Var fixes.

This commit is contained in:
Blaž Kristan 2022-06-09 14:44:48 +02:00
parent 184ff7a3b3
commit a6746f77f0
2 changed files with 130 additions and 129 deletions

View File

@ -2,7 +2,6 @@
#include "wled.h" #include "wled.h"
#include <driver/i2s.h> #include <driver/i2s.h>
#include "audio_source.h"
#ifndef ESP32 #ifndef ESP32
#error This audio reactive usermod does not support the ESP8266. #error This audio reactive usermod does not support the ESP8266.
@ -28,6 +27,8 @@
#define DEBUGSR_PRINTF(x...) #define DEBUGSR_PRINTF(x...)
#endif #endif
#include "audio_source.h"
constexpr i2s_port_t I2S_PORT = I2S_NUM_0; constexpr i2s_port_t I2S_PORT = I2S_NUM_0;
constexpr int BLOCK_SIZE = 128; constexpr int BLOCK_SIZE = 128;
constexpr int SAMPLE_RATE = 10240; // Base sample rate in Hz constexpr int SAMPLE_RATE = 10240; // Base sample rate in Hz
@ -38,6 +39,10 @@ constexpr int SAMPLE_RATE = 10240; // Base sample rate in Hz
//#define MAJORPEAK_SUPPRESS_NOISE // define to activate a dirty hack that ignores the lowest + hightest FFT bins //#define MAJORPEAK_SUPPRESS_NOISE // define to activate a dirty hack that ignores the lowest + hightest FFT bins
byte audioSyncEnabled = 0;
uint16_t audioSyncPort = 11988;
uint8_t inputLevel; // UI slider value
// //
// AGC presets // AGC presets
// Note: in C++, "const" implies "static" - no need to explicitly declare everything as "static const" // Note: in C++, "const" implies "static" - no need to explicitly declare everything as "static const"
@ -92,41 +97,46 @@ const uint16_t samples = 512; // This value MUST ALWAYS be a p
static AudioSource *audioSource; static AudioSource *audioSource;
byte soundSquelch = 10; // default squelch value for volume reactive routines static byte soundSquelch = 10; // default squelch value for volume reactive routines
byte sampleGain = 1; // default sample gain static byte sampleGain = 1; // default sample gain
uint16_t micData; // Analog input for FFT static uint16_t micData; // Analog input for FFT
uint16_t micDataSm; // Smoothed mic data, as it's a bit twitchy static uint16_t micDataSm; // Smoothed mic data, as it's a bit twitchy
static float micDataReal = 0.0f; // future support - this one has the full 24bit MicIn data - lowest 8bit after decimal point
static byte soundAgc = 0; // default Automagic gain control
static float multAgc = 1.0f; // sample * multAgc = sampleAgc. Our multiplier
static uint16_t noiseFloor = 100; // default squelch value for FFT reactive routines
double FFT_MajorPeak = 0; static double FFT_MajorPeak = 0;
double FFT_Magnitude = 0; static double FFT_Magnitude = 0;
//uint16_t mAvg = 0; //static uint16_t mAvg = 0;
// These are the input and output vectors. Input vectors receive computed results from FFT. // These are the input and output vectors. Input vectors receive computed results from FFT.
static double vReal[samplesFFT]; static double vReal[samplesFFT];
static double vImag[samplesFFT]; static double vImag[samplesFFT];
float fftBin[samplesFFT]; static float fftBin[samplesFFT];
// Try and normalize fftBin values to a max of 4096, so that 4096/16 = 256. // Try and normalize fftBin values to a max of 4096, so that 4096/16 = 256.
// Oh, and bins 0,1,2 are no good, so we'll zero them out. // Oh, and bins 0,1,2 are no good, so we'll zero them out.
float fftCalc[16]; static float fftCalc[16];
uint8_t fftResult[16]; // Our calculated result table, which we feed to the animations. static uint8_t fftResult[16]; // Our calculated result table, which we feed to the animations.
//float fftResultMax[16]; // A table used for testing to determine how our post-processing is working. #ifdef SR_DEBUG
float fftAvg[16]; static float fftResultMax[16]; // A table used for testing to determine how our post-processing is working.
