X-Git-Url: https://main.carlh.net/gitweb/?p=libdcp.git;a=blobdiff_plain;f=src%2Fsound_asset.cc;h=1c0a181f81255c72dea107d73b6100c556123e78;hp=70ed0f29cfb279e0a6daff9dc0326affdeac8728;hb=ceaf7bc52712cb60708ed5eb5c62c5e463dd8e89;hpb=ef26e37fce728cba6fb50e9e6743a012b1bbbf64 diff --git a/src/sound_asset.cc b/src/sound_asset.cc index 70ed0f29..1c0a181f 100644 --- a/src/sound_asset.cc +++ b/src/sound_asset.cc @@ -1,260 +1,217 @@ /* - Copyright (C) 2012 Carl Hetherington + Copyright (C) 2012-2021 Carl Hetherington - This program is free software; you can redistribute it and/or modify + This file is part of libdcp. + + libdcp 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 2 of the License, or (at your option) any later version. - This program is distributed in the hope that it will be useful, + libdcp 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 this program; if not, write to the Free Software - Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. - + along with libdcp. If not, see . + + In addition, as a special exception, the copyright holders give + permission to link the code of portions of this program with the + OpenSSL library under certain conditions as described in each + individual source file, and distribute linked combinations + including the two. + + You must obey the GNU General Public License in all respects + for all of the code used other than OpenSSL. If you modify + file(s) with this exception, you may extend this exception to your + version of the file(s), but you are not obligated to do so. If you + do not wish to do so, delete this exception statement from your + version. If you delete this exception statement from all source + files in the program, then also delete it here. */ + /** @file src/sound_asset.cc - * @brief An asset made up of WAV files + * @brief SoundAsset class */ -#include -#include -#include -#include -#include -#include "KM_fileio.h" -#include "AS_DCP.h" + #include "sound_asset.h" #include "util.h" #include "exceptions.h" #include "sound_frame.h" +#include "sound_asset_writer.h" +#include "sound_asset_reader.h" +#include "compose.hpp" +#include "dcp_assert.h" +#include +#include +#include +#include +#include +#include + using std::string; -using std::stringstream; -using std::ostream; using std::vector; using std::list; -using boost::shared_ptr; -using boost::lexical_cast; -using namespace libdcp; - -SoundAsset::SoundAsset (boost::filesystem::path directory, boost::filesystem::path mxf_name) - : MXFAsset (directory, mxf_name) - , _channels (0) - , _sampling_rate (0) -{ - -} - -void -SoundAsset::create (vector const & files) -{ - create (boost::bind (&SoundAsset::path_from_channel, this, _1, files)); -} - -void -SoundAsset::read () +using std::shared_ptr; +using std::dynamic_pointer_cast; +using namespace dcp; + + +SoundAsset::SoundAsset (boost::filesystem::path file) + : Asset (file) + /* XXX: this is a fallback language, which will be used if we can't find the RFC5646SpokenLanguage + * in the MXF header. Perhaps RFC5646SpokenLanguage is optional and we should just not write it + * if we don't know it. + */ + , _language ("en-US") { ASDCP::PCM::MXFReader reader; - if (ASDCP_FAILURE (reader.OpenRead (path().string().c_str()))) { - boost::throw_exception (MXFFileError ("could not open MXF file for reading", path().string())); + auto r = reader.OpenRead (file.string().c_str()); + if (ASDCP_FAILURE(r)) { + boost::throw_exception (MXFFileError("could not open MXF file for reading", file.string(), r)); } ASDCP::PCM::AudioDescriptor desc; - if (ASDCP_FAILURE (reader.FillAudioDescriptor (desc))) { - boost::throw_exception (DCPReadError ("could not read audio MXF information")); + if (ASDCP_FAILURE (reader.FillAudioDescriptor(desc))) { + boost::throw_exception (ReadError("could not read audio MXF