ashikk
/
kaliber
mirror of https://github.com/auygun/kaliber.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
kaliber/src/base/sinc_resampler.cc

523 lines
19 KiB
C++
Raw Blame History

// Copyright 2012 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
//
// Initial input buffer layout, dividing into regions r0_ to r4_ (note: r0_, r3_
// and r4_ will move after the first load):
//
// |----------------|-----------------------------------------|----------------|
//
// request_frames_
// <--------------------------------------------------------->
// r0_ (during first load)
//
// kKernelSize / 2 kKernelSize / 2 kKernelSize / 2 kKernelSize / 2
// <---------------> <---------------> <---------------> <--------------->
// r1_ r2_ r3_ r4_
//
// block_size_ == r4_ - r2_
// <--------------------------------------->
//
// request_frames_
// <------------------ ... ----------------->
// r0_ (during second load)
//
// On the second request r0_ slides to the right by kKernelSize / 2 and r3_, r4_
// and block_size_ are reinitialized via step (3) in the algorithm below.
//
// These new regions remain constant until a Flush() occurs. While complicated,
// this allows us to reduce jitter by always requesting the same amount from the
// provided callback.
//
// The algorithm:
//
// 1) Allocate input_buffer of size: request_frames_ + kKernelSize; this ensures
// there's enough room to read request_frames_ from the callback into region
// r0_ (which will move between the first and subsequent passes).
//
// 2) Let r1_, r2_ each represent half the kernel centered around r0_:
//
// r0_ = input_buffer_ + kKernelSize / 2
// r1_ = input_buffer_
// r2_ = r0_
//
// r0_ is always request_frames_ in size. r1_, r2_ are kKernelSize / 2 in
// size. r1_ must be zero initialized to avoid convolution with garbage (see
// step (5) for why).
//
// 3) Let r3_, r4_ each represent half the kernel right aligned with the end of
// r0_ and choose block_size_ as the distance in frames between r4_ and r2_:
//
// r3_ = r0_ + request_frames_ - kKernelSize
// r4_ = r0_ + request_frames_ - kKernelSize / 2
// block_size_ = r4_ - r2_ = request_frames_ - kKernelSize / 2
//
// 4) Consume request_frames_ frames into r0_.
//
// 5) Position kernel centered at start of r2_ and generate output frames until
// the kernel is centered at the start of r4_ or we've finished generating
// all the output frames.
//
// 6) Wrap left over data from the r3_ to r1_ and r4_ to r2_.
//
// 7) If we're on the second load, in order to avoid overwriting the frames we
// just wrapped from r4_ we need to slide r0_ to the right by the size of
// r4_, which is kKernelSize / 2:
//
// r0_ = r0_ + kKernelSize / 2 = input_buffer_ + kKernelSize
//
// r3_, r4_, and block_size_ then need to be reinitialized, so goto (3).
//
// 8) Else, if we're not on the second load, goto (4).
//
// Note: we're glossing over how the sub-sample handling works with
// |virtual_source_idx_|, etc.
#include "base/sinc_resampler.h"
#include <cmath>
#include <cstring>
#include <limits>
#include "base/log.h"
#if defined(_M_X64) || defined(__x86_64__) || defined(__i386__)
#include <immintrin.h>
// Including these headers directly should generally be avoided. Since
// Chrome is compiled with -msse3 (the minimal requirement), we include the
// headers directly to make the intrinsics available.
#include <avx2intrin.h>
#include <avxintrin.h>
#include <fmaintrin.h>
#elif defined(_M_ARM64) || defined(__aarch64__)
#include <arm_neon.h>
#endif
namespace base {
namespace {
constexpr double kPiDouble = 3.14159265358979323846;
constexpr float kPiFloat = 3.14159265358979323846f;
class ScopedSubnormalFloatDisabler {
public:
ScopedSubnormalFloatDisabler() {
#if defined(_M_X64) || defined(__x86_64__) || defined(__i386__)
// Turn on "subnormals are zero" and "flush to zero" CSR flags.