#endif
static float fftAvg[16];
// Table of linearNoise results to be multiplied by soundSquelch in order to reduce squelch across fftResult bins. // Table of linearNoise results to be multiplied by soundSquelch in order to reduce squelch across fftResult bins.
uint16_t linearNoise[16] = { 34, 28, 26, 25, 20, 12, 9, 6, 4, 4, 3, 2, 2, 2, 2, 2 }; static uint16_t linearNoise[16] = { 34, 28, 26, 25, 20, 12, 9, 6, 4, 4, 3, 2, 2, 2, 2, 2 };
// Table of multiplication factors so that we can even out the frequency response. // Table of multiplication factors so that we can even out the frequency response.
float fftResultPink[16] = { 1.70f, 1.71f, 1.73f, 1.78f, 1.68f, 1.56f, 1.55f, 1.63f, 1.79f, 1.62f, 1.80f, 2.06f, 2.47f, 3.35f, 6.83f, 9.55f }; static float fftResultPink[16] = { 1.70f, 1.71f, 1.73f, 1.78f, 1.68f, 1.56f, 1.55f, 1.63f, 1.79f, 1.62f, 1.80f, 2.06f, 2.47f, 3.35f, 6.83f, 9.55f };
// Create FFT object // Create FFT object
arduinoFFT FFT = arduinoFFT(vReal, vImag, samples, SAMPLE_RATE); static arduinoFFT FFT = arduinoFFT(vReal, vImag, samples, SAMPLE_RATE);
float fftAdd(int from, int to) { float fftAdd(int from, int to) {
int i = from; float result = 0.0f;
float result = 0; for (int i = from; i <= to; i++) {
while (i <= to) { result += fftBin[i];
result += fftBin[i++];
} }
return result; return result;
} }
@ -146,6 +156,7 @@ void FFTcode(void * parameter) {
// Only run the FFT computing code if we're not in Receive mode // Only run the FFT computing code if we're not in Receive mode
if (audioSyncEnabled & (1 << 1)) if (audioSyncEnabled & (1 << 1))
continue; continue;
audioSource->getSamples(vReal, samplesFFT); audioSource->getSamples(vReal, samplesFFT);
// old code - Last sample in vReal is our current mic sample // old code - Last sample in vReal is our current mic sample
@ -248,11 +259,9 @@ void FFTcode(void * parameter) {
vReal[samplesFFT-1] = xtemp[23]; vReal[samplesFFT-1] = xtemp[23];
#endif #endif
for (int i = 0; i < samplesFFT; i++) { // Values for bins 0 and 1 are WAY too large. Might as well start at 3. for (int i = 0; i < samplesFFT; i++) { // Values for bins 0 and 1 are WAY too large. Might as well start at 3.
double t = 0.0; float t = fabs(vReal[i]); // just to be sure - values in fft bins should be positive any way
t = fabs(vReal[i]); // just to be sure - values in fft bins should be positive any way fftBin[i] = t / 16.0; // Reduce magnitude. Want end result to be linear and ~4096 max.
t = t / 16.0; // Reduce magnitude. Want end result to be linear and ~4096 max.
fftBin[i] = t;
} // for() } // for()
@ -286,12 +295,12 @@ void FFTcode(void * parameter) {
// Noise supression of fftCalc bins using soundSquelch adjustment for different input types. // Noise supression of fftCalc bins using soundSquelch adjustment for different input types.
for (int i=0; i < 16; i++) { for (int i=0; i < 16; i++) {
fftCalc[i] = fftCalc[i]-(float)soundSquelch*(float)linearNoise[i]/4.0 <= 0? 0 : fftCalc[i]; fftCalc[i] -= (float)soundSquelch*(float)linearNoise[i]/4.0 <= 0? 0 : fftCalc[i];
} }
// Adjustment for frequency curves. // Adjustment for frequency curves.
for (int i=0; i < 16; i++) { for (int i=0; i < 16; i++) {
fftCalc[i] = fftCalc[i] * fftResultPink[i]; fftCalc[i] *= fftResultPink[i];
} }
// Manual linear adjustment of gain using sampleGain adjustment for different input types. // Manual linear adjustment of gain using sampleGain adjustment for different input types.
