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
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
|
/*
Copyright (C) 2021 Carl Hetherington <cth@carlh.net>
This file is part of DCP-o-matic.
DCP-o-matic 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.
DCP-o-matic 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 DCP-o-matic. If not, see <http://www.gnu.org/licenses/>.
*/
#include "audio_analyser.h"
#include "audio_analysis.h"
#include "audio_buffers.h"
#include "audio_content.h"
#include "audio_filter_graph.h"
#include "audio_point.h"
#include "config.h"
#include "dcpomatic_log.h"
#include "film.h"
#include "filter.h"
#include "playlist.h"
#include <dcp/warnings.h>
extern "C" {
#include <leqm_nrt.h>
LIBDCP_DISABLE_WARNINGS
#include <libavutil/channel_layout.h>
#ifdef DCPOMATIC_HAVE_EBUR128_PATCHED_FFMPEG
#include <libavfilter/f_ebur128.h>
#endif
LIBDCP_ENABLE_WARNINGS
}
using std::make_shared;
using std::max;
using std::shared_ptr;
using std::vector;
using namespace dcpomatic;
static auto constexpr num_points = 1024;
AudioAnalyser::AudioAnalyser(shared_ptr<const Film> film, shared_ptr<const Playlist> playlist, bool whole_film, std::function<void (float)> set_progress)
: _film (film)
, _playlist (playlist)
, _set_progress (set_progress)
#ifdef DCPOMATIC_HAVE_EBUR128_PATCHED_FFMPEG
, _ebur128(film->audio_frame_rate(), film->audio_channels())
#endif
, _sample_peak (film->audio_channels())
, _sample_peak_frame (film->audio_channels())
, _analysis (film->audio_channels())
{
#ifdef DCPOMATIC_HAVE_EBUR128_PATCHED_FFMPEG
_filters.push_back({"ebur128", "ebur128", "audio", "ebur128=peak=true"});
_ebur128.setup(_filters);
#endif
_current = std::vector<AudioPoint>(_film->audio_channels());
if (!whole_film) {
_start = _playlist->start().get_value_or(DCPTime());
}
for (int i = 0; i < film->audio_channels(); ++i) {
_sample_peak[i] = 0;
_sample_peak_frame[i] = 0;
}
auto add_if_required = [](vector<double>& v, size_t i, double db) {
if (v.size() > i) {
v[i] = pow(10, db / 20);
}
};
auto content = _playlist->content();
if (whole_film) {
_leqm_channels = film->audio_channels();
} else {
_leqm_channels = 0;
for (auto channel: content[0]->audio->mapping().mapped_output_channels()) {
/* This means that if, for example, a file only maps C we will
* calculate LEQ(m) for L, R and C. I'm not sure if this is
* right or not.
*/
_leqm_channels = std::min(film->audio_channels(), channel + 1);
}
}
/* XXX: is this right? Especially for more than 5.1? */
vector<double> channel_corrections(_leqm_channels, 1);
add_if_required (channel_corrections, 4, -3); // Ls
add_if_required (channel_corrections, 5, -3); // Rs
add_if_required (channel_corrections, 6, -144); // HI
add_if_required (channel_corrections, 7, -144); // VI
add_if_required (channel_corrections, 8, -3); // Lc
add_if_required (channel_corrections, 9, -3); // Rc
add_if_required (channel_corrections, 10, -3); // Lc
add_if_required (channel_corrections, 11, -3); // Rc
add_if_required (channel_corrections, 12, -144); // DBox
add_if_required (channel_corrections, 13, -144); // Sync
add_if_required (channel_corrections, 14, -144); // Sign Language
add_if_required (channel_corrections, 15, -144); // Unused
_leqm.reset(new leqm_nrt::Calculator(
_leqm_channels,
film->audio_frame_rate(),
24,
channel_corrections,
850, // suggested by leqm_nrt CLI source
64, // suggested by leqm_nrt CLI source
boost::thread::hardware_concurrency()
));
DCPTime const length = _playlist->length (_film);
Frame const len = DCPTime (length - _start).frames_round (film->audio_frame_rate());
_samples_per_point = max (int64_t (1), len / num_points);
}
void
AudioAnalyser::analyse (shared_ptr<AudioBuffers> b, DCPTime time)
{
LOG_DEBUG_AUDIO_ANALYSIS("AudioAnalyser received {} frames at {}", b->frames(), time.to_debug_string());
DCPOMATIC_ASSERT (time >= _start);
/* In bug #2364 we had a lot of frames arriving here (~47s worth) which
* caused an OOM error on Windows. Check for the number of frames being
* reasonable here to make sure we catch this if it happens again.
