Deactivated the intelligibility enhancement functionality by default
NOTRY=true
BUG=
Review-Url: https://codereview.webrtc.org/2272423003
Cr-Commit-Position: refs/heads/master@{#13937}
diff --git a/webrtc/build/common.gypi b/webrtc/build/common.gypi
index 414fd02..c4a548f 100644
--- a/webrtc/build/common.gypi
+++ b/webrtc/build/common.gypi
@@ -118,6 +118,9 @@
# Enables the use of protocol buffers for debug recordings.
'enable_protobuf%': 1,
+ # Disable the code for the intelligibility enhancer by default.
+ 'enable_intelligibility_enhancer%': 0,
+
# Disable these to not build components which can be externally provided.
'build_expat%': 1,
'build_json%': 1,
diff --git a/webrtc/build/webrtc.gni b/webrtc/build/webrtc.gni
index b345d73..a6282b8 100644
--- a/webrtc/build/webrtc.gni
+++ b/webrtc/build/webrtc.gni
@@ -36,6 +36,9 @@
# Enables the use of protocol buffers for debug recordings.
rtc_enable_protobuf = true
+ # Disable the code for the intelligibility enhancer by default.
+ rtc_enable_intelligibility_enhancer = false
+
# Disable these to not build components which can be externally provided.
rtc_build_expat = true
rtc_build_json = true
diff --git a/webrtc/media/BUILD.gn b/webrtc/media/BUILD.gn
index e2d9217..5f9a0af 100644
--- a/webrtc/media/BUILD.gn
+++ b/webrtc/media/BUILD.gn
@@ -141,6 +141,12 @@
]
}
+ if (rtc_enable_intelligibility_enhancer) {
+ defines += [ "WEBRTC_INTELLIGIBILITY_ENHANCER=1" ]
+ } else {
+ defines += [ "WEBRTC_INTELLIGIBILITY_ENHANCER=0" ]
+ }
+
include_dirs = []
if (rtc_build_libyuv) {
deps += [ "$rtc_libyuv_dir" ]
diff --git a/webrtc/media/engine/webrtcvoiceengine.cc b/webrtc/media/engine/webrtcvoiceengine.cc
index a61b8c4..d0ca9ae 100644
--- a/webrtc/media/engine/webrtcvoiceengine.cc
+++ b/webrtc/media/engine/webrtcvoiceengine.cc
@@ -64,6 +64,14 @@
constexpr int kNackRtpHistoryMs = 5000;
+// Check to verify that the define for the intelligibility enhancer is properly
+// set.
+#if !defined(WEBRTC_INTELLIGIBILITY_ENHANCER) || \
+ (WEBRTC_INTELLIGIBILITY_ENHANCER != 0 && \
+ WEBRTC_INTELLIGIBILITY_ENHANCER != 1)
+#error "Set WEBRTC_INTELLIGIBILITY_ENHANCER to either 0 or 1"
+#endif
+
// Codec parameters for Opus.
// draft-spittka-payload-rtp-opus-03
@@ -649,6 +657,11 @@
}
#endif
+#if (WEBRTC_INTELLIGIBILITY_ENHANCER == 0)
+ // Hardcode the intelligibility enhancer to be off.
+ options.intelligibility_enhancer = rtc::Optional<bool>(false);
+#endif
+
webrtc::VoEAudioProcessing* voep = voe_wrapper_->processing();
if (options.echo_cancellation) {
diff --git a/webrtc/media/media.gyp b/webrtc/media/media.gyp
index f3f1102..ecd0a76 100644
--- a/webrtc/media/media.gyp
+++ b/webrtc/media/media.gyp
@@ -125,6 +125,11 @@
'<(DEPTH)/third_party/usrsctp/usrsctp.gyp:usrsctplib',
],
}],
+ ['enable_intelligibility_enhancer==1', {
+ 'defines': ['WEBRTC_INTELLIGIBILITY_ENHANCER=1',],
+ }, {
+ 'defines': ['WEBRTC_INTELLIGIBILITY_ENHANCER=0',],
+ }],
['build_with_chromium==1', {
'dependencies': [
'<(webrtc_root)/modules/modules.gyp:video_capture',
diff --git a/webrtc/modules/BUILD.gn b/webrtc/modules/BUILD.gn
index c04f00c..873c978 100644
--- a/webrtc/modules/BUILD.gn