@ -325,88 +334,6 @@ void FFTcode(void * parameter) {
} // FFTcode() } // FFTcode()
void logAudio() {
#ifdef MIC_LOGGER
//Serial.print("micData:"); Serial.print(micData); Serial.print("\t");
//Serial.print("micDataSm:"); Serial.print(micDataSm); Serial.print("\t");
//Serial.print("micIn:"); Serial.print(micIn); Serial.print("\t");
//Serial.print("micLev:"); Serial.print(micLev); Serial.print("\t");
//Serial.print("sample:"); Serial.print(sample); Serial.print("\t");
//Serial.print("sampleAvg:"); Serial.print(sampleAvg); Serial.print("\t");
Serial.print("sampleReal:"); Serial.print(sampleReal); Serial.print("\t");
//Serial.print("sampleMax:"); Serial.print(sampleMax); Serial.print("\t");
Serial.print("multAgc:"); Serial.print(multAgc, 4); Serial.print("\t");
Serial.print("sampleAgc:"); Serial.print(sampleAgc); Serial.print("\t");
Serial.println(" ");
#endif
#ifdef MIC_SAMPLING_LOG
//------------ Oscilloscope output ---------------------------
Serial.print(targetAgc); Serial.print(" ");
Serial.print(multAgc); Serial.print(" ");
Serial.print(sampleAgc); Serial.print(" ");
Serial.print(sample); Serial.print(" ");
Serial.print(sampleAvg); Serial.print(" ");
Serial.print(micLev); Serial.print(" ");
Serial.print(samplePeak); Serial.print(" "); //samplePeak = 0;
Serial.print(micIn); Serial.print(" ");
Serial.print(100); Serial.print(" ");
Serial.print(0); Serial.print(" ");
Serial.println(" ");
#endif
#ifdef FFT_SAMPLING_LOG
#if 0
for(int i=0; i<16; i++) {
Serial.print(fftResult[i]);
Serial.print("\t");
}
Serial.println("");
#endif
// OPTIONS are in the following format: Description \n Option
//
// Set true if wanting to see all the bands in their own vertical space on the Serial Plotter, false if wanting to see values in Serial Monitor
const bool mapValuesToPlotterSpace = false;
// Set true to apply an auto-gain like setting to to the data (this hasn't been tested recently)
const bool scaleValuesFromCurrentMaxVal = false;
// prints the max value seen in the current data
const bool printMaxVal = false;
// prints the min value seen in the current data
const bool printMinVal = false;
// if !scaleValuesFromCurrentMaxVal, we scale values from [0..defaultScalingFromHighValue] to [0..scalingToHighValue], lower this if you want to see smaller values easier
const int defaultScalingFromHighValue = 256;
// Print values to terminal in range of [0..scalingToHighValue] if !mapValuesToPlotterSpace, or [(i)*scalingToHighValue..(i+1)*scalingToHighValue] if mapValuesToPlotterSpace
const int scalingToHighValue = 256;
// set higher if using scaleValuesFromCurrentMaxVal and you want a small value that's also the current maxVal to look small on the plotter (can't be 0 to avoid divide by zero error)
const int minimumMaxVal = 1;
int maxVal = minimumMaxVal;
int minVal = 0;
for(int i = 0; i < 16; i++) {
if(fftResult[i] > maxVal) maxVal = fftResult[i];
if(fftResult[i] < minVal) minVal = fftResult[i];
}
for(int i = 0; i < 16; i++) {
Serial.print(i); Serial.print(":");
Serial.printf("%04d ", map(fftResult[i], 0, (scaleValuesFromCurrentMaxVal ? maxVal : defaultScalingFromHighValue), (mapValuesToPlotterSpace*i*scalingToHighValue)+0, (mapValuesToPlotterSpace*i*scalingToHighValue)+scalingToHighValue-1));
}
if(printMaxVal) {
Serial.printf("maxVal:%04d ", maxVal + (mapValuesToPlotterSpace ? 16*256 : 0));
}
if(printMinVal) {
Serial.printf("%04d:minVal ", minVal); // printed with value first, then label, so negative values can be seen in Serial Monitor but don't throw off y axis in Serial Plotter
