1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
|
/*
* Copyright (c) 2017 The WebRTC project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "modules/audio_mixer/frame_combiner.h"
#include <algorithm>
#include <array>
#include <cstddef>
#include <cstdint>
#include <memory>
#include <utility>
#include <vector>
#include "api/array_view.h"
#include "api/audio/audio_frame.h"
#include "api/audio/audio_view.h"
#include "api/rtp_packet_info.h"
#include "api/rtp_packet_infos.h"
#include "common_audio/include/audio_util.h"
#include "modules/audio_mixer/audio_frame_manipulator.h"
#include "modules/audio_processing/agc2/limiter.h"
#include "modules/audio_processing/logging/apm_data_dumper.h"
#include "rtc_base/checks.h"
namespace webrtc {
namespace {
void SetAudioFrameFields(ArrayView<const AudioFrame* const> mix_list,
size_t number_of_channels,
int sample_rate,
size_t /* number_of_streams */,
AudioFrame* audio_frame_for_mixing) {
const size_t samples_per_channel =
SampleRateToDefaultChannelSize(sample_rate);
// TODO(minyue): Issue bugs.webrtc.org/3390.
// Audio frame timestamp. The 'timestamp_' field is set to dummy
// value '0', because it is only supported in the one channel case and
// is then updated in the helper functions.
audio_frame_for_mixing->UpdateFrame(
0, nullptr, samples_per_channel, sample_rate, AudioFrame::kUndefined,
AudioFrame::kVadUnknown, number_of_channels);
if (mix_list.empty()) {
audio_frame_for_mixing->elapsed_time_ms_ = -1;
} else {
audio_frame_for_mixing->timestamp_ = mix_list[0]->timestamp_;
audio_frame_for_mixing->elapsed_time_ms_ = mix_list[0]->elapsed_time_ms_;
audio_frame_for_mixing->ntp_time_ms_ = mix_list[0]->ntp_time_ms_;
std::vector<RtpPacketInfo> packet_infos;
for (const auto& frame : mix_list) {
audio_frame_for_mixing->timestamp_ =
std::min(audio_frame_for_mixing->timestamp_, frame->timestamp_);
audio_frame_for_mixing->ntp_time_ms_ =
std::min(audio_frame_for_mixing->ntp_time_ms_, frame->ntp_time_ms_);
audio_frame_for_mixing->elapsed_time_ms_ = std::max(
audio_frame_for_mixing->elapsed_time_ms_, frame->elapsed_time_ms_);
packet_infos.insert(packet_infos.end(), frame->packet_infos_.begin(),
frame->packet_infos_.end());
}
audio_frame_for_mixing->packet_infos_ =
RtpPacketInfos(std::move(packet_infos));
}
}
void MixFewFramesWithNoLimiter(ArrayView<const AudioFrame* const> mix_list,
AudioFrame* audio_frame_for_mixing) {
if (mix_list.empty()) {
audio_frame_for_mixing->Mute();
return;
}
RTC_DCHECK_LE(mix_list.size(), 1);
InterleavedView<int16_t> dst = audio_frame_for_mixing->mutable_data(
mix_list[0]->samples_per_channel_, mix_list[0]->num_channels_);
CopySamples(dst, mix_list[0]->data_view());
}
void MixToFloatFrame(ArrayView<const AudioFrame* const> mix_list,
DeinterleavedView<float>& mixing_buffer) {
const size_t number_of_channels = NumChannels(mixing_buffer);
// Clear the mixing buffer.
mixing_buffer.Clear();
// Convert to FloatS16 and mix.
for (size_t i = 0; i < mix_list.size(); ++i) {
InterleavedView<const int16_t> frame_data = mix_list[i]->data_view();
RTC_CHECK(!frame_data.empty());
for (size_t j = 0; j < number_of_channels; ++j) {
MonoView<float> channel = mixing_buffer[j];
for (size_t k = 0; k < SamplesPerChannel(channel); ++k) {
channel[k] += frame_data[number_of_channels * k + j];
}
}
}
}
void RunLimiter(DeinterleavedView<float> deinterleaved, Limiter* limiter) {
limiter->SetSamplesPerChannel(deinterleaved.samples_per_channel());
limiter->Process(deinterleaved);
}
// Both interleaves and rounds.