+++ b/webrtc/modules/BUILD.gn
@@ -199,8 +199,6 @@
"audio_processing/beamformer/mock_nonlinear_beamformer.h",
"audio_processing/beamformer/nonlinear_beamformer_unittest.cc",
"audio_processing/echo_cancellation_impl_unittest.cc",
- "audio_processing/intelligibility/intelligibility_enhancer_unittest.cc",
- "audio_processing/intelligibility/intelligibility_utils_unittest.cc",
"audio_processing/splitting_filter_unittest.cc",
"audio_processing/transient/dyadic_decimator_unittest.cc",
"audio_processing/transient/file_utils.cc",
@@ -348,6 +346,16 @@
"video_processing/test/video_processing_unittest.h",
]
+ if (rtc_enable_intelligibility_enhancer) {
+ defines += [ "WEBRTC_INTELLIGIBILITY_ENHANCER=1" ]
+ sources += [
+ "audio_processing/intelligibility/intelligibility_enhancer_unittest.cc",
+ "audio_processing/intelligibility/intelligibility_utils_unittest.cc",
+ ]
+ } else {
+ defines += [ "WEBRTC_INTELLIGIBILITY_ENHANCER=0" ]
+ }
+
if (rtc_libvpx_build_vp9) {
sources +=
[ "video_coding/codecs/vp9/vp9_screenshare_layers_unittest.cc" ]
diff --git a/webrtc/modules/audio_processing/BUILD.gn b/webrtc/modules/audio_processing/BUILD.gn
index 21c89d3..97705c6 100644
--- a/webrtc/modules/audio_processing/BUILD.gn
+++ b/webrtc/modules/audio_processing/BUILD.gn
@@ -74,10 +74,6 @@
"high_pass_filter_impl.cc",
"high_pass_filter_impl.h",
"include/audio_processing.h",
- "intelligibility/intelligibility_enhancer.cc",
- "intelligibility/intelligibility_enhancer.h",
- "intelligibility/intelligibility_utils.cc",
- "intelligibility/intelligibility_utils.h",
"level_controller/biquad_filter.cc",
"level_controller/biquad_filter.h",
"level_controller/down_sampler.cc",
@@ -182,6 +178,18 @@
deps += [ ":audioproc_debug_proto" ]
}
+ if (rtc_enable_intelligibility_enhancer) {
+ defines += [ "WEBRTC_INTELLIGIBILITY_ENHANCER=1" ]
+ sources += [
+ "intelligibility/intelligibility_enhancer.cc",
+ "intelligibility/intelligibility_enhancer.h",
+ "intelligibility/intelligibility_utils.cc",
+ "intelligibility/intelligibility_utils.h",
+ ]
+ } else {
+ defines += [ "WEBRTC_INTELLIGIBILITY_ENHANCER=0" ]
+ }
+
if (rtc_prefer_fixed_point) {
defines += [ "WEBRTC_NS_FIXED" ]
sources += [
@@ -481,22 +489,24 @@
}
}
- executable("intelligibility_proc") {
- testonly = true
- sources = [
- "intelligibility/test/intelligibility_proc.cc",
- ]
- deps = [
- ":audio_processing",
- ":audioproc_test_utils",
- "../../system_wrappers:metrics_default",
- "../../test:test_support",
- "//testing/gtest",
- "//third_party/gflags",
- ]
- if (is_clang) {
- # Suppress warnings from the Chromium Clang plugins (bugs.webrtc.org/163).
- configs -= [ "//build/config/clang:find_bad_constructs" ]
+ if (rtc_enable_intelligibility_enhancer) {
+ executable("intelligibility_proc") {
+ testonly = true
+ sources = [
+ "intelligibility/test/intelligibility_proc.cc",
+ ]
+ deps = [
+ ":audio_processing",
+ ":audioproc_test_utils",
+ "../../system_wrappers:metrics_default",
+ "../../test:test_support",
+ "//testing/gtest",
+ "//third_party/gflags",
+ ]
+ if (is_clang) {
+ # Suppress warnings from the Chromium Clang plugins (bugs.webrtc.org/163).
