You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
OpenAudio_ArduinoLibrary/output_i2s2_f32.cpp

465 lines
20 KiB

/*
* ***** output_i2s_f32.cpp *****
*
* Audio Library for Teensy 3.X
* Copyright (c) 2014, Paul Stoffregen, paul@pjrc.com
*
* Development of this audio library was funded by PJRC.COM, LLC by sales of
* Teensy and Audio Adaptor boards. Please support PJRC's efforts to develop
* open source software by purchasing Teensy or other PJRC products.
*
* 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, development funding notice, and this permission
* notice 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.
*/
/*
* Extended by Chip Audette, OpenAudio, May 2019
* Converted to F32 and to variable audio block length
* The F32 conversion is under the MIT License. Use at your own risk.
*/
// Updated OpenAudio F32 with this version from Chip Audette's Tympan Library Jan 2021 RSL
// Removed old commented out code. RSL 30 May 2022
#include "output_i2s2_f32.h"
#include <arm_math.h>
#include <Audio.h> //to get access to Audio/utlity/imxrt_hw.h...do we really need this??? WEA 2020-10-31
float AudioOutputI2S2_F32::setI2SFreq_T3(const float freq_Hz) {
return 0.0f;
}
audio_block_f32_t * AudioOutputI2S2_F32::block_left_1st = NULL;
audio_block_f32_t * AudioOutputI2S2_F32::block_right_1st = NULL;
audio_block_f32_t * AudioOutputI2S2_F32::block_left_2nd = NULL;
audio_block_f32_t * AudioOutputI2S2_F32::block_right_2nd = NULL;
uint16_t AudioOutputI2S2_F32::block_left_offset = 0;
uint16_t AudioOutputI2S2_F32::block_right_offset = 0;
bool AudioOutputI2S2_F32::update_responsibility = false;
DMAChannel AudioOutputI2S2_F32::dma(false);
DMAMEM __attribute__((aligned(32))) static uint64_t i2s2_tx_buffer[AUDIO_BLOCK_SAMPLES];
//DMAMEM static int32_t i2s2_tx_buffer[2*AUDIO_BLOCK_SAMPLES]; //2 channels at 32-bits per sample. Local "audio_block_samples" should be no larger than global "AUDIO_BLOCK_SAMPLES"
float AudioOutputI2S2_F32::sample_rate_Hz = AUDIO_SAMPLE_RATE;
int AudioOutputI2S2_F32::audio_block_samples = AUDIO_BLOCK_SAMPLES;
#if defined(__IMXRT1062__)
#include <utility/imxrt_hw.h> //from Teensy Audio library. For set_audioClock()
#endif
//#for 16-bit transfers
#define I2S_BUFFER_TO_USE_BYTES (AudioOutputI2S2_F32::audio_block_samples*sizeof(i2s2_tx_buffer[0]))
//#for 32-bit transfers
//#define I2S_BUFFER_TO_USE_BYTES (AudioOutputI2S2_F32::audio_block_samples*2*sizeof(i2s2_tx_buffer[0]))
void AudioOutputI2S2_F32::begin(void)
{
bool transferUsing32bit = false;
begin(transferUsing32bit);
}
void AudioOutputI2S2_F32::begin(bool transferUsing32bit) {
dma.begin(true); // Allocate the DMA channel first
block_left_1st = NULL;
block_right_1st = NULL;
AudioOutputI2S2_F32::config_i2s(transferUsing32bit, sample_rate_Hz);
// if AudioInputI2S2 set I2S_TCSR_TE (for clock sync), disable it
I2S2_TCSR = 0;
while (I2S2_TCSR & I2S_TCSR_TE) ; //wait for transmit disabled
CORE_PIN2_CONFIG = 2; //EMC_04, 2=SAI2_TX_DATA, page 428
