Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(345)

Unified Diff: webrtc/modules/audio_processing/intelligibility/intelligibility_proc.cc

Issue 1187033005: Revert of Allow intelligibility to compile in apm (Closed) Base URL: https://chromium.googlesource.com/external/webrtc.git@master
Patch Set: Created 5 years, 6 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
Index: webrtc/modules/audio_processing/intelligibility/intelligibility_proc.cc
diff --git a/webrtc/modules/audio_processing/intelligibility/intelligibility_proc.cc b/webrtc/modules/audio_processing/intelligibility/intelligibility_proc.cc
index dbe3d35e019fc3a80388566ab92701a3507b6570..b0ea2dfee0315b19aa779ac3521e70d52ee5ff7c 100644
--- a/webrtc/modules/audio_processing/intelligibility/intelligibility_proc.cc
+++ b/webrtc/modules/audio_processing/intelligibility/intelligibility_proc.cc
@@ -7,12 +7,6 @@
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
-
-//
-// Command line tool for speech intelligibility enhancement. Provides for
-// running and testing intelligibility_enhancer as an independent process.
-// Use --help for options.
-//
#include <arpa/inet.h>
#include <fcntl.h>
@@ -30,71 +24,53 @@
#include <complex>
#include "gflags/gflags.h"
-#include "testing/gtest/include/gtest/gtest.h"
#include "webrtc/base/checks.h"
#include "webrtc/common_audio/real_fourier.h"
#include "webrtc/modules/audio_processing/intelligibility/intelligibility_enhancer.h"
#include "webrtc/modules/audio_processing/intelligibility/intelligibility_utils.h"
#include "webrtc/system_wrappers/interface/critical_section_wrapper.h"
+#include "webrtc/system_wrappers/interface/scoped_ptr.h"
-// PCM data simulating streams
const int16_t* in_ipcm;
int16_t* out_ipcm;
const int16_t* noise_ipcm;
+
float* in_fpcm;
float* out_fpcm;
float* noise_fpcm;
-
-// Current locations in streams
float* noise_cursor;
float* clear_cursor;
-int samples; // Number of samples in input PCM file
-int fragment_size; // Number of samples to process at a time
- // to simulate APM stream processing
+int samples;
+int fragment_size;
using std::complex;
-
-namespace webrtc {
-
using webrtc::RealFourier;
using webrtc::IntelligibilityEnhancer;
-DEFINE_int32(clear_type,
- webrtc::intelligibility::VarianceArray::kStepInfinite,
+DEFINE_int32(clear_type, webrtc::intelligibility::VarianceArray::kStepInfinite,
"Variance algorithm for clear data.");
-DEFINE_double(clear_alpha, 0.9, "Variance decay factor for clear data.");
-DEFINE_int32(clear_window,
- 475,
+DEFINE_double(clear_alpha, 0.9,
+ "Variance decay factor for clear data.");
+DEFINE_int32(clear_window, 475,
"Window size for windowed variance for clear data.");
-DEFINE_int32(sample_rate,
- 16000,
+DEFINE_int32(sample_rate, 16000,
"Audio sample rate used in the input and output files.");
-DEFINE_int32(ana_rate,
- 800,
+DEFINE_int32(ana_rate, 800,
"Analysis rate; gains recalculated every N blocks.");
-DEFINE_int32(
- var_rate,
- 2,
- "Variance clear rate; history is forgotten every N gain recalculations.");
+DEFINE_int32(var_rate, 2,
+ "Variance clear rate; history is forgotten every N gain recalculations.");
DEFINE_double(gain_limit, 1000.0, "Maximum gain change in one block.");
DEFINE_bool(repeat, false, "Repeat input file ad nauseam.");
DEFINE_string(clear_file, "speech.pcm", "Input file with clear speech.");
DEFINE_string(noise_file, "noise.pcm", "Input file with noise data.");
-DEFINE_string(out_file,
- "proc_enhanced.pcm",
- "Enhanced output. Use '-' to "
+DEFINE_string(out_file, "proc_enhanced.pcm", "Enhanced output. Use '-' to "