+ configs -= [ "//build/config/clang:find_bad_constructs" ]
+ }
}
}
diff --git a/webrtc/modules/audio_processing/audio_processing.gypi b/webrtc/modules/audio_processing/audio_processing.gypi
index 3cfc727..14e1b66 100644
--- a/webrtc/modules/audio_processing/audio_processing.gypi
+++ b/webrtc/modules/audio_processing/audio_processing.gypi
@@ -85,10 +85,6 @@
'high_pass_filter_impl.cc',
'high_pass_filter_impl.h',
'include/audio_processing.h',
- 'intelligibility/intelligibility_enhancer.cc',
- 'intelligibility/intelligibility_enhancer.h',
- 'intelligibility/intelligibility_utils.cc',
- 'intelligibility/intelligibility_utils.h',
'level_controller/biquad_filter.cc',
'level_controller/biquad_filter.h',
'level_controller/down_sampler.cc',
@@ -184,6 +180,17 @@
'dependencies': ['audioproc_debug_proto'],
'defines': ['WEBRTC_AUDIOPROC_DEBUG_DUMP'],
}],
+ ['enable_intelligibility_enhancer==1', {
+ 'defines': ['WEBRTC_INTELLIGIBILITY_ENHANCER=1',],
+ 'sources': [
+ 'intelligibility/intelligibility_enhancer.cc',
+ 'intelligibility/intelligibility_enhancer.h',
+ 'intelligibility/intelligibility_utils.cc',
+ 'intelligibility/intelligibility_utils.h',
+ ],
+ }, {
+ 'defines': ['WEBRTC_INTELLIGIBILITY_ENHANCER=0',],
+ }],
['prefer_fixed_point==1', {
'defines': ['WEBRTC_NS_FIXED'],
'sources': [
diff --git a/webrtc/modules/audio_processing/audio_processing_impl.cc b/webrtc/modules/audio_processing/audio_processing_impl.cc
index 3b3a951..011325f 100644
--- a/webrtc/modules/audio_processing/audio_processing_impl.cc
+++ b/webrtc/modules/audio_processing/audio_processing_impl.cc
@@ -30,7 +30,9 @@
#include "webrtc/modules/audio_processing/gain_control_for_experimental_agc.h"
#include "webrtc/modules/audio_processing/gain_control_impl.h"
#include "webrtc/modules/audio_processing/high_pass_filter_impl.h"
+#if WEBRTC_INTELLIGIBILITY_ENHANCER
#include "webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h"
+#endif
#include "webrtc/modules/audio_processing/level_controller/level_controller.h"
#include "webrtc/modules/audio_processing/level_estimator_impl.h"
#include "webrtc/modules/audio_processing/noise_suppression_impl.h"
@@ -50,6 +52,14 @@
#endif
#endif // WEBRTC_AUDIOPROC_DEBUG_DUMP
+// Check to verify that the define for the intelligibility enhancer is properly
+// set.
+#if !defined(WEBRTC_INTELLIGIBILITY_ENHANCER) || \
+ (WEBRTC_INTELLIGIBILITY_ENHANCER != 0 && \
+ WEBRTC_INTELLIGIBILITY_ENHANCER != 1)
+#error "Set WEBRTC_INTELLIGIBILITY_ENHANCER to either 0 or 1"
+#endif
+
#define RETURN_ON_ERR(expr) \
do { \
int err = (expr); \
@@ -124,7 +134,9 @@
// Accessed internally from both render and capture.
std::unique_ptr<TransientSuppressor> transient_suppressor;
+#if WEBRTC_INTELLIGIBILITY_ENHANCER
std::unique_ptr<IntelligibilityEnhancer> intelligibility_enhancer;
+#endif
};
struct AudioProcessingImpl::ApmPrivateSubmodules {
@@ -321,7 +333,9 @@
InitializeExperimentalAgc();
InitializeTransient();
InitializeBeamformer();
+#if WEBRTC_INTELLIGIBILITY_ENHANCER
InitializeIntelligibility();
+#endif
InitializeHighPassFilter();
InitializeNoiseSuppression();
InitializeLevelEstimator();
@@ -423,12 +437,14 @@
InitializeLevelController();
}
+#if WEBRTC_INTELLIGIBILITY_ENHANCER
if(capture_nonlocked_.intelligibility_enabled !=
config.Get<Intelligibility>().enabled) {
capture_nonlocked_.intelligibility_enabled =
config.Get<Intelligibility>().enabled;
InitializeIntelligibility();
}
+#endif
#ifdef WEBRTC_ANDROID_PLATFORM_BUILD
if (capture_nonlocked_.beamformer_enabled !=
@@ -725,6 +741,7 @@
ca->CopyLowPassToReference();
}
public_submodules_->noise_suppression->ProcessCaptureAudio(ca);
+#if WEBRTC_INTELLIGIBILITY_ENHANCER
if (capture_nonlocked_.intelligibility_enabled) {
RTC_DCHECK(public_submodules_->noise_suppression->is_enabled());
int gain_db = public_submodules_->gain_control->is_enabled() ?
@@ -737,6 +754,7 @@
public_submodules_->intelligibility_enhancer->SetCaptureNoiseEstimate(
public_submodules_->noise_suppression->NoiseEstimate(), gain);
}
+#endif
// Ensure that the stream delay was set before the call to the
// AECM ProcessCaptureAudio function.