orig_state_ = _mm_getcsr();
_mm_setcsr(orig_state_ | 0x8040);
#endif
}
ScopedSubnormalFloatDisabler(const ScopedSubnormalFloatDisabler&) = delete;
~ScopedSubnormalFloatDisabler() {
#if defined(_M_X64) || defined(__x86_64__) || defined(__i386__)
_mm_setcsr(orig_state_);
#endif
}
ScopedSubnormalFloatDisabler& operator=(const ScopedSubnormalFloatDisabler&) =
delete;
#if defined(_M_X64) || defined(__x86_64__) || defined(__i386__)
private:
unsigned int orig_state_;
#endif
};
double SincScaleFactor(double io_ratio) {
// |sinc_scale_factor| is basically the normalized cutoff frequency of the
// low-pass filter.
double sinc_scale_factor = io_ratio > 1.0 ? 1.0 / io_ratio : 1.0;
// The sinc function is an idealized brick-wall filter, but since we're
// windowing it the transition from pass to stop does not happen right away.
// So we should adjust the low pass filter cutoff slightly downward to avoid
// some aliasing at the very high-end.
// TODO(crogers): this value is empirical and to be more exact should vary
// depending on kKernelSize.
sinc_scale_factor *= 0.9;
return sinc_scale_factor;
}
int CalculateChunkSize(int block_size_, double io_ratio) {
return block_size_ / io_ratio;
}
} // namespace
// If we know the minimum architecture at compile time, avoid CPU detection.
void SincResampler::InitializeCPUSpecificFeatures() {
#if defined(_M_ARM64) || defined(__aarch64__)
convolve_proc_ = Convolve_NEON;
#elif defined(_M_X64) || defined(__x86_64__) || defined(__i386__)
#if 0 // TODO
// Using AVX2 instead of SSE2 when AVX2/FMA3 supported.
if (cpu.has_avx2() && cpu.has_fma3())
convolve_proc_ = Convolve_AVX2;
else if (cpu.has_sse2())
convolve_proc_ = Convolve_SSE;
else
convolve_proc_ = Convolve_C;
#endif
convolve_proc_ = Convolve_SSE;
#else
// Unknown architecture.
convolve_proc_ = Convolve_C;
#endif
}
SincResampler::SincResampler(double io_sample_rate_ratio, int request_frames)
: io_sample_rate_ratio_(io_sample_rate_ratio),
request_frames_(request_frames),
input_buffer_size_(request_frames_ + kKernelSize),
// Create input buffers with a 32-byte alignment for SIMD optimizations.
kernel_storage_(static_cast<float*>(
base::AlignedAlloc<32>(sizeof(float) * kKernelStorageSize))),
kernel_pre_sinc_storage_(static_cast<float*>(
base::AlignedAlloc<32>(sizeof(float) * kKernelStorageSize))),
kernel_window_storage_(static_cast<float*>(
base::AlignedAlloc<32>(sizeof(float) * kKernelStorageSize))),
input_buffer_(static_cast<float*>(
base::AlignedAlloc<32>(sizeof(float) * input_buffer_size_))),
r1_(input_buffer_.get()),
r2_(input_buffer_.get() + kKernelSize / 2) {
CHECK(request_frames > kKernelSize * 3 / 2)
<< "request_frames must be greater than 1.5 kernels to allow sufficient "
"data for resampling";
// This means that after the first call to Flush we will have
// block_size_ > kKernelSize and r2_ < r3_.
InitializeCPUSpecificFeatures();
DCHECK(convolve_proc_);
CHECK(request_frames_ > 0);
Flush();
memset(kernel_storage_.get(), 0,
sizeof(*kernel_storage_.get()) * kKernelStorageSize);
memset(kernel_pre_sinc_storage_.get(), 0,
sizeof(*kernel_pre_sinc_storage_.get()) * kKernelStorageSize);
memset(kernel_window_storage_.get(), 0,
sizeof(*kernel_window_storage_.get()) * kKernelStorageSize);
InitializeKernel();
}
SincResampler::~SincResampler() = default;
void SincResampler::UpdateRegions(bool second_load) {
// Setup various region pointers in the buffer (see diagram above). If we're
// on the second load we need to slide r0_ to the right by kKernelSize / 2.
r0_ = input_buffer_.get() + (second_load ? kKernelSize : kKernelSize / 2);
r3_ = r0_ + request_frames_ - kKernelSize;
r4_ = r0_ + request_frames_ - kKernelSize / 2;
block_size_ = r4_ - r2_;
chunk_size_ = CalculateChunkSize(block_size_, io_sample_rate_ratio_);
// r1_ at the beginning of the buffer.