}
if(mapValuesToPlotterSpace)
Serial.printf("max:%04d ", (printMaxVal ? 17 : 16)*256); // print line above the maximum value we expect to see on the plotter to avoid autoscaling y axis
else
Serial.printf("max:%04d ", 256);
Serial.println();
#endif // FFT_SAMPLING_LOG
} // logAudio()
//class name. Use something descriptive and leave the ": public Usermod" part :) //class name. Use something descriptive and leave the ": public Usermod" part :)
class AudioReactive : public Usermod { class AudioReactive : public Usermod {
@ -437,12 +364,7 @@ class AudioReactive : public Usermod {
int8_t i2sckPin = I2S_CKPIN; int8_t i2sckPin = I2S_CKPIN;
#endif #endif
//byte soundAgc = 0; // default Automagic gain control
//uint16_t noiseFloor = 100; // default squelch value for FFT reactive routines
WiFiUDP fftUdp; WiFiUDP fftUdp;
byte audioSyncEnabled = 0;
uint16_t audioSyncPort = 11988;
struct audioSyncPacket { struct audioSyncPacket {
char header[6] = UDP_SYNC_HEADER; char header[6] = UDP_SYNC_HEADER;
uint8_t myVals[32]; // 32 Bytes uint8_t myVals[32]; // 32 Bytes
@ -475,9 +397,7 @@ class AudioReactive : public Usermod {
int rawSampleAgc = 0; // Our AGC sample - raw int rawSampleAgc = 0; // Our AGC sample - raw
long timeOfPeak = 0; long timeOfPeak = 0;
long lastTime = 0; long lastTime = 0;
float micDataReal = 0.0; // future support - this one has the full 24bit MicIn data - lowest 8bit after decimal point
float micLev = 0.f; // Used to convert returned value to have '0' as minimum. A leveller float micLev = 0.f; // Used to convert returned value to have '0' as minimum. A leveller
float multAgc = 1.0f; // sample * multAgc = sampleAgc. Our multiplier
float sampleAvg = 0.f; // Smoothed Average float sampleAvg = 0.f; // Smoothed Average
float beat = 0.f; // beat Detection float beat = 0.f; // beat Detection
@ -499,6 +419,87 @@ class AudioReactive : public Usermod {
} }
} }
void logAudio() {
#ifdef MIC_LOGGER
//Serial.print("micData:"); Serial.print(micData); Serial.print("\t");
//Serial.print("micDataSm:"); Serial.print(micDataSm); Serial.print("\t");
//Serial.print("micIn:"); Serial.print(micIn); Serial.print("\t");
//Serial.print("micLev:"); Serial.print(micLev); Serial.print("\t");
//Serial.print("sample:"); Serial.print(sample); Serial.print("\t");
//Serial.print("sampleAvg:"); Serial.print(sampleAvg); Serial.print("\t");
Serial.print("sampleReal:"); Serial.print(sampleReal); Serial.print("\t");
//Serial.print("sampleMax:"); Serial.print(sampleMax); Serial.print("\t");
Serial.print("multAgc:"); Serial.print(multAgc, 4); Serial.print("\t");
Serial.print("sampleAgc:"); Serial.print(sampleAgc); Serial.print("\t");
Serial.println(" ");
#endif
#ifdef MIC_SAMPLING_LOG
//------------ Oscilloscope output ---------------------------
Serial.print(targetAgc); Serial.print(" ");
Serial.print(multAgc); Serial.print(" ");
Serial.print(sampleAgc); Serial.print(" ");
Serial.print(sample); Serial.print(" ");
Serial.print(sampleAvg); Serial.print(" ");
Serial.print(micLev); Serial.print(" ");
Serial.print(samplePeak); Serial.print(" "); //samplePeak = 0;
Serial.print(micIn); Serial.print(" ");
Serial.print(100); Serial.print(" ");
Serial.print(0); Serial.print(" ");
Serial.println(" ");
#endif
#ifdef FFT_SAMPLING_LOG
#if 0
for(int i=0; i<16; i++) {
Serial.print(fftResult[i]);
Serial.print("\t");
}
Serial.println("");
#endif
// OPTIONS are in the following format: Description \n Option
//
// Set true if wanting to see all the bands in their own vertical space on the Serial Plotter, false if wanting to see values in Serial Monitor
