mirror of
https://github.com/mollyim/webrtc.git
synced 2025-05-12 21:30:45 +01:00

Sample rates not divisible by 100, in particular 11025 Hz and 22050 Hz, have long been used with APM in Chrome, but the support has never been stated explicitly. This CL makes minor modifications to the APM API to clarify how rates are handled when 10 ms is not an integer number of samples. Unit tests are also extended to cover this case better. This does not update all references to 10 ms and implicit floor(sample_rate/100) computations, but it does at least take us closer to a correct API. Note that not all code needs to support these sample rates. For example, audio processing submodules only need to operate on the native APM rates 16000, 32000, 48000. Bug: chromium:1332484 Change-Id: I1dad15468f6ccb9c0d4d09c5819fe87f8388d5b8 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/268769 Reviewed-by: Henrik Andreassson <henrika@webrtc.org> Commit-Queue: Sam Zackrisson <saza@webrtc.org> Reviewed-by: Ivo Creusen <ivoc@webrtc.org> Cr-Commit-Position: refs/heads/main@{#37682}
276 lines
10 KiB
C++
276 lines
10 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.
|
|
*/
|
|
|
|
#include "audio/remix_resample.h"
|
|
|
|
#include <cmath>
|
|
|
|
#include "common_audio/resampler/include/push_resampler.h"
|
|
#include "rtc_base/arraysize.h"
|
|
#include "rtc_base/checks.h"
|
|
#include "test/gtest.h"
|
|
|
|
namespace webrtc {
|
|
namespace voe {
|
|
namespace {
|
|
|
|
int GetFrameSize(int sample_rate_hz) {
|
|
return sample_rate_hz / 100;
|
|
}
|
|
|
|
class UtilityTest : public ::testing::Test {
|
|
protected:
|
|
UtilityTest() {
|
|
src_frame_.sample_rate_hz_ = 16000;
|
|
src_frame_.samples_per_channel_ = src_frame_.sample_rate_hz_ / 100;
|
|
src_frame_.num_channels_ = 1;
|
|
dst_frame_.CopyFrom(src_frame_);
|
|
golden_frame_.CopyFrom(src_frame_);
|
|
}
|
|
|
|
void RunResampleTest(int src_channels,
|
|
int src_sample_rate_hz,
|
|
int dst_channels,
|
|
int dst_sample_rate_hz);
|
|
|
|
PushResampler<int16_t> resampler_;
|
|
AudioFrame src_frame_;
|
|
AudioFrame dst_frame_;
|
|
AudioFrame golden_frame_;
|
|
};
|
|
|
|
// Sets the signal value to increase by `data` with every sample. Floats are
|
|
// used so non-integer values result in rounding error, but not an accumulating
|
|
// error.
|
|
void SetMonoFrame(float data, int sample_rate_hz, AudioFrame* frame) {
|
|
frame->Mute();
|
|
frame->num_channels_ = 1;
|
|
frame->sample_rate_hz_ = sample_rate_hz;
|
|
frame->samples_per_channel_ = GetFrameSize(sample_rate_hz);
|
|
int16_t* frame_data = frame->mutable_data();
|
|
for (size_t i = 0; i < frame->samples_per_channel_; i++) {
|
|
frame_data[i] = static_cast<int16_t>(data * i);
|
|
}
|
|
}
|
|
|
|
// Keep the existing sample rate.
|
|
void SetMonoFrame(float data, AudioFrame* frame) {
|
|
SetMonoFrame(data, frame->sample_rate_hz_, frame);
|
|
}
|
|
|
|
// Sets the signal value to increase by `left` and `right` with every sample in
|
|
// each channel respectively.
|
|
void SetStereoFrame(float left,
|
|
float right,
|
|
int sample_rate_hz,
|
|
AudioFrame* frame) {
|
|
frame->Mute();
|
|
frame->num_channels_ = 2;
|
|
frame->sample_rate_hz_ = sample_rate_hz;
|
|
frame->samples_per_channel_ = GetFrameSize(sample_rate_hz);
|
|
int16_t* frame_data = frame->mutable_data();
|
|
for (size_t i = 0; i < frame->samples_per_channel_; i++) {
|
|
frame_data[i * 2] = static_cast<int16_t>(left * i);
|
|
frame_data[i * 2 + 1] = static_cast<int16_t>(right * i);
|
|
}
|
|
}
|
|
|
|
// Keep the existing sample rate.
|
|
void SetStereoFrame(float left, float right, AudioFrame* frame) {
|
|
SetStereoFrame(left, right, frame->sample_rate_hz_, frame);
|
|
}
|
|
|
|
// Sets the signal value to increase by `ch1`, `ch2`, `ch3`, `ch4` with every
|
|
// sample in each channel respectively.