*/
DCPOMATIC_ASSERT(b->frames() < 480000);
#ifdef DCPOMATIC_HAVE_EBUR128_PATCHED_FFMPEG
if (Config::instance()->analyse_ebur128 ()) {
_ebur128.process(b);
}
#endif
int const frames = b->frames ();
vector<double> interleaved(frames * _leqm_channels);
for (int j = 0; j < _leqm_channels; ++j) {
float const* data = b->data(j);
for (int i = 0; i < frames; ++i) {
float s = data[i];
interleaved[i * _leqm_channels + j] = s;
float as = fabsf (s);
if (as < 10e-7) {
/* We may struggle to serialise and recover inf or -inf, so prevent such
values by replacing with this (140dB down) */
s = as = 10e-7;
}
_current[j][AudioPoint::RMS] += pow (s, 2);
_current[j][AudioPoint::PEAK] = max (_current[j][AudioPoint::PEAK], as);
if (as > _sample_peak[j]) {
_sample_peak[j] = as;
_sample_peak_frame[j] = _done + i;
}
if (((_done + i) % _samples_per_point) == 0) {
_current[j][AudioPoint::RMS] = sqrt (_current[j][AudioPoint::RMS] / _samples_per_point);
_analysis.add_point (j, _current[j]);
_current[j] = AudioPoint ();
}
}
}
_leqm->add(interleaved);
_done += frames;
DCPTime const length = _playlist->length (_film);
_set_progress ((time.seconds() - _start.seconds()) / (length.seconds() - _start.seconds()));
LOG_DEBUG_AUDIO_ANALYSIS_NC("Frames processed");
}
void
AudioAnalyser::finish ()
{
vector<AudioAnalysis::PeakTime> sample_peak;
for (int i = 0; i < _film->audio_channels(); ++i) {
sample_peak.push_back (
AudioAnalysis::PeakTime(_sample_peak[i], DCPTime(_sample_peak_frame[i], _film->audio_frame_rate()))
);
}
_analysis.set_sample_peak (sample_peak);
#ifdef DCPOMATIC_HAVE_EBUR128_PATCHED_FFMPEG
if (Config::instance()->analyse_ebur128 ()) {
void* eb = _ebur128.get("Parsed_ebur128_0")->priv;
vector<float> true_peak;
for (int i = 0; i < _film->audio_channels(); ++i) {
true_peak.push_back (av_ebur128_get_true_peaks(eb)[i]);
}
_analysis.set_true_peak (true_peak);
_analysis.set_integrated_loudness (av_ebur128_get_integrated_loudness(eb));
_analysis.set_loudness_range (av_ebur128_get_loudness_range(eb));
}
#endif
if (_playlist->content().size() == 1) {
/* If there was only one piece of content in this analysis we may later need to know what its
gain was when we analysed it.
*/
if (auto ac = _playlist->content().front()->audio) {
_analysis.set_analysis_gain (ac->gain());
}
}
_analysis.set_samples_per_point (_samples_per_point);
_analysis.set_sample_rate (_film->audio_frame_rate ());
_analysis.set_leqm (_leqm->leq_m());
}
|