const bool mapValuesToPlotterSpace = false;
// Set true to apply an auto-gain like setting to to the data (this hasn't been tested recently)
const bool scaleValuesFromCurrentMaxVal = false;
// prints the max value seen in the current data
const bool printMaxVal = false;
// prints the min value seen in the current data
const bool printMinVal = false;
// if !scaleValuesFromCurrentMaxVal, we scale values from [0..defaultScalingFromHighValue] to [0..scalingToHighValue], lower this if you want to see smaller values easier
const int defaultScalingFromHighValue = 256;
// Print values to terminal in range of [0..scalingToHighValue] if !mapValuesToPlotterSpace, or [(i)*scalingToHighValue..(i+1)*scalingToHighValue] if mapValuesToPlotterSpace
const int scalingToHighValue = 256;
// set higher if using scaleValuesFromCurrentMaxVal and you want a small value that's also the current maxVal to look small on the plotter (can't be 0 to avoid divide by zero error)
const int minimumMaxVal = 1;
int maxVal = minimumMaxVal;
int minVal = 0;
for(int i = 0; i < 16; i++) {
if(fftResult[i] > maxVal) maxVal = fftResult[i];
if(fftResult[i] < minVal) minVal = fftResult[i];
}
for(int i = 0; i < 16; i++) {
Serial.print(i); Serial.print(":");
Serial.printf("%04d ", map(fftResult[i], 0, (scaleValuesFromCurrentMaxVal ? maxVal : defaultScalingFromHighValue), (mapValuesToPlotterSpace*i*scalingToHighValue)+0, (mapValuesToPlotterSpace*i*scalingToHighValue)+scalingToHighValue-1));
}
if(printMaxVal) {
Serial.printf("maxVal:%04d ", maxVal + (mapValuesToPlotterSpace ? 16*256 : 0));
}
if(printMinVal) {
Serial.printf("%04d:minVal ", minVal); // printed with value first, then label, so negative values can be seen in Serial Monitor but don't throw off y axis in Serial Plotter
}
if(mapValuesToPlotterSpace)
Serial.printf("max:%04d ", (printMaxVal ? 17 : 16)*256); // print line above the maximum value we expect to see on the plotter to avoid autoscaling y axis
else
Serial.printf("max:%04d ", 256);
Serial.println();
#endif // FFT_SAMPLING_LOG
} // logAudio()
/* /*
* A "PI control" multiplier to automatically adjust sound sensitivity. * A "PI control" multiplier to automatically adjust sound sensitivity.
* *

View File

@ -135,13 +135,13 @@ public:
esp_err_t err = i2s_driver_install(I2S_NUM_0, &_config, 0, nullptr); esp_err_t err = i2s_driver_install(I2S_NUM_0, &_config, 0, nullptr);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to install i2s driver: %d\n", err); DEBUGSR_PRINTF("Failed to install i2s driver: %d\n", err);
return; return;
} }
err = i2s_set_pin(I2S_NUM_0, &_pinConfig); err = i2s_set_pin(I2S_NUM_0, &_pinConfig);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to set i2s pin config: %d\n", err); DEBUGSR_PRINTF("Failed to set i2s pin config: %d\n", err);
return; return;
} }
@ -152,7 +152,7 @@ public:
_initialized = false; _initialized = false;
esp_err_t err = i2s_driver_uninstall(I2S_NUM_0); esp_err_t err = i2s_driver_uninstall(I2S_NUM_0);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to uninstall i2s driver: %d\n", err); DEBUGSR_PRINTF("Failed to uninstall i2s driver: %d\n", err);
return; return;
} }
pinManager.deallocatePin(i2swsPin, PinOwner::DigitalMic); pinManager.deallocatePin(i2swsPin, PinOwner::DigitalMic);
@ -174,13 +174,13 @@ public:
err = i2s_read(I2S_NUM_0, (void *)newSamples, sizeof(newSamples), &bytes_read, portMAX_DELAY); err = i2s_read(I2S_NUM_0, (void *)newSamples, sizeof(newSamples), &bytes_read, portMAX_DELAY);