DCHECK(r1_ == input_buffer_.get());
// r1_ left of r2_, r4_ left of r3_ and size correct.
DCHECK(r2_ - r1_ == r4_ - r3_);
// r2_ left of r3.
DCHECK(r2_ < r3_);
}
void SincResampler::InitializeKernel() {
// Blackman window parameters.
static const double kAlpha = 0.16;
static const double kA0 = 0.5 * (1.0 - kAlpha);
static const double kA1 = 0.5;
static const double kA2 = 0.5 * kAlpha;
// Generates a set of windowed sinc() kernels.
// We generate a range of sub-sample offsets from 0.0 to 1.0.
const double sinc_scale_factor = SincScaleFactor(io_sample_rate_ratio_);
for (int offset_idx = 0; offset_idx <= kKernelOffsetCount; ++offset_idx) {
const float subsample_offset =
static_cast<float>(offset_idx) / kKernelOffsetCount;
for (int i = 0; i < kKernelSize; ++i) {
const int idx = i + offset_idx * kKernelSize;
const float pre_sinc =
base::kPiFloat * (i - kKernelSize / 2 - subsample_offset);
kernel_pre_sinc_storage_[idx] = pre_sinc;
// Compute Blackman window, matching the offset of the sinc().
const float x = (i - subsample_offset) / kKernelSize;
const float window =
static_cast<float>(kA0 - kA1 * cos(2.0 * base::kPiDouble * x) +
kA2 * cos(4.0 * base::kPiDouble * x));
kernel_window_storage_[idx] = window;
// Compute the sinc with offset, then window the sinc() function and store
// at the correct offset.
kernel_storage_[idx] = static_cast<float>(
window * (pre_sinc ? sin(sinc_scale_factor * pre_sinc) / pre_sinc
: sinc_scale_factor));
}
}
}
void SincResampler::SetRatio(double io_sample_rate_ratio) {
if (fabs(io_sample_rate_ratio_ - io_sample_rate_ratio) <
std::numeric_limits<double>::epsilon()) {
return;
}
io_sample_rate_ratio_ = io_sample_rate_ratio;
chunk_size_ = CalculateChunkSize(block_size_, io_sample_rate_ratio_);
// Optimize reinitialization by reusing values which are independent of
// |sinc_scale_factor|. Provides a 3x speedup.
const double sinc_scale_factor = SincScaleFactor(io_sample_rate_ratio_);
for (int offset_idx = 0; offset_idx <= kKernelOffsetCount; ++offset_idx) {
for (int i = 0; i < kKernelSize; ++i) {
const int idx = i + offset_idx * kKernelSize;
const float window = kernel_window_storage_[idx];
const float pre_sinc = kernel_pre_sinc_storage_[idx];
kernel_storage_[idx] = static_cast<float>(
window * (pre_sinc ? sin(sinc_scale_factor * pre_sinc) / pre_sinc
: sinc_scale_factor));
}
}
}
void SincResampler::Resample(int frames, float* destination, ReadCB read_cb) {
int remaining_frames = frames;
// Step (1) -- Prime the input buffer at the start of the input stream.
if (!buffer_primed_ && remaining_frames) {
read_cb(request_frames_, r0_);
buffer_primed_ = true;
}
// Step (2) -- Resample!
while (remaining_frames) {
// Silent audio can contain non-zero samples small enough to result in
// subnormals internally. Disabling subnormals can be significantly faster.
{
ScopedSubnormalFloatDisabler disable_subnormals;
while (virtual_source_idx_ < block_size_) {
// |virtual_source_idx_| lies in between two kernel offsets so figure
// out what they are.
const int source_idx = static_cast<int>(virtual_source_idx_);
const double virtual_offset_idx =
(virtual_source_idx_ - source_idx) * kKernelOffsetCount;
const int offset_idx = static_cast<int>(virtual_offset_idx);
// We'll compute "convolutions" for the two kernels which straddle
// |virtual_source_idx_|.
const float* k1 = kernel_storage_.get() + offset_idx * kKernelSize;
const float* k2 = k1 + kKernelSize;
// Ensure |k1|, |k2| are 32-byte aligned for SIMD usage. Should always
// be true so long as kKernelSize is a multiple of 32.