dma.TCD->SADDR = i2s2_tx_buffer;
dma.TCD->SOFF = 4;
dma.TCD->ATTR = DMA_TCD_ATTR_SSIZE(2) | DMA_TCD_ATTR_DSIZE(2);
dma.TCD->NBYTES_MLNO = 4;
//dma.TCD->SLAST = -sizeof(i2s2_tx_buffer);//orig from Teensy Audio Library 2020-10-31
dma.TCD->SLAST = -I2S_BUFFER_TO_USE_BYTES;
dma.TCD->DOFF = 0;
//dma.TCD->CITER_ELINKNO = sizeof(i2s2_tx_buffer) / 2; //orig from Teensy Audio Library 2020-10-31
dma.TCD->CITER_ELINKNO = I2S_BUFFER_TO_USE_BYTES / 4;
dma.TCD->DLASTSGA = 0;
//dma.TCD->BITER_ELINKNO = sizeof(i2s2_tx_buffer) / 2;//orig from Teensy Audio Library 2020-10-31
dma.TCD->BITER_ELINKNO = I2S_BUFFER_TO_USE_BYTES / 4;
dma.TCD->CSR = DMA_TCD_CSR_INTHALF | DMA_TCD_CSR_INTMAJOR;
dma.TCD->DADDR = (void *)((uint32_t)&I2S2_TDR0 + 0);
dma.triggerAtHardwareEvent(DMAMUX_SOURCE_SAI2_TX);
dma.enable(); //newer location of this line in Teensy Audio library
// I2S1_RCSR |= I2S_RCSR_RE | I2S_RCSR_BCE; // from output_i2s_f32.cpp
// I2S1_TCSR = I2S_TCSR_TE | I2S_TCSR_BCE | I2S_TCSR_FRDE; // from output_i2s_f32.cpp
I2S2_TCSR |= I2S_TCSR_TE | I2S_TCSR_BCE | I2S_TCSR_FRDE | I2S_TCSR_FR;
update_responsibility = update_setup();
dma.attachInterrupt(AudioOutputI2S2_F32::isr);
//dma.enable(); //original location of this line in older Tympan_Library
enabled = 1;
//AudioInputI2S_F32::begin_guts();
}
void AudioOutputI2S2_F32::isr(void)
{
int32_t *dest;
audio_block_f32_t *blockL, *blockR;
uint32_t saddr, offsetL, offsetR;
saddr = (uint32_t)(dma.TCD->SADDR);
dma.clearInterrupt();
//if (saddr < (uint32_t)i2s2_tx_buffer + sizeof(i2s2_tx_buffer) / 2) { //original 16-bit
if (saddr < (uint32_t)i2s2_tx_buffer + I2S_BUFFER_TO_USE_BYTES / 2) { //are we transmitting the first half or second half of the buffer?
// DMA is transmitting the first half of the buffer
// so we must fill the second half
//dest = (int16_t *)&i2s2_tx_buffer[AUDIO_BLOCK_SAMPLES/2]; //original Teensy Audio
dest = (int32_t *)&i2s2_tx_buffer[audio_block_samples/2]; //this will be diff if we were to do 32-bit samples
if (AudioOutputI2S2_F32::update_responsibility) AudioStream_F32::update_all();
} else {
// DMA is transmitting the second half of the buffer
// so we must fill the first half
dest = (int32_t *)i2s2_tx_buffer;
}
blockL = AudioOutputI2S2_F32::block_left_1st;
blockR = AudioOutputI2S2_F32::block_right_1st;
offsetL = AudioOutputI2S2_F32::block_left_offset;
offsetR = AudioOutputI2S2_F32::block_right_offset;
int32_t *d = dest;
if (blockL && blockR) {
//memcpy_tointerleaveLR(dest, blockL->data + offsetL, blockR->data + offsetR);
//memcpy_tointerleaveLRwLen(dest, blockL->data + offsetL, blockR->data + offsetR, audio_block_samples/2);
float32_t *pL = blockL->data + offsetL;
float32_t *pR = blockR->data + offsetR;
for (int i=0; i < audio_block_samples/2; i++) {
*d++ = (int32_t) *pL++;
*d++ = (int32_t) *pR++; //interleave
//*d++ = 0;
//*d++ = 0;
}
offsetL += audio_block_samples / 2;
offsetR += audio_block_samples / 2;
} else if (blockL) {
//memcpy_tointerleaveLR(dest, blockL->data + offsetL, blockR->data + offsetR);