if ((err != ESP_OK)){ if ((err != ESP_OK)){
Serial.printf("Failed to get samples: %d\n", err); DEBUGSR_PRINTF("Failed to get samples: %d\n", err);
return; return;
} }
// For correct operation, we need to read exactly sizeof(samples) bytes from i2s // For correct operation, we need to read exactly sizeof(samples) bytes from i2s
if(bytes_read != sizeof(newSamples)) { if(bytes_read != sizeof(newSamples)) {
Serial.printf("Failed to get enough samples: wanted: %d read: %d\n", sizeof(newSamples), bytes_read); DEBUGSR_PRINTF("Failed to get enough samples: wanted: %d read: %d\n", sizeof(newSamples), bytes_read);
return; return;
} }
@ -349,7 +349,7 @@ public:
// Determine Analog channel. Only Channels on ADC1 are supported // Determine Analog channel. Only Channels on ADC1 are supported
int8_t channel = digitalPinToAnalogChannel(audioPin); int8_t channel = digitalPinToAnalogChannel(audioPin);
if (channel > 9) { if (channel > 9) {
Serial.printf("Incompatible GPIO used for audio in: %d\n", audioPin); DEBUGSR_PRINTF("Incompatible GPIO used for audio in: %d\n", audioPin);
return; return;
} else { } else {
adc_gpio_init(ADC_UNIT_1, adc_channel_t(channel)); adc_gpio_init(ADC_UNIT_1, adc_channel_t(channel));
@ -358,14 +358,14 @@ public:
// Install Driver // Install Driver
esp_err_t err = i2s_driver_install(I2S_NUM_0, &_config, 0, nullptr); esp_err_t err = i2s_driver_install(I2S_NUM_0, &_config, 0, nullptr);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to install i2s driver: %d\n", err); DEBUGSR_PRINTF("Failed to install i2s driver: %d\n", err);
return; return;
} }
// Enable I2S mode of ADC // Enable I2S mode of ADC
err = i2s_set_adc_mode(ADC_UNIT_1, adc1_channel_t(channel)); err = i2s_set_adc_mode(ADC_UNIT_1, adc1_channel_t(channel));
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to set i2s adc mode: %d\n", err); DEBUGSR_PRINTF("Failed to set i2s adc mode: %d\n", err);
return; return;
} }
@ -374,7 +374,7 @@ public:
// fingers crossed // fingers crossed
err = i2s_adc_enable(I2S_NUM_0); err = i2s_adc_enable(I2S_NUM_0);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to enable i2s adc: %d\n", err); DEBUGSR_PRINTF("Failed to enable i2s adc: %d\n", err);
//return; //return;
} }
#endif #endif
@ -393,7 +393,7 @@ public:
esp_err_t err = i2s_adc_enable(I2S_NUM_0); esp_err_t err = i2s_adc_enable(I2S_NUM_0);
//esp_err_t err = i2s_start(I2S_NUM_0); //esp_err_t err = i2s_start(I2S_NUM_0);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to enable i2s adc: %d\n", err); DEBUGSR_PRINTF("Failed to enable i2s adc: %d\n", err);
return; return;
} }
#endif #endif
@ -404,7 +404,7 @@ public:
err = i2s_adc_disable(I2S_NUM_0); err = i2s_adc_disable(I2S_NUM_0);
//err = i2s_stop(I2S_NUM_0); //err = i2s_stop(I2S_NUM_0);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to disable i2s adc: %d\n", err); DEBUGSR_PRINTF("Failed to disable i2s adc: %d\n", err);
return; return;
} }
#endif #endif
@ -420,13 +420,13 @@ public:
// fingers crossed // fingers crossed
err = i2s_adc_disable(I2S_NUM_0); err = i2s_adc_disable(I2S_NUM_0);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to disable i2s adc: %d\n", err); DEBUGSR_PRINTF("Failed to disable i2s adc: %d\n", err);
//return; //return;
} }
#endif #endif
err = i2s_driver_uninstall(I2S_NUM_0); err = i2s_driver_uninstall(I2S_NUM_0);
if (err != ESP_OK) { if (err != ESP_OK) {
Serial.printf("Failed to uninstall i2s driver: %d\n", err); DEBUGSR_PRINTF("Failed to uninstall i2s driver: %d\n", err);
return; return;
} }
} }