cf808d2366
This change instroduces a mode where the Accelerate operation will be more aggressive. When enabled, it will allow acceleration at lower correlation levels, and possibly remove multiple pitch periods at once. The feature is enabled through NetEq::Config, and is off by default. This means that bit-exactness tests are currently not affected. A unit test was added for the Accelerate class, with and without fast mode enabled. BUG=4691 R=minyue@webrtc.org Review URL: https://webrtc-codereview.appspot.com/50039004 Cr-Commit-Position: refs/heads/master@{#9295}
117 lines
4.1 KiB
C++
117 lines
4.1 KiB
C++
/*
|
|
* Copyright (c) 2012 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.
|
|
*/
|
|
|
|
#ifndef WEBRTC_MODULES_AUDIO_CODING_NETEQ_TIME_STRETCH_H_
|
|
#define WEBRTC_MODULES_AUDIO_CODING_NETEQ_TIME_STRETCH_H_
|
|
|
|
#include <assert.h>
|
|
#include <string.h> // memset, size_t
|
|
|
|
#include "webrtc/base/constructormagic.h"
|
|
#include "webrtc/modules/audio_coding/neteq/audio_multi_vector.h"
|
|
#include "webrtc/typedefs.h"
|
|
|
|
namespace webrtc {
|
|
|
|
// Forward declarations.
|
|
class BackgroundNoise;
|
|
|
|
// This is the base class for Accelerate and PreemptiveExpand. This class
|
|
// cannot be instantiated, but must be used through either of the derived
|
|
// classes.
|
|
class TimeStretch {
|
|
public:
|
|
enum ReturnCodes {
|
|
kSuccess = 0,
|
|
kSuccessLowEnergy = 1,
|
|
kNoStretch = 2,
|
|
kError = -1
|
|
};
|
|
|
|
TimeStretch(int sample_rate_hz, size_t num_channels,
|
|
const BackgroundNoise& background_noise)
|
|
: sample_rate_hz_(sample_rate_hz),
|
|
fs_mult_(sample_rate_hz / 8000),
|
|
num_channels_(static_cast<int>(num_channels)),
|
|
master_channel_(0), // First channel is master.
|
|
background_noise_(background_noise),
|
|
max_input_value_(0) {
|
|
assert(sample_rate_hz_ == 8000 ||
|
|
sample_rate_hz_ == 16000 ||
|
|
sample_rate_hz_ == 32000 ||
|
|
sample_rate_hz_ == 48000);
|
|
assert(num_channels_ > 0);
|
|
assert(static_cast<int>(master_channel_) < num_channels_);
|
|
memset(auto_correlation_, 0, sizeof(auto_correlation_));
|
|
}
|
|
|
|
virtual ~TimeStretch() {}
|
|
|
|
// This method performs the processing common to both Accelerate and
|
|
// PreemptiveExpand.
|
|
ReturnCodes Process(const int16_t* input,
|
|
size_t input_len,
|
|
bool fast_mode,
|
|
AudioMultiVector* output,
|
|
int16_t* length_change_samples);
|
|
|
|
protected:
|
|
// Sets the parameters |best_correlation| and |peak_index| to suitable
|
|
// values when the signal contains no active speech. This method must be
|
|
// implemented by the sub-classes.
|
|
virtual void SetParametersForPassiveSpeech(size_t input_length,
|
|
int16_t* best_correlation,
|
|
int* peak_index) const = 0;
|
|
|
|
// Checks the criteria for performing the time-stretching operation and,
|
|
// if possible, performs the time-stretching. This method must be implemented
|
|
// by the sub-classes.
|
|
virtual ReturnCodes CheckCriteriaAndStretch(
|
|
const int16_t* input,
|
|
size_t input_length,
|
|
size_t peak_index,
|
|
int16_t best_correlation,
|
|
bool active_speech,
|
|
bool fast_mode,
|
|
AudioMultiVector* output) const = 0;
|
|
|
|
static const int kCorrelationLen = 50;
|
|
static const int kLogCorrelationLen = 6; // >= log2(kCorrelationLen).
|
|
static const int kMinLag = 10;
|
|
static const int kMaxLag = 60;
|
|
static const int kDownsampledLen = kCorrelationLen + kMaxLag;
|
|
static const int kCorrelationThreshold = 14746; // 0.9 in Q14.
|
|
|
|
const int sample_rate_hz_;
|
|
const int fs_mult_; // Sample rate multiplier = sample_rate_hz_ / 8000.
|
|
const int num_channels_;
|
|
const size_t master_channel_;
|
|
const BackgroundNoise& background_noise_;
|
|
int16_t max_input_value_;
|
|
int16_t downsampled_input_[kDownsampledLen];
|
|
// Adding 1 to the size of |auto_correlation_| because of how it is used
|
|
// by the peak-detection algorithm.
|
|
int16_t auto_correlation_[kCorrelationLen + 1];
|
|
|
|
private:
|
|
// Calculates the auto-correlation of |downsampled_input_| and writes the
|
|
// result to |auto_correlation_|.
|
|
void AutoCorrelation();
|
|
|
|
// Performs a simple voice-activity detection based on the input parameters.
|
|
bool SpeechDetection(int32_t vec1_energy, int32_t vec2_energy,
|
|
int peak_index, int scaling) const;
|
|
|
|
DISALLOW_COPY_AND_ASSIGN(TimeStretch);
|
|
};
|
|
|
|
} // namespace webrtc
|
|
#endif // WEBRTC_MODULES_AUDIO_CODING_NETEQ_TIME_STRETCH_H_
|