DCHECK(0u == reinterpret_cast<uintptr_t>(k1) & 0x1F);
DCHECK(0u == reinterpret_cast<uintptr_t>(k2) & 0x1F);
// Initialize input pointer based on quantized |virtual_source_idx_|.
const float* input_ptr = r1_ + source_idx;
// Figure out how much to weight each kernel's "convolution".
const double kernel_interpolation_factor =
virtual_offset_idx - offset_idx;
*destination++ =
convolve_proc_(input_ptr, k1, k2, kernel_interpolation_factor);
// Advance the virtual index.
virtual_source_idx_ += io_sample_rate_ratio_;
if (!--remaining_frames)
return;
}
}
// Wrap back around to the start.
DCHECK(virtual_source_idx_ > block_size_);
virtual_source_idx_ -= block_size_;
// Step (3) -- Copy r3_, r4_ to r1_, r2_.
// This wraps the last input frames back to the start of the buffer.
memcpy(r1_, r3_, sizeof(*input_buffer_.get()) * kKernelSize);
// Step (4) -- Reinitialize regions if necessary.
if (r0_ == r2_)
UpdateRegions(true);
// Step (5) -- Refresh the buffer with more input.
read_cb(request_frames_, r0_);
}
}
void SincResampler::PrimeWithSilence() {
// By enforcing the buffer hasn't been primed, we ensure the input buffer has
// already been zeroed during construction or by a previous Flush() call.
DCHECK(!buffer_primed_);
DCHECK(input_buffer_[0] == 0.0f);
UpdateRegions(true);
}
void SincResampler::Flush() {
virtual_source_idx_ = 0;
buffer_primed_ = false;
memset(input_buffer_.get(), 0,
sizeof(*input_buffer_.get()) * input_buffer_size_);
UpdateRegions(false);
}
int SincResampler::GetMaxInputFramesRequested(
int output_frames_requested) const {
const int num_chunks = static_cast<int>(
std::ceil(static_cast<float>(output_frames_requested) / chunk_size_));
return num_chunks * request_frames_;
}
double SincResampler::BufferedFrames() const {
return buffer_primed_ ? request_frames_ - virtual_source_idx_ : 0;
}
float SincResampler::Convolve_C(const float* input_ptr,
const float* k1,
const float* k2,
double kernel_interpolation_factor) {
float sum1 = 0;
float sum2 = 0;
// Generate a single output sample. Unrolling this loop hurt performance in
// local testing.
int n = kKernelSize;
while (n--) {
sum1 += *input_ptr * *k1++;
sum2 += *input_ptr++ * *k2++;
}
// Linearly interpolate the two "convolutions".
return static_cast<float>((1.0 - kernel_interpolation_factor) * sum1 +
kernel_interpolation_factor * sum2);
}
#if defined(_M_X64) || defined(__x86_64__) || defined(__i386__)
float SincResampler::Convolve_SSE(const float* input_ptr,
const float* k1,
const float* k2,
double kernel_interpolation_factor) {
__m128 m_input;
__m128 m_sums1 = _mm_setzero_ps();
__m128 m_sums2 = _mm_setzero_ps();
// Based on |input_ptr| alignment, we need to use loadu or load. Unrolling
// these loops hurt performance in local testing.
if (reinterpret_cast<uintptr_t>(input_ptr) & 0x0F) {
for (int i = 0; i < kKernelSize; i += 4) {
m_input = _mm_loadu_ps(input_ptr + i);
m_sums1 = _mm_add_ps(m_sums1, _mm_mul_ps(m_input, _mm_load_ps(k1 + i)));
m_sums2 = _mm_add_ps(m_sums2, _mm_mul_ps(m_input, _mm_load_ps(k2 + i)));
}
} else {
for (int i = 0; i < kKernelSize; i += 4) {
m_input = _mm_load_ps(input_ptr + i);
m_sums1 = _mm_add_ps(m_sums1, _mm_mul_ps(m_input, _mm_load_ps(k1 + i)));
m_sums2 = _mm_add_ps(m_sums2, _mm_mul_ps(m_input, _mm_load_ps(k2 + i)));
}
}
// Linearly interpolate the two "convolutions".
