mirror of
https://gitlab.freedesktop.org/pipewire/pipewire.git
synced 2025-11-02 09:01:50 -05:00
audioconvert: compile c version separately
Also compile the c versions in a separate module with their own flags.
This commit is contained in:
parent
fe6ebd0e8a
commit
d260cb19be
14 changed files with 1471 additions and 1255 deletions
449
spa/plugins/audioconvert/channelmix-ops-c.c
Normal file
449
spa/plugins/audioconvert/channelmix-ops-c.c
Normal file
|
|
@ -0,0 +1,449 @@
|
|||
/* Spa
|
||||
*
|
||||
* Copyright © 2018 Wim Taymans
|
||||
*
|
||||
* Permission is hereby granted, free of charge, to any person obtaining a
|
||||
* copy of this software and associated documentation files (the "Software"),
|
||||
* to deal in the Software without restriction, including without limitation
|
||||
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
|
||||
* and/or sell copies of the Software, and to permit persons to whom the
|
||||
* Software is furnished to do so, subject to the following conditions:
|
||||
*
|
||||
* The above copyright notice and this permission notice (including the next
|
||||
* paragraph) shall be included in all copies or substantial portions of the
|
||||
* Software.
|
||||
*
|
||||
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
||||
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
||||
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
|
||||
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
||||
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
||||
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
|
||||
* DEALINGS IN THE SOFTWARE.
|
||||
*/
|
||||
|
||||
#include "channelmix-ops.h"
|
||||
|
||||
void
|
||||
channelmix_copy_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, n;
|
||||
float **d = (float **)dst;
|
||||
const float **s = (const float **)src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
memset(d[i], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
spa_memcpy(d[i], s[i], n_samples * sizeof(float));
|
||||
}
|
||||
else {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
for (n = 0; n < n_samples; n++)
|
||||
d[i][n] = s[i][n] * v;
|
||||
}
|
||||
}
|
||||
|
||||
#define _M(ch) (1UL << SPA_AUDIO_CHANNEL_ ## ch)
|
||||
|
||||
void
|
||||
channelmix_f32_n_m_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, j, n;
|
||||
float **d = (float **) dst;
|
||||
const float **s = (const float **) src;
|
||||
const float *m = matrix;
|
||||
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
for (i = 0; i < n_dst; i++) {
|
||||
float sum = 0.0f;
|
||||
for (j = 0; j < n_src; j++)
|
||||
sum += s[j][n] * m[i * n_src + j];
|
||||
d[i][n] = sum * v;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
#define MASK_MONO _M(FC)|_M(MONO)|_M(UNKNOWN)
|
||||
#define MASK_STEREO _M(FL)|_M(FR)|_M(UNKNOWN)
|
||||
|
||||
void
|
||||
channelmix_f32_1_2_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int n;
|
||||
float **d = (float **)dst;
|
||||
const float **s = (const float **)src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
memset(d[0], 0, n_samples * sizeof(float));
|
||||
memset(d[1], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (n = 0; n < n_samples; n++)
|
||||
d[0][n] = d[1][n] = s[0][n];
|
||||
}
|
||||
else {
|
||||
for (n = 0; n < n_samples; n++)
|
||||
d[0][n] = d[1][n] = s[0][n] * v;
|
||||
}
|
||||
}
|
||||
|
||||
void
|
||||
channelmix_f32_2_1_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int n;
|
||||
float **d = (float **)dst;
|
||||
const float **s = (const float **)src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
memset(d[0], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else {
|
||||
const float f = v * 0.5f;
|
||||
for (n = 0; n < n_samples; n++)
|
||||
d[0][n] = (s[0][n] + s[1][n]) * f;
|
||||
}
|
||||
}
|
||||
|
||||
void
|
||||
channelmix_f32_4_1_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int n;
|
||||
float **d = (float **)dst;
|
||||
const float **s = (const float **)src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
memset(d[0], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else {
|
||||
const float f = v * 0.25f;
|
||||
for (n = 0; n < n_samples; n++)
