// Copyright 2013 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "media/base/audio_buffer.h"
#include "base/logging.h"
#include "media/base/audio_bus.h"
#include "media/base/buffers.h"
#include "media/base/limits.h"
namespace media {
AudioBuffer::AudioBuffer(SampleFormat sample_format,
int channel_count,
int frame_count,
bool create_buffer,
const uint8* const* data,
const base::TimeDelta timestamp,
const base::TimeDelta duration)
: sample_format_(sample_format),
channel_count_(channel_count),
adjusted_frame_count_(frame_count),
trim_start_(0),
end_of_stream_(!create_buffer && data == NULL && frame_count == 0),
timestamp_(timestamp),
duration_(duration) {
CHECK_GE(channel_count, 0);
CHECK_LE(channel_count, limits::kMaxChannels);
CHECK_GE(frame_count, 0);
int bytes_per_channel = SampleFormatToBytesPerChannel(sample_format);
DCHECK_LE(bytes_per_channel, kChannelAlignment);
int data_size = frame_count * bytes_per_channel;
// Empty buffer?
if (!create_buffer)
return;
if (sample_format == kSampleFormatPlanarF32 ||
sample_format == kSampleFormatPlanarS16) {
// Planar data, so need to allocate buffer for each channel.
// Determine per channel data size, taking into account alignment.
int block_size_per_channel =
(data_size + kChannelAlignment - 1) & ~(kChannelAlignment - 1);
DCHECK_GE(block_size_per_channel, data_size);
// Allocate a contiguous buffer for all the channel data.
data_.reset(static_cast<uint8*>(base::AlignedAlloc(
channel_count * block_size_per_channel, kChannelAlignment)));
channel_data_.reserve(channel_count);
// Copy each channel's data into the appropriate spot.
for (int i = 0; i < channel_count; ++i) {
channel_data_.push_back(data_.get() + i * block_size_per_channel);
if (data)
memcpy(channel_data_[i], data[i], data_size);
}
return;
}
// Remaining formats are interleaved data.
DCHECK(sample_format_ == kSampleFormatU8 ||
sample_format_ == kSampleFormatS16 ||
sample_format_ == kSampleFormatS32 ||
sample_format_ == kSampleFormatF32) << sample_format_;
// Allocate our own buffer and copy the supplied data into it. Buffer must
// contain the data for all channels.
data_size *= channel_count;
data_.reset(
static_cast<uint8*>(base::AlignedAlloc(data_size, kChannelAlignment)));
channel_data_.reserve(1);
channel_data_.push_back(data_.get());
if (data)
memcpy(data_.get(), data[0], data_size);
}
AudioBuffer::~AudioBuffer() {}
// static
scoped_refptr<AudioBuffer> AudioBuffer::CopyFrom(
SampleFormat sample_format,
int channel_count,
int frame_count,
const uint8* const* data,
const base::TimeDelta timestamp,
const base::TimeDelta duration) {
// If you hit this CHECK you likely have a bug in a demuxer. Go fix it.
CHECK_GT(frame_count, 0); // Otherwise looks like an EOF buffer.
CHECK(data[0]);
return make_scoped_refptr(new AudioBuffer(sample_format,
channel_count,
frame_count,
true,
data,
timestamp,
duration));
}
// static
scoped_refptr<AudioBuffer> AudioBuffer::CreateBuffer(SampleFormat sample_format,
int channel_count,
int frame_count) {
CHECK_GT(frame_count, 0); // Otherwise looks like an EOF buffer.
return make_scoped_refptr(new AudioBuffer(sample_format,
channel_count,
frame_count,
true,
NULL,
kNoTimestamp(),
kNoTimestamp()));
}
// static
scoped_refptr<AudioBuffer> AudioBuffer::CreateEmptyBuffer(
int channel_count,
int frame_count,
const base::TimeDelta timestamp,
const base::TimeDelta duration) {
CHECK_GT(frame_count, 0); // Otherwise looks like an EOF buffer.
// Since data == NULL, format doesn't matter.
return make_scoped_refptr(new AudioBuffer(kSampleFormatF32,
channel_count,
frame_count,
false,
NULL,
timestamp,
duration));
}
// static
scoped_refptr<AudioBuffer> AudioBuffer::CreateEOSBuffer() {
return make_scoped_refptr(new AudioBuffer(
kUnknownSampleFormat, 1, 0, false, NULL, kNoTimestamp(), kNoTimestamp()));
}
// Convert int16 values in the range [kint16min, kint16max] to [-1.0, 1.0].
static inline float ConvertS16ToFloat(int16 value) {
return value * (value < 0 ? -1.0f / kint16min : 1.0f / kint16max);
}
void AudioBuffer::ReadFrames(int frames_to_copy,
int source_frame_offset,
int dest_frame_offset,
AudioBus* dest) {
// Deinterleave each channel (if necessary) and convert to 32bit
// floating-point with nominal range -1.0 -> +1.0 (if necessary).