float32_t *pL = blockL->data + offsetL;
for (int i=0; i < audio_block_samples / 2 * 2; i+=2) { *(d+i) = (int32_t) *pL++; } //interleave
offsetL += audio_block_samples / 2;
} else if (blockR) {
float32_t *pR = blockR->data + offsetR;
for (int i=0; i < audio_block_samples /2 * 2; i+=2) { *(d+i) = (int32_t) *pR++; } //interleave
offsetR += audio_block_samples / 2;
} else {
//memset(dest,0,AUDIO_BLOCK_SAMPLES * 2);
memset(dest,0,audio_block_samples * 4);
return;
}
arm_dcache_flush_delete(dest, sizeof(i2s2_tx_buffer) / 2 );
//if (offsetL < AUDIO_BLOCK_SAMPLES) { //orig Teensy Audio
if (offsetL < (uint16_t)audio_block_samples) {
AudioOutputI2S2_F32::block_left_offset = offsetL;
} else {
AudioOutputI2S2_F32::block_left_offset = 0;
AudioStream_F32::release(blockL);
AudioOutputI2S2_F32::block_left_1st = AudioOutputI2S2_F32::block_left_2nd;
AudioOutputI2S2_F32::block_left_2nd = NULL;
}
//if (offsetR < AUDIO_BLOCK_SAMPLES) { //orig Teensy Audio
if (offsetR < (uint16_t)audio_block_samples) {
AudioOutputI2S2_F32::block_right_offset = offsetR;
} else {
AudioOutputI2S2_F32::block_right_offset = 0;
AudioStream_F32::release(blockR);
AudioOutputI2S2_F32::block_right_1st = AudioOutputI2S2_F32::block_right_2nd;
AudioOutputI2S2_F32::block_right_2nd = NULL;
}
}
#define F32_TO_I16_NORM_FACTOR (32767) //which is 2^15-1
void AudioOutputI2S2_F32::scale_f32_to_i16(float32_t *p_f32, float32_t *p_i16, int len) {
for (int i=0; i<len; i++) { *p_i16++ = max(-F32_TO_I16_NORM_FACTOR,min(F32_TO_I16_NORM_FACTOR,(*p_f32++) * F32_TO_I16_NORM_FACTOR)); }
}
#define F32_TO_I24_NORM_FACTOR (8388607) //which is 2^23-1
void AudioOutputI2S2_F32::scale_f32_to_i24( float32_t *p_f32, float32_t *p_i24, int len) {
for (int i=0; i<len; i++) { *p_i24++ = max(-F32_TO_I24_NORM_FACTOR,min(F32_TO_I24_NORM_FACTOR,(*p_f32++) * F32_TO_I24_NORM_FACTOR)); }
}
#define F32_TO_I32_NORM_FACTOR (2147483647) //which is 2^31-1
//define F32_TO_I32_NORM_FACTOR (8388607) //which is 2^23-1
void AudioOutputI2S2_F32::scale_f32_to_i32( float32_t *p_f32, float32_t *p_i32, int len) {
for (int i=0; i<len; i++) { *p_i32++ = max(-F32_TO_I32_NORM_FACTOR,min(F32_TO_I32_NORM_FACTOR,(*p_f32++) * F32_TO_I32_NORM_FACTOR)); }
//for (int i=0; i<len; i++) { *p_i32++ = (*p_f32++) * F32_TO_I32_NORM_FACTOR + 512.f*8388607.f; }
}
//update has to be carefully coded so that, if audio_blocks are not available, the code exits
//gracefully and won't hang. That'll cause the whole system to hang, which would be very bad.
//static int count = 0;
void AudioOutputI2S2_F32::update(void)
{
// null audio device: discard all incoming data
//if (!active) return;
//audio_block_t *block = receiveReadOnly();
//if (block) release(block);
audio_block_f32_t *block_f32;
audio_block_f32_t *block_f32_scaled = AudioStream_F32::allocate_f32();
audio_block_f32_t *block2_f32_scaled = AudioStream_F32::allocate_f32();
if ((!block_f32_scaled) || (!block2_f32_scaled)) {
//couldn't get some working memory. Return.