|
|
void SetQuadFrame(float ch1,
|
|
float ch2,
|
|
float ch3,
|
|
float ch4,
|
|
int sample_rate_hz,
|
|
AudioFrame* frame) {
|
|
frame->Mute();
|
|
frame->num_channels_ = 4;
|
|
frame->sample_rate_hz_ = sample_rate_hz;
|
|
frame->samples_per_channel_ = GetFrameSize(sample_rate_hz);
|
|
int16_t* frame_data = frame->mutable_data();
|
|
for (size_t i = 0; i < frame->samples_per_channel_; i++) {
|
|
frame_data[i * 4] = static_cast<int16_t>(ch1 * i);
|
|
frame_data[i * 4 + 1] = static_cast<int16_t>(ch2 * i);
|
|
frame_data[i * 4 + 2] = static_cast<int16_t>(ch3 * i);
|
|
frame_data[i * 4 + 3] = static_cast<int16_t>(ch4 * i);
|
|
}
|
|
}
|
|
|
|
void VerifyParams(const AudioFrame& ref_frame, const AudioFrame& test_frame) {
|
|
EXPECT_EQ(ref_frame.num_channels_, test_frame.num_channels_);
|
|
EXPECT_EQ(ref_frame.samples_per_channel_, test_frame.samples_per_channel_);
|
|
EXPECT_EQ(ref_frame.sample_rate_hz_, test_frame.sample_rate_hz_);
|
|
}
|
|
|
|
// Computes the best SNR based on the error between `ref_frame` and
|
|
// `test_frame`. It allows for up to a `max_delay` in samples between the
|
|
// signals to compensate for the resampling delay.
|
|
float ComputeSNR(const AudioFrame& ref_frame,
|
|
const AudioFrame& test_frame,
|
|
size_t max_delay) {
|
|
VerifyParams(ref_frame, test_frame);
|
|
float best_snr = 0;
|
|
size_t best_delay = 0;
|
|
for (size_t delay = 0; delay <= max_delay; delay++) {
|
|
float mse = 0;
|
|
float variance = 0;
|
|
const int16_t* ref_frame_data = ref_frame.data();
|
|
const int16_t* test_frame_data = test_frame.data();
|
|
for (size_t i = 0;
|
|
i < ref_frame.samples_per_channel_ * ref_frame.num_channels_ - delay;
|
|
i++) {
|
|
int error = ref_frame_data[i] - test_frame_data[i + delay];
|
|
mse += error * error;
|
|
variance += ref_frame_data[i] * ref_frame_data[i];
|
|
}
|
|
float snr = 100; // We assign 100 dB to the zero-error case.
|
|
if (mse > 0)
|
|
snr = 10 * std::log10(variance / mse);
|
|
if (snr > best_snr) {
|
|
best_snr = snr;
|
|
best_delay = delay;
|
|
}
|
|
}
|
|
printf("SNR=%.1f dB at delay=%zu\n", best_snr, best_delay);
|
|
return best_snr;
|
|
}
|
|
|
|
void VerifyFramesAreEqual(const AudioFrame& ref_frame,
|
|
const AudioFrame& test_frame) {
|
|
VerifyParams(ref_frame, test_frame);
|
|
const int16_t* ref_frame_data = ref_frame.data();
|
|
const int16_t* test_frame_data = test_frame.data();
|
|
for (size_t i = 0;
|
|
i < ref_frame.samples_per_channel_ * ref_frame.num_channels_; i++) {
|
|
EXPECT_EQ(ref_frame_data[i], test_frame_data[i]);
|
|
}
|
|
}
|
|
|
|
void UtilityTest::RunResampleTest(int src_channels,
|
|
int src_sample_rate_hz,
|
|
int dst_channels,
|
|
int dst_sample_rate_hz) {
|
|
PushResampler<int16_t> resampler; // Create a new one with every test.
|
|
const int16_t kSrcCh1 = 30; // Shouldn't overflow for any used sample rate.