"pipe through aplay internally.");
-// Constant IntelligibilityEnhancer constructor parameters.
-const int kErbResolution = 2;
-const int kNumChannels = 1;
-
-// Converts output stream to Sun AU format and writes to file descriptor |fd|.
-// Can be used to pipe output directly into aplay.
-// TODO(ekmeyerson): Modify to write WAV instead.
+// Write an Sun AU-formatted audio chunk into file descriptor |fd|. Can be used
+// to pipe the audio stream directly into aplay.
void writeau(int fd) {
uint32_t thing;
@@ -116,14 +92,12 @@
write(fd, out_ipcm, sizeof(*out_ipcm) * samples);
}
-// void function for gtest
-void void_main(int argc, char* argv[]) {
- google::SetUsageMessage(
- "\n\nVariance algorithm types are:\n"
- " 0 - infinite/normal,\n"
- " 1 - exponentially decaying,\n"
- " 2 - rolling window.\n"
- "\nInput files must be little-endian 16-bit signed raw PCM.\n");
+int main(int argc, char* argv[]) {
+ google::SetUsageMessage("\n\nVariance algorithm types are:\n"
+ " 0 - infinite/normal,\n"
+ " 1 - exponentially decaying,\n"
+ " 2 - rolling window.\n"
+ "\nInput files must be little-endian 16-bit signed raw PCM.\n");
google::ParseCommandLineFlags(&argc, &argv, true);
const char* in_name = FLAGS_clear_file.c_str();
@@ -133,15 +107,10 @@
int in_fd, out_fd, noise_fd;
FILE* aplay_file = nullptr;
- // Load settings and set up PCMs.
+ fragment_size = FLAGS_sample_rate / 100;
- fragment_size = FLAGS_sample_rate / 100; // Mirror real time APM chunk size.
- // Duplicates chunk_length_ in
- // IntelligibilityEnhancer.
-
- ASSERT_EQ(stat(in_name, &in_stat), 0) << "Empty speech input.";
- ASSERT_EQ(stat(noise_name, &noise_stat), 0) << "Empty noise input.";
-
+ stat(in_name, &in_stat);
+ stat(noise_name, &noise_stat);
samples = in_stat.st_size / sizeof(*in_ipcm);
in_fd = open(in_name, O_RDONLY);
@@ -154,10 +123,10 @@
}
noise_fd = open(noise_name, O_RDONLY);
- in_ipcm = static_cast<int16_t*>(
- mmap(nullptr, in_stat.st_size, PROT_READ, MAP_PRIVATE, in_fd, 0));
- noise_ipcm = static_cast<int16_t*>(
- mmap(nullptr, noise_stat.st_size, PROT_READ, MAP_PRIVATE, noise_fd, 0));
+ in_ipcm = static_cast<int16_t*>(mmap(nullptr, in_stat.st_size, PROT_READ,
+ MAP_PRIVATE, in_fd, 0));
+ noise_ipcm = static_cast<int16_t*>(mmap(nullptr, noise_stat.st_size,
+ PROT_READ, MAP_PRIVATE, noise_fd, 0));
out_ipcm = new int16_t[samples];
out_fpcm = new float[samples];
in_fpcm = new float[samples];
@@ -167,17 +136,18 @@
noise_fpcm[i] = noise_ipcm[i % (noise_stat.st_size / sizeof(*noise_ipcm))];
}
- // Run intelligibility enhancement.
-
- IntelligibilityEnhancer enh(
- kErbResolution,
- FLAGS_sample_rate,
- kNumChannels,
- FLAGS_clear_type, static_cast<float>(FLAGS_clear_alpha),
- FLAGS_clear_window, FLAGS_ana_rate, FLAGS_var_rate, FLAGS_gain_limit);
+ //feenableexcept(FE_INVALID | FE_OVERFLOW);
+ IntelligibilityEnhancer enh(2,
+ FLAGS_sample_rate, 1,
+ FLAGS_clear_type,
+ static_cast<float>(FLAGS_clear_alpha),
+ FLAGS_clear_window,
+ FLAGS_ana_rate,
+ FLAGS_var_rate,
+ FLAGS_gain_limit);
// Slice the input into smaller chunks, as the APM would do, and feed them
- // through the enhancer. Repeat indefinitely if FLAGS_repeat is set.
+ // into the enhancer. Repeat indefinitely if FLAGS_repeat is set.
do {
noise_cursor = noise_fpcm;
clear_cursor = in_fpcm;
@@ -211,11 +181,7 @@
close(out_fd);
}
close(in_fd);
+
+ return 0;
}
-} // namespace webrtc
-
-int main(int argc, char* argv[]) {
- webrtc::void_main(argc, argv);
- return 0;
-}

Powered by Google App Engine
This is Rietveld 408576698