if (block_f32_scaled) AudioStream_F32::release(block_f32_scaled);
if (block2_f32_scaled) AudioStream_F32::release(block2_f32_scaled);
return;
}
//now that we have our working memory, proceed with getting the audio data and processing
block_f32 = receiveReadOnly_f32(0); // input 0 = left channel
if (block_f32) {
if (block_f32->length != audio_block_samples) {
Serial.print("AudioOutputI2S2_F32: *** WARNING ***: audio_block says len = ");
Serial.print(block_f32->length);
Serial.print(", but I2S settings want it to be = ");
Serial.println(audio_block_samples);
}
//Serial.print("AudioOutputI2S2_F32: audio_block_samples = ");
//Serial.println(audio_block_samples);
// Optional scaling for easy volume control. Leave outputScale==1.0f for default
if(outputScale<1.0f || outputScale>1.0f)
arm_scale_f32 (block_f32->data, outputScale, block_f32->data, block_f32->length);
//scale F32 to Int32
//block_f32_scaled = AudioStream_F32::allocate_f32();
scale_f32_to_i32(block_f32->data, block_f32_scaled->data, audio_block_samples);
//scale_f32_to_i16(block_f32->data, block_f32_scaled->data, audio_block_samples);
//now process the data blocks
__disable_irq();
if (block_left_1st == NULL) {
block_left_1st = block_f32_scaled;
block_left_offset = 0;
__enable_irq();
} else if (block_left_2nd == NULL) {
block_left_2nd = block_f32_scaled;
__enable_irq();
} else {
audio_block_f32_t *tmp = block_left_1st;
block_left_1st = block_left_2nd;
block_left_2nd = block_f32_scaled;
block_left_offset = 0;
__enable_irq();
AudioStream_F32::release(tmp);
}
AudioStream_F32::transmit(block_f32,0); AudioStream_F32::release(block_f32); //echo the incoming audio out the outputs
} else {
//this branch should never get called, but if it does, let's release the buffer that was never used
AudioStream_F32::release(block_f32_scaled);
}
block_f32_scaled = block2_f32_scaled; //this is simply renaming the pre-allocated buffer
block_f32 = receiveReadOnly_f32(1); // input 1 = right channel
if (block_f32) {
// Optional scaling for easy volume control. Leave outputScale==1.0f for default
if(outputScale<1.0f || outputScale>1.0f)
arm_scale_f32 (block_f32->data, outputScale, block_f32->data, block_f32->length);
//scale F32 to Int32
//block_f32_scaled = AudioStream_F32::allocate_f32();
scale_f32_to_i32(block_f32->data, block_f32_scaled->data, audio_block_samples);
//scale_f32_to_i16(block_f32->data, block_f32_scaled->data, audio_block_samples);
__disable_irq();
if (block_right_1st == NULL) {
block_right_1st = block_f32_scaled;
block_right_offset = 0;
__enable_irq();
} else if (block_right_2nd == NULL) {
block_right_2nd = block_f32_scaled;
__enable_irq();
} else {
audio_block_f32_t *tmp = block_right_1st;
block_right_1st = block_right_2nd;
block_right_2nd = block_f32_scaled;
block_right_offset = 0;