information")); } _sampling_rate = desc.AudioSamplingRate.Numerator / desc.AudioSamplingRate.Denominator; _channels = desc.ChannelCount; - _edit_rate = desc.EditRate.Numerator; - assert (desc.EditRate.Denominator == 1); - _intrinsic_duration = desc.ContainerDuration; -} - -boost::filesystem::path -SoundAsset::path_from_channel (Channel channel, vector const & files) -{ - unsigned int const c = int (channel); - assert (c < files.size ()); - return files[c]; -} + _edit_rate = Fraction (desc.EditRate.Numerator, desc.EditRate.Denominator); -void -SoundAsset::create (boost::function get_path) -{ - ASDCP::Rational asdcp_edit_rate (_edit_rate, 1); - - assert (_channels > 0); - ASDCP::PCM::WAVParser pcm_parser_channel[_channels]; - if (pcm_parser_channel[0].OpenRead (get_path(LEFT).string().c_str(), asdcp_edit_rate)) { - boost::throw_exception (FileError ("could not open WAV file for reading", get_path(LEFT))); - } - - ASDCP::PCM::AudioDescriptor audio_desc; - pcm_parser_channel[0].FillAudioDescriptor (audio_desc); - audio_desc.ChannelCount = 0; - audio_desc.BlockAlign = 0; - audio_desc.EditRate = asdcp_edit_rate; - audio_desc.AvgBps = audio_desc.AvgBps * _channels; - - Channel channels[] = { - LEFT, - RIGHT, - CENTRE, - LFE, - LS, - RS, - /* XXX: not quite sure what these should be yet */ - CHANNEL_7, - CHANNEL_8 - }; - - assert (int(_channels) <= int(sizeof(channels) / sizeof(Channel))); - - ASDCP::PCM::FrameBuffer frame_buffer_channel[_channels]; - ASDCP::PCM::AudioDescriptor audio_desc_channel[_channels]; - - for (int i = 0; i < _channels; ++i) { - - boost::filesystem::path const path = get_path (channels[i]); - - if (ASDCP_FAILURE (pcm_parser_channel[i].OpenRead (path.string().c_str(), asdcp_edit_rate))) { - boost::throw_exception (FileError ("could not open WAV file for reading", path)); - } - - pcm_parser_channel[i].FillAudioDescriptor (audio_desc_channel[i]); - frame_buffer_channel[i].Capacity (ASDCP::PCM::CalcFrameBufferSize (audio_desc_channel[i])); - - audio_desc.ChannelCount += audio_desc_channel[i].ChannelCount; - audio_desc.BlockAlign += audio_desc_channel[i].BlockAlign; - } - - ASDCP::PCM::FrameBuffer frame_buffer; - frame_buffer.Capacity (ASDCP::PCM::CalcFrameBufferSize (audio_desc)); - frame_buffer.Size (ASDCP::PCM::CalcFrameBufferSize (audio_desc)); - - ASDCP::WriterInfo writer_info; - MXFAsset::fill_writer_info (&writer_info, _uuid, _interop, _metadata); + _intrinsic_duration = desc.ContainerDuration; - ASDCP::PCM::MXFWriter mxf_writer; - if (ASDCP_FAILURE (mxf_writer.OpenWrite (path().string().c_str(), writer_info, audio_desc))) { - boost::throw_exception (FileError ("could not open audio MXF for writing", path().string())); + ASDCP::WriterInfo info; + if (ASDCP_FAILURE (reader.FillWriterInfo(info))) { + boost::throw_exception (ReadError("could not read audio MXF information")); } - - for (int i = 0; i < _intrinsic_duration; ++i) { - for (int j = 0; j < _channels; ++j) { - memset (frame_buffer_channel[j].Data(), 0, frame_buffer_channel[j].Capacity()); - if (ASDCP_FAILURE (pcm_parser_channel[j].ReadFrame (frame_buffer_channel[j]))) { - boost::throw_exception (MiscError ("could not read audio frame")); - } - } - - byte_t *data_s = frame_buffer.Data(); - byte_t *data_e = data_s + frame_buffer.Capacity(); - byte_t sample_size = ASDCP::PCM::CalcSampleSize (audio_desc_channel[0]); - int offset = 0; - - while (data_s < data_e) { - for (int j = 0; j < _channels; ++j) { - byte_t* frame = frame_buffer_channel[j].Data() + offset; - memcpy (data_s, frame, sample_size); - data_s += sample_size; - } - offset += sample_size; - } - - if (ASDCP_FAILURE (mxf_writer.WriteFrame (frame_buffer, _encryption_context, 0))) { - boost::throw_exception (MiscError ("could not write audio MXF frame")); - } - - if (_progress) { - (*_progress) (0.5 * float (i) / _intrinsic_duration); + ASDCP::MXF::SoundfieldGroupLabelSubDescriptor* soundfield; + auto rr = reader.OP1aHeader().GetMDObjectByType( + asdcp_smpte_dict->ul(ASDCP::MDD_SoundfieldGroupLabelSubDescriptor), + reinterpret_cast(&soundfield) + ); + + if (KM_SUCCESS(rr)) { + if (!soundfield->RFC5646SpokenLanguage.empty()) { + char buffer[64]; + soundfield->RFC5646SpokenLanguage.get().EncodeString(buffer, sizeof(buffer)); + _language = buffer; } } - if (ASDCP_FAILURE (mxf_writer.Finalize())) { - boost::throw_exception (MiscError ("could not finalise audio MXF")); - } + _id = read_writer_info (info); } -string -SoundAsset::cpl_node_name () const + +SoundAsset::SoundAsset (Fraction edit_rate, int sampling_rate, int channels, LanguageTag language, Standard standard) + : MXF (standard) + , _edit_rate (edit_rate) + , _channels (channels) + , _sampling_rate (sampling_rate) + , _language (language.to_string()) { - return "MainSound"; + } + bool -SoundAsset::equals (shared_ptr other, EqualityOptions opt, boost::function note) const +SoundAsset::equals (shared_ptr other, EqualityOptions opt, NoteHandler note) const { - if (!MXFAsset::equals (other, opt, note)) { - return false; - } - ASDCP::PCM::MXFReader reader_A; - if (ASDCP_FAILURE (reader_A.OpenRead (path().string().c_str()))) { - boost::throw_exception (MXFFileError ("could not open MXF file for reading", path().string())); + DCP_ASSERT (file()); + auto r = reader_A.OpenRead (file()->string().c_str()); + if (ASDCP_FAILURE(r)) { + boost::throw_exception (MXFFileError("could not open MXF file for reading", file()->string(), r)); } ASDCP::PCM::MXFReader reader_B; - if (ASDCP_FAILURE (reader_B.OpenRead (other->path().string().c_str()))) { - boost::throw_exception (MXFFileError ("could not open MXF file for reading", path().string())); + r = reader_B.OpenRead (other->file()->string().c_str()); + if (ASDCP_FAILURE (r)) { + boost::throw_exception (MXFFileError("could not open MXF file for reading", other->file()->string(), r)); } ASDCP::PCM::AudioDescriptor desc_A; - if (ASDCP_FAILURE (reader_A.FillAudioDescriptor (desc_A))) { - boost::throw_exception (DCPReadError ("could not read audio MXF information")); + if (ASDCP_FAILURE (reader_A.FillAudioDescriptor(desc_A))) { + boost::throw_exception (ReadError ("could not read audio MXF information")); } ASDCP::PCM::AudioDescriptor desc_B; - if (ASDCP_FAILURE (reader_B.FillAudioDescriptor (desc_B))) { - boost::throw_exception (DCPReadError ("could not read audio MXF information")); + if (ASDCP_FAILURE (reader_B.FillAudioDescriptor(desc_B))) { + boost::throw_exception (ReadError ("could not read audio MXF information")); } - - if ( - desc_A.EditRate != desc_B.EditRate || - desc_A.AudioSamplingRate != desc_B.AudioSamplingRate || - desc_A.Locked != desc_B.Locked || - desc_A.ChannelCount != desc_B.ChannelCount || - desc_A.QuantizationBits != desc_B.QuantizationBits || - desc_A.BlockAlign != desc_B.BlockAlign || - desc_A.AvgBps != desc_B.AvgBps || - desc_A.LinkedTrackID != desc_B.LinkedTrackID || - desc_A.ContainerDuration != desc_B.ContainerDuration -// desc_A.ChannelFormat != desc_B.ChannelFormat || - ) { - - note (ERROR, "audio MXF picture descriptors differ"); + + if (desc_A.EditRate != desc_B.EditRate) { + note ( + NoteType::ERROR, + String::compose ( + "audio edit rates differ: %1/%2 cf %3/%4", + desc_A.EditRate.Numerator, desc_A.EditRate.Denominator, desc_B.EditRate.Numerator, desc_B.EditRate.Denominator + ) + ); + return false; + } else if (desc_A.AudioSamplingRate != desc_B.AudioSamplingRate) { + note ( + NoteType::ERROR, + String::compose ( + "audio sampling rates differ: %1 cf %2", + desc_A.AudioSamplingRate.Numerator, desc_A.AudioSamplingRate.Denominator, + desc_B.AudioSamplingRate.Numerator, desc_B.AudioSamplingRate.Numerator + ) + ); + return