m_sums1 = _mm_mul_ps(
m_sums1,
_mm_set_ps1(static_cast<float>(1.0 - kernel_interpolation_factor)));
m_sums2 = _mm_mul_ps(
m_sums2, _mm_set_ps1(static_cast<float>(kernel_interpolation_factor)));
m_sums1 = _mm_add_ps(m_sums1, m_sums2);
// Sum components together.
float result;
m_sums2 = _mm_add_ps(_mm_movehl_ps(m_sums1, m_sums1), m_sums1);
_mm_store_ss(&result,
_mm_add_ss(m_sums2, _mm_shuffle_ps(m_sums2, m_sums2, 1)));
return result;
}
__attribute__((target("avx2,fma"))) float SincResampler::Convolve_AVX2(
const float* input_ptr,
const float* k1,
const float* k2,
double kernel_interpolation_factor) {
__m256 m_input;
__m256 m_sums1 = _mm256_setzero_ps();
__m256 m_sums2 = _mm256_setzero_ps();
// Based on |input_ptr| alignment, we need to use loadu or load. Unrolling
// these loops has not been tested or benchmarked.
bool aligned_input = (reinterpret_cast<uintptr_t>(input_ptr) & 0x1F) == 0;
if (!aligned_input) {
for (size_t i = 0; i < kKernelSize; i += 8) {
m_input = _mm256_loadu_ps(input_ptr + i);
m_sums1 = _mm256_fmadd_ps(m_input, _mm256_load_ps(k1 + i), m_sums1);
m_sums2 = _mm256_fmadd_ps(m_input, _mm256_load_ps(k2 + i), m_sums2);
}
} else {
for (size_t i = 0; i < kKernelSize; i += 8) {
m_input = _mm256_load_ps(input_ptr + i);
m_sums1 = _mm256_fmadd_ps(m_input, _mm256_load_ps(k1 + i), m_sums1);
m_sums2 = _mm256_fmadd_ps(m_input, _mm256_load_ps(k2 + i), m_sums2);
}
}
// Linearly interpolate the two "convolutions".
__m128 m128_sums1 = _mm_add_ps(_mm256_extractf128_ps(m_sums1, 0),
_mm256_extractf128_ps(m_sums1, 1));
__m128 m128_sums2 = _mm_add_ps(_mm256_extractf128_ps(m_sums2, 0),
_mm256_extractf128_ps(m_sums2, 1));
m128_sums1 = _mm_mul_ps(
m128_sums1,
_mm_set_ps1(static_cast<float>(1.0 - kernel_interpolation_factor)));
m128_sums2 = _mm_mul_ps(
m128_sums2, _mm_set_ps1(static_cast<float>(kernel_interpolation_factor)));
m128_sums1 = _mm_add_ps(m128_sums1, m128_sums2);
// Sum components together.
float result;
m128_sums2 = _mm_add_ps(_mm_movehl_ps(m128_sums1, m128_sums1), m128_sums1);
_mm_store_ss(&result, _mm_add_ss(m128_sums2,
_mm_shuffle_ps(m128_sums2, m128_sums2, 1)));
return result;
}
#elif defined(_M_ARM64) || defined(__aarch64__)
float SincResampler::Convolve_NEON(const float* input_ptr,
const float* k1,
const float* k2,
double kernel_interpolation_factor) {
float32x4_t m_input;
float32x4_t m_sums1 = vmovq_n_f32(0);
float32x4_t m_sums2 = vmovq_n_f32(0);
const float* upper = input_ptr + kKernelSize;
for (; input_ptr < upper;) {
m_input = vld1q_f32(input_ptr);
input_ptr += 4;
m_sums1 = vmlaq_f32(m_sums1, m_input, vld1q_f32(k1));
k1 += 4;
m_sums2 = vmlaq_f32(m_sums2, m_input, vld1q_f32(k2));
k2 += 4;
}
// Linearly interpolate the two "convolutions".
m_sums1 = vmlaq_f32(
vmulq_f32(m_sums1, vmovq_n_f32(1.0 - kernel_interpolation_factor)),
m_sums2, vmovq_n_f32(kernel_interpolation_factor));
// Sum components together.
float32x2_t m_half = vadd_f32(vget_high_f32(m_sums1), vget_low_f32(m_sums1));
return vget_lane_f32(vpadd_f32(m_half, m_half), 0);
}
#endif
} // namespace base
Reference in New Issue View Git Blame Copy Permalink