void InterleaveToAudioFrame(DeinterleavedView<float> deinterleaved,
AudioFrame* audio_frame_for_mixing) {
InterleavedView<int16_t> mixing_data = audio_frame_for_mixing->mutable_data(
deinterleaved.samples_per_channel(), deinterleaved.num_channels());
// Put data in the result frame.
for (size_t i = 0; i < mixing_data.num_channels(); ++i) {
auto channel = deinterleaved[i];
for (size_t j = 0; j < mixing_data.samples_per_channel(); ++j) {
mixing_data[mixing_data.num_channels() * j + i] =
FloatS16ToS16(channel[j]);
}
}
}
} // namespace
constexpr size_t FrameCombiner::kMaximumNumberOfChannels;
constexpr size_t FrameCombiner::kMaximumChannelSize;
FrameCombiner::FrameCombiner(bool use_limiter)
: data_dumper_(new ApmDataDumper(0)),
limiter_(data_dumper_.get(), kMaximumChannelSize, "AudioMixer"),
use_limiter_(use_limiter) {
static_assert(kMaximumChannelSize * kMaximumNumberOfChannels <=
AudioFrame::kMaxDataSizeSamples,
"");
}
FrameCombiner::~FrameCombiner() = default;
void FrameCombiner::Combine(ArrayView<AudioFrame* const> mix_list,
size_t number_of_channels,
int sample_rate,
size_t number_of_streams,
AudioFrame* audio_frame_for_mixing) {
RTC_DCHECK(audio_frame_for_mixing);
RTC_DCHECK_GT(sample_rate, 0);
// Note: `mix_list` is allowed to be empty.
// See FrameCombiner.CombiningZeroFramesShouldProduceSilence.
// Make sure to cap `number_of_channels` to the kMaximumNumberOfChannels
// limits since processing from hereon out will be bound by them.
number_of_channels = std::min(number_of_channels, kMaximumNumberOfChannels);
SetAudioFrameFields(mix_list, number_of_channels, sample_rate,
number_of_streams, audio_frame_for_mixing);
size_t samples_per_channel = SampleRateToDefaultChannelSize(sample_rate);
#if RTC_DCHECK_IS_ON
for (const auto* frame : mix_list) {
RTC_DCHECK_EQ(samples_per_channel, frame->samples_per_channel_);
RTC_DCHECK_EQ(sample_rate, frame->sample_rate_hz_);
}
#endif
// The 'num_channels_' field of frames in 'mix_list' could be
// different from 'number_of_channels'.
for (auto* frame : mix_list) {
RemixFrame(number_of_channels, frame);
}
if (number_of_streams <= 1) {
MixFewFramesWithNoLimiter(mix_list, audio_frame_for_mixing);
return;
}
// Make sure that the size of the view based on the desired
// `samples_per_channel` and `number_of_channels` doesn't exceed the size of
// the `mixing_buffer_` buffer.
RTC_DCHECK_LE(samples_per_channel, kMaximumChannelSize);
// Since the above check is a DCHECK only, clamp down on `samples_per_channel`
// to make sure we don't exceed the buffer size in non-dcheck builds.
// See also FrameCombinerDeathTest.DebugBuildCrashesWithHighRate.
samples_per_channel = std::min(samples_per_channel, kMaximumChannelSize);
DeinterleavedView<float> deinterleaved(
mixing_buffer_.data(), samples_per_channel, number_of_channels);
MixToFloatFrame(mix_list, deinterleaved);
if (use_limiter_) {
RunLimiter(deinterleaved, &limiter_);
}
InterleaveToAudioFrame(deinterleaved, audio_frame_for_mixing);
}
} // namespace webrtc
|