|
||||
d[0][n] = (s[0][n] + s[1][n] + s[2][n] + s[3][n]) * f;
|
||||
}
|
||||
}
|
||||
|
||||
void
|
||||
channelmix_f32_3p1_1_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int n;
|
||||
float **d = (float **)dst;
|
||||
const float **s = (const float **)src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
memset(d[0], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else {
|
||||
const float f = v * 0.5f;
|
||||
for (n = 0; n < n_samples; n++)
|
||||
d[0][n] = (s[0][n] + s[1][n] + s[2][n]) * f;
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
#define MASK_QUAD _M(FL)|_M(FR)|_M(RL)|_M(RR)|_M(UNKNOWN)
|
||||
|
||||
void
|
||||
channelmix_f32_2_4_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, n;
|
||||
float **d = (float **)dst;
|
||||
const float **s = (const float **)src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
memset(d[i], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
d[0][n] = d[2][n] = s[0][n];
|
||||
d[1][n] = d[3][n] = s[1][n];
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
d[0][n] = d[2][n] = s[0][n] * v;
|
||||
d[1][n] = d[3][n] = s[1][n] * v;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
#define MASK_3_1 _M(FL)|_M(FR)|_M(FC)|_M(LFE)
|
||||
void
|
||||
channelmix_f32_2_3p1_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, n;
|
||||
float **d = (float **)dst;
|
||||
const float **s = (const float **)src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
memset(d[i], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
d[0][n] = s[0][n];
|
||||
d[1][n] = s[1][n];
|
||||
d[2][n] = (s[0][n] + s[1][n]) * 0.5f;
|
||||
d[3][n] = 0.0f;
|
||||
}
|
||||
}
|
||||
else {
|
||||
const float f = 0.5f * v;
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
d[0][n] = s[0][n] * v;
|
||||
d[1][n] = s[1][n] * v;
|
||||
d[2][n] = (s[0][n] + s[1][n]) * f;
|
||||
d[3][n] = 0.0f;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
#define MASK_5_1 _M(FL)|_M(FR)|_M(FC)|_M(LFE)|_M(SL)|_M(SR)|_M(RL)|_M(RR)
|
||||
void
|
||||
channelmix_f32_2_5p1_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, n;
|
||||
float **d = (float **)dst;
|
||||
const float **s = (const float **)src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
memset(d[i], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
d[0][n] = d[4][n] = s[0][n];
|
||||
d[1][n] = d[5][n] = s[1][n];
|
||||
d[2][n] = (s[0][n] + s[1][n]) * 0.5f;
|
||||
d[3][n] = 0.0f;
|
||||
}
|
||||
}
|
||||
else {
|
||||
const float f = 0.5f * v;
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
d[0][n] = d[4][n] = s[0][n] * v;
|
||||
d[1][n] = d[5][n] = s[1][n] * v;
|
||||
d[2][n] = (s[0][n] + s[1][n]) * f;
|
||||
d[3][n] = 0.0f;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* FL+FR+FC+LFE+SL+SR -> FL+FR */
|
||||
void
|
||||
channelmix_f32_5p1_2_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int n;
|
||||
float **d = (float **) dst;
|
||||
const float **s = (const float **) src;
|
||||
const float *m = matrix;
|
||||
const float clev = m[2];
|
||||
const float llev = m[3];
|
||||
const float slev = m[4];
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
memset(d[0], 0, n_samples * sizeof(float));
|
||||
memset(d[1], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
const float ctr = clev * s[2][n] + llev * s[3][n];
|
||||
d[0][n] = s[0][n] + ctr + (slev * s[4][n]);
|
||||
d[1][n] = s[1][n] + ctr + (slev * s[5][n]);
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
const float ctr = clev * s[2][n] + llev * s[3][n];
|
||||
d[0][n] = (s[0][n] + ctr + (slev * s[4][n])) * v;
|
||||
d[1][n] = (s[1][n] + ctr + (slev * s[5][n])) * v;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* FL+FR+FC+LFE+SL+SR -> FL+FR+FC+LFE*/
|
||||
void
|
||||
channelmix_f32_5p1_3p1_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, n;
|
||||
float **d = (float **) dst;
|
||||
const float **s = (const float **) src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
memset(d[i], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else {
|
||||
const float f1 = 0.5f * v;
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
d[0][n] = (s[0][n] + s[4][n]) * f1;
|
||||
d[1][n] = (s[1][n] + s[5][n]) * f1;