|
|
const int16_t kSrcCh2 = 15;
|
|
const int16_t kSrcCh3 = 22;
|
|
const int16_t kSrcCh4 = 8;
|
|
const float resampling_factor =
|
|
(1.0 * src_sample_rate_hz) / dst_sample_rate_hz;
|
|
const float dst_ch1 = resampling_factor * kSrcCh1;
|
|
const float dst_ch2 = resampling_factor * kSrcCh2;
|
|
const float dst_ch3 = resampling_factor * kSrcCh3;
|
|
const float dst_ch4 = resampling_factor * kSrcCh4;
|
|
const float dst_stereo_to_mono = (dst_ch1 + dst_ch2) / 2;
|
|
const float dst_quad_to_mono = (dst_ch1 + dst_ch2 + dst_ch3 + dst_ch4) / 4;
|
|
const float dst_quad_to_stereo_ch1 = (dst_ch1 + dst_ch2) / 2;
|
|
const float dst_quad_to_stereo_ch2 = (dst_ch3 + dst_ch4) / 2;
|
|
if (src_channels == 1)
|
|
SetMonoFrame(kSrcCh1, src_sample_rate_hz, &src_frame_);
|
|
else if (src_channels == 2)
|
|
SetStereoFrame(kSrcCh1, kSrcCh2, src_sample_rate_hz, &src_frame_);
|
|
else
|
|
SetQuadFrame(kSrcCh1, kSrcCh2, kSrcCh3, kSrcCh4, src_sample_rate_hz,
|
|
&src_frame_);
|
|
|
|
if (dst_channels == 1) {
|
|
SetMonoFrame(0, dst_sample_rate_hz, &dst_frame_);
|
|
if (src_channels == 1)
|
|
SetMonoFrame(dst_ch1, dst_sample_rate_hz, &golden_frame_);
|
|
else if (src_channels == 2)
|
|
SetMonoFrame(dst_stereo_to_mono, dst_sample_rate_hz, &golden_frame_);
|
|
else
|
|
SetMonoFrame(dst_quad_to_mono, dst_sample_rate_hz, &golden_frame_);
|
|
} else {
|
|
SetStereoFrame(0, 0, dst_sample_rate_hz, &dst_frame_);
|
|
if (src_channels == 1)
|
|
SetStereoFrame(dst_ch1, dst_ch1, dst_sample_rate_hz, &golden_frame_);
|
|
else if (src_channels == 2)
|
|
SetStereoFrame(dst_ch1, dst_ch2, dst_sample_rate_hz, &golden_frame_);
|
|
else
|
|
SetStereoFrame(dst_quad_to_stereo_ch1, dst_quad_to_stereo_ch2,
|
|
dst_sample_rate_hz, &golden_frame_);
|
|
}
|
|
|
|
// The sinc resampler has a known delay, which we compute here. Multiplying by
|
|
// two gives us a crude maximum for any resampling, as the old resampler
|
|
// typically (but not always) has lower delay.
|
|
static const size_t kInputKernelDelaySamples = 16;
|
|
const size_t max_delay = static_cast<size_t>(
|
|
static_cast<double>(dst_sample_rate_hz) / src_sample_rate_hz *
|
|
kInputKernelDelaySamples * dst_channels * 2);
|
|
printf("(%d, %d Hz) -> (%d, %d Hz) ", // SNR reported on the same line later.
|
|
src_channels, src_sample_rate_hz, dst_channels, dst_sample_rate_hz);
|
|
RemixAndResample(src_frame_, &resampler, &dst_frame_);
|
|
|
|
if (src_sample_rate_hz == 96000 && dst_sample_rate_hz <= 11025) {
|
|
// The sinc resampler gives poor SNR at this extreme conversion, but we
|
|
// expect to see this rarely in practice.
|
|
EXPECT_GT(ComputeSNR(golden_frame_, dst_frame_, max_delay), 14.0f);
|
|
} else {
|
|
EXPECT_GT(ComputeSNR(golden_frame_, dst_frame_, max_delay), 46.0f);
|
|
}
|
|
}
|
|
|
|
TEST_F(UtilityTest, RemixAndResampleCopyFrameSucceeds) {
|
|
// Stereo -> stereo.
|
|
SetStereoFrame(10, 10, &src_frame_);
|
|
SetStereoFrame(0, 0, &dst_frame_);
|
|
RemixAndResample(src_frame_, &resampler_, &dst_frame_);
|
|
VerifyFramesAreEqual(src_frame_, dst_frame_);
|
|
|
|
// Mono -> mono.
|
|
SetMonoFrame(20, &src_frame_);
|
|
SetMonoFrame(0, &dst_frame_);
|
|
RemixAndResample(src_frame_, &resampler_, &dst_frame_);
|
|
VerifyFramesAreEqual(src_frame_, dst_frame_);
|
|
}
|
|
|
|
TEST_F(UtilityTest, RemixAndResampleMixingOnlySucceeds) {
|
|
// Stereo -> mono.
|
|
SetStereoFrame(0, 0, &dst_frame_);
|
|
SetMonoFrame(10, &src_frame_);
|
|
SetStereoFrame(10, 10, &golden_frame_);
|
|
RemixAndResample(src_frame_, &resampler_, &dst_frame_);
|
|
VerifyFramesAreEqual(dst_frame_, golden_frame_);
|
|
|
|
// Mono -> stereo.
|
|
SetMonoFrame(0, &dst_frame_);
|
|
SetStereoFrame(10, 20, &src_frame_);
|
|
SetMonoFrame(15, &golden_frame_);
|
|
RemixAndResample(src_frame_, &resampler_, &dst_frame_);
|
|
VerifyFramesAreEqual(golden_frame_, dst_frame_);
|
|
}
|
|
|
|
TEST_F(UtilityTest, RemixAndResampleSucceeds) {
|
|
const int kSampleRates[] = {8000, 11025, 16000, 22050,
|
|
32000, 44100, 48000, 96000};
|
|
const int kSrcChannels[] = {1, 2, 4};
|
|
const int kDstChannels[] = {1, 2};
|
|
|
|
for (int src_rate : kSampleRates) {
|
|
for (int dst_rate : kSampleRates) {
|
|
for (size_t src_channels : kSrcChannels) {
|
|
for (size_t dst_channels : kDstChannels) {
|
|
RunResampleTest(src_channels, src_rate, dst_channels, dst_rate);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
} // namespace
|
|
} // namespace voe
|
|
} // namespace webrtc
|