptr decode(AVPacket& pkt) - { - buffer1_.resize(AVCODEC_MAX_AUDIO_FRAME_SIZE*2); - int written_bytes = buffer1_.size() - FF_INPUT_BUFFER_PADDING_SIZE; - - int ret = THROW_ON_ERROR2(avcodec_decode_audio3(codec_context_.get(), reinterpret_cast(buffer1_.data()), &written_bytes, &pkt), "[audio_decoder]"); - - // There might be several frames in one packet. - pkt.size -= ret; - pkt.data += ret; - - buffer1_.resize(written_bytes); - - buffer1_ = resampler_.resample(std::move(buffer1_)); - - const auto n_samples = buffer1_.size() / av_get_bytes_per_sample(AV_SAMPLE_FMT_S32); - const auto samples = reinterpret_cast(buffer1_.data()); - - ++file_frame_number_; - - return std::make_shared(samples, samples + n_samples); - } - - bool ready() const - { - return packets_.size() > 10; - } - - uint32_t nb_frames() const - { - return 0;//std::max(nb_frames_, file_frame_number_); - } -}; - -audio_decoder::audio_decoder(const safe_ptr& context, const core::video_format_desc& format_desc) : impl_(new implementation(context, format_desc)){} -void audio_decoder::push(const std::shared_ptr& packet){impl_->push(packet);} -bool audio_decoder::ready() const{return impl_->ready();} -std::shared_ptr audio_decoder::poll(){return impl_->poll();} -uint32_t audio_decoder::nb_frames() const{return impl_->nb_frames();} -uint32_t audio_decoder::file_frame_number() const{return impl_->file_frame_number_;} - -}} \ No newline at end of file +/* +* Copyright 2013 Sveriges Television AB http://casparcg.com/ +* +* This file is part of CasparCG (www.casparcg.com). +* +* CasparCG is free software: you can redistribute it and/or modify +* it under the terms of the GNU General Public License as published by +* the Free Software Foundation, either version 3 of the License, or +* (at your option) any later version. +* +* CasparCG is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +* GNU General Public License for more details. +* +* You should have received a copy of the GNU General Public License +* along with CasparCG. If not, see . +* +* Author: Robert Nagy, ronag89@gmail.com +*/ + +#include "../../StdAfx.h" + +#include "audio_decoder.h" + +#include "../util/util.h" +#include "../../ffmpeg_error.h" + +#include +#include + +#include + +#include + +#if defined(_MSC_VER) +#pragma warning (push) +#pragma warning (disable : 4244) +#endif +extern "C" +{ + #include + #include + #include +} +#if defined(_MSC_VER) +#pragma warning (pop) +#endif + +namespace caspar { namespace ffmpeg { + +struct audio_decoder::implementation : boost::noncopyable +{ + int index_ = -1; + const spl::shared_ptr codec_context_; + const int out_samplerate_; + + cache_aligned_vector buffer_; + + std::queue> packets_; + + std::shared_ptr swr_ { + swr_alloc_set_opts( + nullptr, + codec_context_->channel_layout + ? codec_context_->channel_layout + : av_get_default_channel_layout(codec_context_->channels), + AV_SAMPLE_FMT_S32, + out_samplerate_, + codec_context_->channel_layout + ? codec_context_->channel_layout + : av_get_default_channel_layout(codec_context_->channels), + codec_context_->sample_fmt, + codec_context_->sample_rate, + 0, + nullptr), + [](SwrContext* p) + { + swr_free(&p); + } + }; + +public: + explicit implementation(const spl::shared_ptr& context, int out_samplerate) + : codec_context_(open_codec(*context, AVMEDIA_TYPE_AUDIO, index_, false)) + , out_samplerate_(out_samplerate) + , buffer_(10 * out_samplerate_ * codec_context_->channels) // 10 seconds of audio + { + if(!swr_) + CASPAR_THROW_EXCEPTION(bad_alloc()); + + THROW_ON_ERROR2(swr_init(swr_.get()), "[audio_decoder]"); + + codec_context_->refcounted_frames = 1; + } + + void push(const std::shared_ptr& packet) + { + if(!packet) + return; + + if(packet->stream_index == index_ || packet->data == nullptr) + packets_.push(spl::make_shared_ptr(packet)); + } + + std::shared_ptr poll() + { + if(packets_.empty()) + return nullptr; + + auto packet = packets_.front(); + + if(packet->data == nullptr) + { + packets_.pop(); + avcodec_flush_buffers(codec_context_.get()); + return flush_audio(); + } + + auto audio = decode(*packet); + + if(packet->size == 0) + packets_.pop(); + + return audio; + } + + std::shared_ptr decode(AVPacket& pkt) + { + auto decoded_frame = create_frame(); + + int got_frame = 0; + auto len = THROW_ON_ERROR2(avcodec_decode_audio4(codec_context_.get(), decoded_frame.get(), &got_frame, &pkt), "[audio_decoder]"); + + if (len == 0) + { + pkt.size = 0; + return nullptr; + } + + pkt.data += len; + pkt.size -= len; + + if (!got_frame) + return nullptr; + + const uint8_t **in = const_cast(decoded_frame->extended_data); + uint8_t* out[] = { reinterpret_cast(buffer_.data()) }; + + const auto channel_samples = swr_convert( + swr_.get(), + out, + static_cast(buffer_.size()) / codec_context_->channels, + in, + decoded_frame->nb_samples); + + return std::make_shared( + buffer_.begin(), + buffer_.begin() + channel_samples * decoded_frame->channels); + } + + bool ready() const + { + return packets_.size() > 10; + } + + std::wstring print() const + { + return L"[audio-decoder] " + u16(codec_context_->codec->long_name); + } +}; + +audio_decoder::audio_decoder(const spl::shared_ptr& context, int out_samplerate) : impl_(new implementation(context, out_samplerate)){} +void audio_decoder::push(const std::shared_ptr& packet){impl_->push(packet);} +bool audio_decoder::ready() const{return impl_->ready();} +std::shared_ptr audio_decoder::poll() { return impl_->poll(); } +int audio_decoder::num_channels() const { return impl_->codec_context_->channels; } +uint64_t audio_decoder::ffmpeg_channel_layout() const { return impl_->codec_context_->channel_layout; } +std::wstring audio_decoder::print() const{return impl_->print();} + +}}