__enable_irq();
AudioStream_F32::release(tmp);
}
AudioStream_F32::transmit(block_f32,1); AudioStream_F32::release(block_f32); //echo the incoming audio out the outputs
} else {
//this branch should never get called, but if it does, let's release the buffer that was never used
AudioStream_F32::release(block_f32_scaled);
}
}
void AudioOutputI2S2_F32::config_i2s(void) { config_i2s(false, AudioOutputI2S2_F32::sample_rate_Hz); }
void AudioOutputI2S2_F32::config_i2s(bool transferUsing32bit) { config_i2s(transferUsing32bit, AudioOutputI2S2_F32::sample_rate_Hz); }
void AudioOutputI2S2_F32::config_i2s(float fs_Hz) { config_i2s(false, fs_Hz); }
void AudioOutputI2S2_F32::config_i2s(bool transferUsing32bit, float fs_Hz)
{
CCM_CCGR5 |= CCM_CCGR5_SAI2(CCM_CCGR_ON);
// if either transmitter or receiver is enabled, do nothing
if (I2S2_TCSR & I2S_TCSR_TE) return;
if (I2S2_RCSR & I2S_RCSR_RE) return;
//PLL:
//int fs = AUDIO_SAMPLE_RATE_EXACT; //original from Teensy Audio Library
int fs = fs_Hz;
// PLL between 27*24 = 648MHz und 54*24=1296MHz
int n1 = 4; //SAI prescaler 4 => (n1*n2) = multiple of 4
int n2 = 1 + (24000000 * 27) / (fs * 256 * n1);
double C = ((double)fs * 256 * n1 * n2) / 24000000;
int c0 = C;
int c2 = 10000;
int c1 = C * c2 - (c0 * c2);
set_audioClock(c0, c1, c2);
// clear SAI2_CLK register locations
CCM_CSCMR1 = (CCM_CSCMR1 & ~(CCM_CSCMR1_SAI2_CLK_SEL_MASK))
| CCM_CSCMR1_SAI2_CLK_SEL(2);
CCM_CS2CDR = (CCM_CS2CDR & ~(CCM_CS2CDR_SAI2_CLK_PRED_MASK | CCM_CS2CDR_SAI2_CLK_PODF_MASK))
| CCM_CS2CDR_SAI2_CLK_PRED(n1-1) // &0x07
| CCM_CS2CDR_SAI2_CLK_PODF(n2-1); // &0x3f
// Select MCLK
IOMUXC_GPR_GPR1 = (IOMUXC_GPR_GPR1
& ~(IOMUXC_GPR_GPR1_SAI2_MCLK3_SEL_MASK))
| (IOMUXC_GPR_GPR1_SAI2_MCLK_DIR | IOMUXC_GPR_GPR1_SAI2_MCLK3_SEL(0));
// CORE_PIN23_CONFIG = 3; //1:MCLK
// CORE_PIN21_CONFIG = 3; //1:RX_BCLK
// CORE_PIN20_CONFIG = 3; //1:RX_SYNC
CORE_PIN33_CONFIG = 2; //EMC_07, 2=SAI2_MCLK
CORE_PIN4_CONFIG = 2; //EMC_06, 2=SAI2_TX_BCLK
CORE_PIN3_CONFIG = 2; //EMC_05, 2=SAI2_TX_SYNC, page 429
int rsync = 1;
int tsync = 0;
I2S2_TMR = 0;
//I2S2_TCSR = (1<<25); //Reset
I2S2_TCR1 = I2S_TCR1_RFW(1);
I2S2_TCR2 = I2S_TCR2_SYNC(tsync) | I2S_TCR2_BCP // sync=0; tx is async;
| (I2S_TCR2_BCD | I2S_TCR2_DIV((1)) | I2S_TCR2_MSEL(1));
I2S2_TCR3 = I2S_TCR3_TCE;
I2S2_TCR4 = I2S_TCR4_FRSZ((2-1)) | I2S_TCR4_SYWD((32-1)) | I2S_TCR4_MF
| I2S_TCR4_FSD | I2S_TCR4_FSE | I2S_TCR4_FSP;
I2S2_TCR5 = I2S_TCR5_WNW((32-1)) | I2S_TCR5_W0W((32-1)) | I2S_TCR5_FBT((32-1));
I2S2_RMR = 0;
//I2S2_RCSR = (1<<25); //Reset
I2S2_RCR1 = I2S_RCR1_RFW(1);
I2S2_RCR2 = I2S_RCR2_SYNC(rsync) | I2S_RCR2_BCP // sync=0; rx is async;
| (I2S_RCR2_BCD | I2S_RCR2_DIV((1)) | I2S_RCR2_MSEL(1));
I2S2_RCR3 = I2S_RCR3_RCE;
I2S2_RCR4 = I2S_RCR4_FRSZ((2-1)) | I2S_RCR4_SYWD((32-1)) | I2S_RCR4_MF
| I2S_RCR4_FSE | I2S_RCR4_FSP | I2S_RCR4_FSD;