false; + } else if (desc_A.Locked != desc_B.Locked) { + note (NoteType::ERROR, String::compose ("audio locked flags differ: %1 cf %2", desc_A.Locked, desc_B.Locked)); + return false; + } else if (desc_A.ChannelCount != desc_B.ChannelCount) { + note (NoteType::ERROR, String::compose ("audio channel counts differ: %1 cf %2", desc_A.ChannelCount, desc_B.ChannelCount)); + return false; + } else if (desc_A.QuantizationBits != desc_B.QuantizationBits) { + note (NoteType::ERROR, String::compose ("audio bits per sample differ: %1 cf %2", desc_A.QuantizationBits, desc_B.QuantizationBits)); + return false; + } else if (desc_A.BlockAlign != desc_B.BlockAlign) { + note (NoteType::ERROR, String::compose ("audio bytes per sample differ: %1 cf %2", desc_A.BlockAlign, desc_B.BlockAlign)); + return false; + } else if (desc_A.AvgBps != desc_B.AvgBps) { + note (NoteType::ERROR, String::compose ("audio average bps differ: %1 cf %2", desc_A.AvgBps, desc_B.AvgBps)); + return false; + } else if (desc_A.LinkedTrackID != desc_B.LinkedTrackID) { + note (NoteType::ERROR, String::compose ("audio linked track IDs differ: %1 cf %2", desc_A.LinkedTrackID, desc_B.LinkedTrackID)); + return false; + } else if (desc_A.ContainerDuration != desc_B.ContainerDuration) { + note (NoteType::ERROR, String::compose ("audio container durations differ: %1 cf %2", desc_A.ContainerDuration, desc_B.ContainerDuration)); return false; + } else if (desc_A.ChannelFormat != desc_B.ChannelFormat) { + /* XXX */ } - - ASDCP::PCM::FrameBuffer buffer_A (1 * Kumu::Megabyte); - ASDCP::PCM::FrameBuffer buffer_B (1 * Kumu::Megabyte); - + + auto other_sound = dynamic_pointer_cast (other); + + auto reader = start_read (); + auto other_reader = other_sound->start_read (); + for (int i = 0; i < _intrinsic_duration; ++i) { - if (ASDCP_FAILURE (reader_A.ReadFrame (i, buffer_A))) { - boost::throw_exception (DCPReadError ("could not read audio frame")); - } - - if (ASDCP_FAILURE (reader_B.ReadFrame (i, buffer_B))) { - boost::throw_exception (DCPReadError ("could not read audio frame")); - } - - if (buffer_A.Size() != buffer_B.Size()) { - note (ERROR, "sizes of audio data for frame " + lexical_cast(i) + " differ"); + + auto frame_A = reader->get_frame (i); + auto frame_B = other_reader->get_frame (i); + + if (frame_A->size() != frame_B->size()) { + note (NoteType::ERROR, String::compose ("sizes of audio data for frame %1 differ", i)); return false; } - - if (memcmp (buffer_A.RoData(), buffer_B.RoData(), buffer_A.Size()) != 0) { - for (uint32_t i = 0; i < buffer_A.Size(); ++i) { - int const d = abs (buffer_A.RoData()[i] - buffer_B.RoData()[i]); - if (d > opt.max_audio_sample_error) { - note (ERROR, "PCM data difference of " + lexical_cast (d)); - return false; + + if (memcmp (frame_A->data(), frame_B->data(), frame_A->size()) != 0) { + for (int sample = 0; sample < frame_A->samples(); ++sample) { + for (int channel = 0; channel < frame_A->channels(); ++channel) { + int32_t const d = abs(frame_A->get(channel, sample) - frame_B->get(channel, sample)); + if (d > opt.max_audio_sample_error) { + note (NoteType::ERROR, String::compose ("PCM data difference of %1", d)); + return false; + } } } } @@ -263,115 +220,43 @@ SoundAsset::equals (shared_ptr other, EqualityOptions opt, boost::f return true; } -shared_ptr -SoundAsset::get_frame (int n) const -{ - /* XXX: should add on entry point here? */ - return shared_ptr (new SoundFrame (path().string(), n, _decryption_context)); -} shared_ptr -SoundAsset::start_write () -{ - /* XXX: can't we use a shared_ptr here? */ - return shared_ptr (new SoundAssetWriter (this)); -} - -struct SoundAssetWriter::ASDCPState -{ - ASDCP::PCM::MXFWriter mxf_writer; - ASDCP::PCM::FrameBuffer frame_buffer; - ASDCP::WriterInfo writer_info; - ASDCP::PCM::AudioDescriptor