@@ -936,11 +954,13 @@
ra->SplitIntoFrequencyBands();
}
+#if WEBRTC_INTELLIGIBILITY_ENHANCER
if (capture_nonlocked_.intelligibility_enabled) {
public_submodules_->intelligibility_enhancer->ProcessRenderAudio(
ra->split_channels_f(kBand0To8kHz), capture_nonlocked_.split_rate,
ra->num_channels());
}
+#endif
RETURN_ON_ERR(public_submodules_->echo_cancellation->ProcessRenderAudio(ra));
RETURN_ON_ERR(
@@ -1172,7 +1192,11 @@
}
bool AudioProcessingImpl::is_rev_processed() const {
+#if WEBRTC_INTELLIGIBILITY_ENHANCER
return capture_nonlocked_.intelligibility_enabled;
+#else
+ return false;
+#endif
}
bool AudioProcessingImpl::rev_synthesis_needed() const {
@@ -1237,12 +1261,14 @@
}
void AudioProcessingImpl::InitializeIntelligibility() {
+#if WEBRTC_INTELLIGIBILITY_ENHANCER
if (capture_nonlocked_.intelligibility_enabled) {
public_submodules_->intelligibility_enhancer.reset(
new IntelligibilityEnhancer(capture_nonlocked_.split_rate,
render_.render_audio->num_channels(),
NoiseSuppressionImpl::num_noise_bins()));
}
+#endif
}
void AudioProcessingImpl::InitializeHighPassFilter() {
diff --git a/webrtc/modules/audio_processing/audio_processing_tests.gypi b/webrtc/modules/audio_processing/audio_processing_tests.gypi
index 87598ed..d68fed3 100644
--- a/webrtc/modules/audio_processing/audio_processing_tests.gypi
+++ b/webrtc/modules/audio_processing/audio_processing_tests.gypi
@@ -61,22 +61,29 @@
'beamformer/nonlinear_beamformer_test.cc',
],
}, # nonlinear_beamformer_test
- {
- 'target_name': 'intelligibility_proc',
- 'type': 'executable',
- 'dependencies': [
- 'audioproc_test_utils',
- '<(DEPTH)/third_party/gflags/gflags.gyp:gflags',
- '<(DEPTH)/testing/gtest.gyp:gtest',
- '<(webrtc_root)/modules/modules.gyp:audio_processing',
- '<(webrtc_root)/test/test.gyp:test_support',
- ],
- 'sources': [
- 'intelligibility/test/intelligibility_proc.cc',
- ],
- }, # intelligibility_proc
],
'conditions': [
+ ['enable_intelligibility_enhancer==1', {
+ 'defines': ['WEBRTC_INTELLIGIBILITY_ENHANCER=1',],
+ 'targets': [
+ {
+ 'target_name': 'intelligibility_proc',
+ 'type': 'executable',
+ 'dependencies': [
+ 'audioproc_test_utils',
+ '<(DEPTH)/third_party/gflags/gflags.gyp:gflags',
+ '<(DEPTH)/testing/gtest.gyp:gtest',
+ '<(webrtc_root)/modules/modules.gyp:audio_processing',
+ '<(webrtc_root)/test/test.gyp:test_support',
+ ],
+ 'sources': [
+ 'intelligibility/test/intelligibility_proc.cc',
+ ],
+ },
+ ],
+ }, {
+ 'defines': ['WEBRTC_INTELLIGIBILITY_ENHANCER=0',],
+ }],
['enable_protobuf==1', {
'targets': [
{
diff --git a/webrtc/modules/modules.gyp b/webrtc/modules/modules.gyp
index 582b750..16fb818 100644
--- a/webrtc/modules/modules.gyp
+++ b/webrtc/modules/modules.gyp
@@ -249,8 +249,6 @@
'audio_processing/beamformer/mock_nonlinear_beamformer.h',
'audio_processing/beamformer/nonlinear_beamformer_unittest.cc',
'audio_processing/echo_cancellation_impl_unittest.cc',
- 'audio_processing/intelligibility/intelligibility_enhancer_unittest.cc',
- 'audio_processing/intelligibility/intelligibility_utils_unittest.cc',
'audio_processing/splitting_filter_unittest.cc',
'audio_processing/transient/dyadic_decimator_unittest.cc',
'audio_processing/transient/file_utils.cc',
@@ -398,6 +396,15 @@
'video_processing/test/video_processing_unittest.h',
],
'conditions': [
+ ['enable_intelligibility_enhancer==1', {
+ 'defines': ['WEBRTC_INTELLIGIBILITY_ENHANCER=1',],
+ 'sources': [
+ 'audio_processing/intelligibility/intelligibility_enhancer_unittest.cc',
+ 'audio_processing/intelligibility/intelligibility_utils_unittest.cc',
+ ],
+ }, {
+ 'defines': ['WEBRTC_INTELLIGIBILITY_ENHANCER=0',],
+ }],
['libvpx_build_vp9==1', {
'sources': [
'video_coding/codecs/vp9/vp9_screenshare_layers_unittest.cc',