|
||||
d[2][n] = s[2][n] * v;
|
||||
d[3][n] = s[3][n] * v;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* FL+FR+FC+LFE+SL+SR -> FL+FR+RL+RR*/
|
||||
void
|
||||
channelmix_f32_5p1_4_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, n;
|
||||
float **d = (float **) dst;
|
||||
const float **s = (const float **) src;
|
||||
const float *m = matrix;
|
||||
const float clev = m[2];
|
||||
const float llev = m[3];
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
memset(d[i], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
const float ctr = s[2][n] * clev + s[3][n] * llev;
|
||||
d[0][n] = s[0][n] + ctr;
|
||||
d[1][n] = s[1][n] + ctr;
|
||||
d[2][n] = s[4][n];
|
||||
d[3][n] = s[5][n];
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
const float ctr = s[2][n] * clev + s[3][n] * llev;
|
||||
d[0][n] = (s[0][n] + ctr) * v;
|
||||
d[1][n] = (s[1][n] + ctr) * v;
|
||||
d[2][n] = s[4][n] * v;
|
||||
d[3][n] = s[5][n] * v;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
#define MASK_7_1 _M(FL)|_M(FR)|_M(FC)|_M(LFE)|_M(SL)|_M(SR)|_M(RL)|_M(RR)
|
||||
|
||||
/* FL+FR+FC+LFE+SL+SR+RL+RR -> FL+FR */
|
||||
void
|
||||
channelmix_f32_7p1_2_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int n;
|
||||
float **d = (float **) dst;
|
||||
const float **s = (const float **) src;
|
||||
const float *m = matrix;
|
||||
const float clev = m[2];
|
||||
const float llev = m[3];
|
||||
const float slev = m[4];
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
memset(d[0], 0, n_samples * sizeof(float));
|
||||
memset(d[1], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
const float ctr = clev * s[2][n] + llev * s[3][n];
|
||||
d[0][n] = s[0][n] + ctr + (slev * (s[4][n] + s[6][n]));
|
||||
d[1][n] = s[1][n] + ctr + (slev * (s[5][n] + s[7][n]));
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
const float ctr = clev * s[2][n] + llev * s[3][n];
|
||||
d[0][n] = (s[0][n] + ctr + (slev * (s[4][n] + s[6][n]))) * v;
|
||||
d[1][n] = (s[1][n] + ctr + (slev * (s[5][n] + s[6][n]))) * v;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* FL+FR+FC+LFE+SL+SR+RL+RR -> FL+FR+FC+LFE*/
|
||||
void
|
||||
channelmix_f32_7p1_3p1_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, n;
|
||||
float **d = (float **) dst;
|
||||
const float **s = (const float **) src;
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
memset(d[i], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else {
|
||||
const float f1 = 0.5 * v;
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
d[0][n] = s[0][n] + (s[4][n] + s[6][n]) * f1;
|
||||
d[1][n] = s[1][n] + (s[5][n] + s[7][n]) * f1;
|
||||
d[2][n] = s[2][n] * v;
|
||||
d[3][n] = s[3][n] * v;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* FL+FR+FC+LFE+SL+SR+RL+RR -> FL+FR+RL+RR*/
|
||||
void
|
||||
channelmix_f32_7p1_4_c(void *data, int n_dst, void * SPA_RESTRICT dst[n_dst],
|
||||
int n_src, const void * SPA_RESTRICT src[n_src],
|
||||
const void *matrix, float v, int n_samples)
|
||||
{
|
||||
int i, n;
|
||||
float **d = (float **) dst;
|
||||
const float **s = (const float **) src;
|
||||
const float *m = matrix;
|
||||
const float clev = m[2];
|
||||
const float llev = m[3];
|
||||
const float slev = m[4];
|
||||
|
||||
if (v <= VOLUME_MIN) {
|
||||
for (i = 0; i < n_dst; i++)
|
||||
memset(d[i], 0, n_samples * sizeof(float));
|
||||
}
|
||||
else if (v == VOLUME_NORM) {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
const float ctr = s[2][n] * clev + s[3][n] * llev;
|
||||
const float sl = s[4][n] * slev;
|
||||
const float sr = s[5][n] * slev;
|
||||
d[0][n] = s[0][n] + ctr + sl;
|
||||
d[1][n] = s[1][n] + ctr + sr;
|
||||
d[2][n] = s[6][n] + sl;
|
||||
d[3][n] = s[7][n] + sr;
|
||||
}
|
||||
}
|
||||
else {
|
||||
for (n = 0; n < n_samples; n++) {
|
||||
const float ctr = s[2][n] * clev + s[3][n] * llev;
|
||||
const float sl = s[4][n] * slev;
|
||||
const float sr = s[5][n] * slev;
|
||||
d[0][n] = (s[0][n] + ctr + sl) * v;
|
||||
d[1][n] = (s[1][n] + ctr + sr) * v;
|
||||
d[2][n] = (s[6][n] + sl) * v;
|
||||
d[3][n] = (s[7][n] + sr) * v;
|
||||
}
|
||||
}
|
||||
}
|
||||
Loading…
Add table
Add a link
Reference in a new issue