audio_desc; - ASDCP::AESEncContext* encryption_context; -}; - -SoundAssetWriter::SoundAssetWriter (SoundAsset* a) - : _state (new SoundAssetWriter::ASDCPState) - , _asset (a) - , _finalized (false) - , _frames_written (0) - , _frame_buffer_offset (0) +SoundAsset::start_write (boost::filesystem::path file, vector active_channels, bool atmos_sync) { - _state->encryption_context = a->encryption_context (); - - /* Derived from ASDCP::Wav::SimpleWaveHeader::FillADesc */ - _state->audio_desc.EditRate = ASDCP::Rational (_asset->edit_rate(), 1); - _state->audio_desc.AudioSamplingRate = ASDCP::Rational (_asset->sampling_rate(), 1); - _state->audio_desc.Locked = 0; - _state->audio_desc.ChannelCount = _asset->channels (); - _state->audio_desc.QuantizationBits = 24; - _state->audio_desc.BlockAlign = 3 * _asset->channels(); - _state->audio_desc.AvgBps = _asset->sampling_rate() * _state->audio_desc.BlockAlign; - _state->audio_desc.LinkedTrackID = 0; - _state->audio_desc.ChannelFormat = ASDCP::PCM::CF_NONE; - - _state->frame_buffer.Capacity (ASDCP::PCM::CalcFrameBufferSize (_state->audio_desc)); - _state->frame_buffer.Size (ASDCP::PCM::CalcFrameBufferSize (_state->audio_desc)); - memset (_state->frame_buffer.Data(), 0, _state->frame_buffer.Capacity()); - - _asset->fill_writer_info (&_state->writer_info, _asset->uuid (), _asset->interop(), _asset->metadata()); - - if (ASDCP_FAILURE (_state->mxf_writer.OpenWrite (_asset->path().string().c_str(), _state->writer_info, _state->audio_desc))) { - boost::throw_exception (FileError ("could not open audio MXF for writing", _asset->path().string())); + if (atmos_sync && _channels < 14) { + throw MiscError ("Insufficient channels to write ATMOS sync (there must be at least 14)"); } -} - -void -SoundAssetWriter::write (float const * const * data, int frames) -{ - for (int i = 0; i < frames; ++i) { - byte_t* out = _state->frame_buffer.Data() + _frame_buffer_offset; - - /* Write one sample per channel */ - for (int j = 0; j < _asset->channels(); ++j) { - int32_t const s = data[j][i] * (1 << 23); - *out++ = (s & 0xff); - *out++ = (s & 0xff00) >> 8; - *out++ = (s & 0xff0000) >> 16; - } - _frame_buffer_offset += 3 * _asset->channels(); - - assert (_frame_buffer_offset <= int (_state->frame_buffer.Capacity())); - - /* Finish the MXF frame if required */ - if (_frame_buffer_offset == int (_state->frame_buffer.Capacity())) { - write_current_frame (); - _frame_buffer_offset = 0; - memset (_state->frame_buffer.Data(), 0, _state->frame_buffer.Capacity()); - } - } + return shared_ptr (new SoundAssetWriter(this, file, active_channels, atmos_sync)); } -void -SoundAssetWriter::write_current_frame () -{ - if (ASDCP_FAILURE (_state->mxf_writer.WriteFrame (_state->frame_buffer, _state->encryption_context, 0))) { - boost::throw_exception (MiscError ("could not write audio MXF frame")); - } - ++_frames_written; +shared_ptr +SoundAsset::start_read () const +{ + return shared_ptr (new SoundAssetReader(this, key(), standard())); } -void -SoundAssetWriter::finalize () + +string +SoundAsset::static_pkl_type (Standard standard) { - if (_frame_buffer_offset > 0) { - write_current_frame (); - } - - if (ASDCP_FAILURE (_state->mxf_writer.Finalize())) { - boost::throw_exception (MiscError ("could not finalise audio MXF")); + switch (standard) { + case Standard::INTEROP: + return "application/x-smpte-mxf;asdcpKind=Sound"; + case Standard::SMPTE: + return "application/mxf"; + default: + DCP_ASSERT (false); } - - _finalized = true; - _asset->set_intrinsic_duration (_frames_written); - _asset->set_duration (_frames_written); } -string -SoundAsset::key_type () const + +bool +SoundAsset::valid_mxf (boost::filesystem::path file) { - return "MDAK"; + ASDCP::PCM::MXFReader reader; + Kumu::Result_t r = reader.OpenRead (file.string().c_str()); + return !ASDCP_FAILURE (r); }