I2S2_RCR5 = I2S_RCR5_WNW((32-1)) | I2S_RCR5_W0W((32-1)) | I2S_RCR5_FBT((32-1));
}
/******************************************************************/
// From Chip: The I2SSlave functionality has NOT been extended to
// allow for different block sizes or sample rates (2020-10-31)
void AudioOutputI2S2slave_F32::begin(void)
{
dma.begin(true); // Allocate the DMA channel first
//pinMode(2, OUTPUT);
block_left_1st = NULL;
block_right_1st = NULL;
AudioOutputI2S2slave_F32::config_i2s();
CORE_PIN2_CONFIG = 2; //1:TX_DATA0
dma.TCD->SADDR = i2s2_tx_buffer;
dma.TCD->SOFF = 2;
dma.TCD->ATTR = DMA_TCD_ATTR_SSIZE(1) | DMA_TCD_ATTR_DSIZE(1);
dma.TCD->NBYTES_MLNO = 2;
dma.TCD->SLAST = -sizeof(i2s2_tx_buffer);
//dma.TCD->DADDR = (void *)((uint32_t)&I2S1_TDR1 + 2);
dma.TCD->DOFF = 0;
dma.TCD->CITER_ELINKNO = sizeof(i2s2_tx_buffer) / 2;
dma.TCD->DLASTSGA = 0;
dma.TCD->BITER_ELINKNO = sizeof(i2s2_tx_buffer) / 2;
//dma.triggerAtHardwareEvent(DMAMUX_SOURCE_SAI2_TX);
dma.TCD->DADDR = (void *)((uint32_t)&I2S2_TDR0 + 2);
dma.TCD->CSR = DMA_TCD_CSR_INTHALF | DMA_TCD_CSR_INTMAJOR;
dma.triggerAtHardwareEvent(DMAMUX_SOURCE_SAI2_TX);
dma.enable();
// I2S2_RCSR |= I2S_RCSR_RE | I2S_RCSR_BCE;
// I2S2_TCSR = I2S_TCSR_TE | I2S_TCSR_BCE | I2S_TCSR_FRDE;
I2S2_TCSR |= I2S_TCSR_TE | I2S_TCSR_BCE | I2S_TCSR_FRDE | I2S_TCSR_FR;
update_responsibility = update_setup();
//dma.enable();
dma.attachInterrupt(AudioOutputI2S2_F32::isr);
}
void AudioOutputI2S2slave_F32::config_i2s(void)
{
CCM_CCGR5 |= CCM_CCGR5_SAI2(CCM_CCGR_ON);
// if either transmitter or receiver is enabled, do nothing
if (I2S2_TCSR & I2S_TCSR_TE) return;
if (I2S2_RCSR & I2S_RCSR_RE) return;
// not using MCLK in slave mode - hope that's ok?
// //CORE_PIN23_CONFIG = 3; // AD_B1_09 ALT3=SAI1_MCLK
// CORE_PIN21_CONFIG = 3; // AD_B1_11 ALT3=SAI1_RX_BCLK
// CORE_PIN20_CONFIG = 3; // AD_B1_10 ALT3=SAI1_RX_SYNC
CORE_PIN4_CONFIG = 2; //2:TX_BCLK
CORE_PIN3_CONFIG = 2; //2:TX_SYNC
IOMUXC_SAI2_RX_BCLK_SELECT_INPUT = 0; //page 422 bitclock IOMUXC_SW_MUX_CTL_PAD_GPIO_EMC_06
IOMUXC_SAI2_RX_SYNC_SELECT_INPUT = 0; //IOMUXC_SW_MUX_CTL_PAD_GPIO_EMC_05
int rsync = 1;
int tsync = 0;
// uint32_t noBits = 32;
// uint32_t noBitsM1=noBits-1;
// configure transmitter
I2S2_TMR = 0;
I2S2_TCR1 = I2S_TCR1_RFW(1); // watermark at half fifo size
I2S2_TCR2 = I2S_TCR2_SYNC(tsync) | I2S_TCR2_BCP;
I2S2_TCR3 = I2S_TCR3_TCE;
I2S2_TCR4 = I2S_TCR4_FRSZ(1) | I2S_TCR4_SYWD(31) | I2S_TCR4_MF
| I2S_TCR4_FSE | I2S_TCR4_FSP;
I2S2_TCR5 = I2S_TCR5_WNW(31) | I2S_TCR5_W0W(31) | I2S_TCR5_FBT(31);
// configure receiver
I2S2_RMR = 0;
I2S2_RCR1 = I2S_RCR1_RFW(1);
I2S2_RCR2 = I2S_RCR2_SYNC(rsync) | I2S_TCR2_BCP;
I2S2_RCR3 = I2S_RCR3_RCE;
I2S2_RCR4 = I2S_RCR4_FRSZ(1) | I2S_RCR4_SYWD(31) | I2S_RCR4_MF
| I2S_RCR4_FSE | I2S_RCR4_FSP;
I2S2_RCR5 = I2S_RCR5_WNW(31) | I2S_RCR5_W0W(31) | I2S_RCR5_FBT(31);
}