// |dest| must have the same number of channels, and the number of frames
// specified must be in range.
DCHECK(!end_of_stream());
DCHECK_EQ(dest->channels(), channel_count_);
DCHECK_LE(source_frame_offset + frames_to_copy, adjusted_frame_count_);
DCHECK_LE(dest_frame_offset + frames_to_copy, dest->frames());
// Move the start past any frames that have been trimmed.
source_frame_offset += trim_start_;
if (!data_) {
// Special case for an empty buffer.
dest->ZeroFramesPartial(dest_frame_offset, frames_to_copy);
return;
}
if (sample_format_ == kSampleFormatPlanarF32) {
// Format is planar float32. Copy the data from each channel as a block.
for (int ch = 0; ch < channel_count_; ++ch) {
const float* source_data =
reinterpret_cast<const float*>(channel_data_[ch]) +
source_frame_offset;
memcpy(dest->channel(ch) + dest_frame_offset,
source_data,
sizeof(float) * frames_to_copy);
}
return;
}
if (sample_format_ == kSampleFormatPlanarS16) {
// Format is planar signed16. Convert each value into float and insert into
// output channel data.
for (int ch = 0; ch < channel_count_; ++ch) {
const int16* source_data =
reinterpret_cast<const int16*>(channel_data_[ch]) +
source_frame_offset;
float* dest_data = dest->channel(ch) + dest_frame_offset;
for (int i = 0; i < frames_to_copy; ++i) {
dest_data[i] = ConvertS16ToFloat(source_data[i]);
}
}
return;
}
if (sample_format_ == kSampleFormatF32) {
// Format is interleaved float32. Copy the data into each channel.
const float* source_data = reinterpret_cast<const float*>(data_.get()) +
source_frame_offset * channel_count_;
for (int ch = 0; ch < channel_count_; ++ch) {
float* dest_data = dest->channel(ch) + dest_frame_offset;
for (int i = 0, offset = ch; i < frames_to_copy;
++i, offset += channel_count_) {
dest_data[i] = source_data[offset];
}
}
return;
}
// Remaining formats are integer interleaved data. Use the deinterleaving code
// in AudioBus to copy the data.
DCHECK(sample_format_ == kSampleFormatU8 ||
sample_format_ == kSampleFormatS16 ||
sample_format_ == kSampleFormatS32);
int bytes_per_channel = SampleFormatToBytesPerChannel(sample_format_);
int frame_size = channel_count_ * bytes_per_channel;
const uint8* source_data = data_.get() + source_frame_offset * frame_size;
dest->FromInterleavedPartial(
source_data, dest_frame_offset, frames_to_copy, bytes_per_channel);
}
void AudioBuffer::TrimStart(int frames_to_trim) {
CHECK_GE(frames_to_trim, 0);
CHECK_LE(frames_to_trim, adjusted_frame_count_);
// Adjust timestamp_ and duration_ to reflect the smaller number of frames.
double offset = static_cast<double>(duration_.InMicroseconds()) *
frames_to_trim / adjusted_frame_count_;
base::TimeDelta offset_as_time =
base::TimeDelta::FromMicroseconds(static_cast<int64>(offset));
timestamp_ += offset_as_time;
duration_ -= offset_as_time;
// Finally adjust the number of frames in this buffer and where the start
// really is.
adjusted_frame_count_ -= frames_to_trim;
trim_start_ += frames_to_trim;
}
void AudioBuffer::TrimEnd(int frames_to_trim) {
CHECK_GE(frames_to_trim, 0);
CHECK_LE(frames_to_trim, adjusted_frame_count_);
// Adjust duration_ only to reflect the smaller number of frames.
double offset = static_cast<double>(duration_.InMicroseconds()) *
frames_to_trim / adjusted_frame_count_;
base::TimeDelta offset_as_time =
base::TimeDelta::FromMicroseconds(static_cast<int64>(offset));
duration_ -= offset_as_time;
// Finally adjust the number of frames in this buffer.
adjusted_frame_count_ -= frames_to_trim;
}
} // namespace media