Which filter should I use for Echo Cancellation?

[ziv2013]

Problem Description

Hi,

Thanks for your lib. It's very useful.

I have build a device with one I2S Mic and one I2S speaker with ESP32S3. Based "streams-i2s-i2s" example, it's done quickly.

But I found the MIC is very sensitive, there would be echo when I set the speaker load.

Which filter should I use for Echo Cancellation?

Ziv

Device Description

Self made board, core is ESP32-S3

Sketch

/**
   @file streams-i2s-i2s-2.ino
   @brief Copy audio from I2S to I2S: We use 2 different i2s ports!
   @author Phil Schatzmann
   @copyright GPLv3
*/
#include <Wire.h>
#include "AudioTools.h"

AudioInfo IN_info(16000, 2, 32);
AudioInfo OUT_info(16000, 2, 32);
I2SStream in;
I2SStream out;
StreamCopy copier(out, in); // copies sound into i2s
//FormatConverterStream converter(in);  // or use converter(out)
//StreamCopy copier(out, converter);       //        copier(converter, sound);
// HT513 音量
uint16_t Volume;
// 最小音量
#define LOWESTVOLUME 1
#define TOLENCE 16
#define HT513_ADDR_L 0x6c
/**
   @brief  ht513写寄存器
   @param  addr 寄存器地址
   @param  val 要写的值
   @retval None
*/
void HT513_WriteOneByte(uint8_t addr, uint8_t val)
{
  Wire.beginTransmission(HT513_ADDR_L);
  Wire.write(addr);
  Wire.write(val);
  int ack = Wire.endTransmission(true);
  Serial.print("Ack ");
  Serial.println(ack, HEX);
}

/**
   @brief  ht513读寄存器
   @param  addr 寄存器地址
   @retval 读取到的寄存器值
*/
uint8_t HT513_ReadOneByte(uint8_t addr)
{
  uint8_t temp = 0;

  Wire.beginTransmission(HT513_ADDR_L);
  Wire.write(addr);
  Wire.endTransmission(false);
  uint8_t bytesReceived = 0;
  bytesReceived = Wire.requestFrom(HT513_ADDR_L, (uint8_t)1, true);
  if (bytesReceived == 1) {
    temp = Wire.read();
  }
  else {
    Serial.println("Read Error ");
  }

  return temp;
}

// Arduino Setup
void setup(void) {
  delay(5000);
  // HT513 SD Pin 需要设置为 High
  pinMode(8, OUTPUT);
  digitalWrite(8, HIGH);
  analogReadResolution(9);
  // Open Serial
  Serial.begin(115200);
  // change to Warning to improve the quality
  //AudioToolsLogger.begin(Serial, AudioToolsLogLevel::Info);

  Wire.begin(18, 17);
  int nDevices;
  byte error, address;
  Serial.println("Scanning...");
  nDevices = 0;
  for ( address = 1; address < 127; address++ ) {
    Wire.beginTransmission(address);
    error = Wire.endTransmission();
    if (error == 0) {
      Serial.print("I2C device found at address 0x");
      if (address < 16) {
        Serial.print("0");
      }
      Serial.println(address, HEX);
      nDevices++;
    }
    else if (error == 4) {
      Serial.print("Unknow error at address 0x");
      if (address < 16) {
        Serial.print("0");
      }
      Serial.println(address, HEX);
    }
  }
  if (nDevices == 0) {
    Serial.println("No I2C devices found\n");
  }
  else {
    Serial.println("done\n");
  }

  //  设置 SD 为LOW
  HT513_WriteOneByte(0x12, 0b11110000);

  // 设置数据格式为 I2S, 32Bits
  HT513_WriteOneByte(0x13, 0b00000000);

  // 读取音量设置
  Volume = analogRead(3);
  uint8_t Vol=(Volume,0,511,0x07,0xff);
  HT513_WriteOneByte(0x16, Vol);
  HT513_WriteOneByte(0x15, Vol);
  Serial.println(Volume, HEX);

  // 调整声道
  HT513_WriteOneByte(0x17, 0b10110000);

  Serial.println("++++++++++++++++");
  //  设置 SD 为HIGH
  HT513_WriteOneByte(0x12, 0b11110100);
  uint8_t Value = HT513_ReadOneByte(0x12);
  Serial.println(Value, HEX);
  Value = HT513_ReadOneByte(0x13);
  Serial.println(Value, HEX);
  Value = HT513_ReadOneByte(0x16);
  Serial.println(Value, HEX);
  Value = HT513_ReadOneByte(0x17);
  Serial.println(Value, HEX);

  // Define Converter
  //converter.begin(IN_info, OUT_info);

  // start I2S in
  Serial.println("starting I2S...");
  auto config_in = in.defaultConfig(RX_MODE);
  config_in.copyFrom(IN_info);
  config_in.i2s_format = I2S_STD_FORMAT;
  config_in.is_master = true;
  config_in.port_no = 1;
  config_in.pin_bck = 37;
  config_in.pin_ws = 38;
  config_in.pin_data = 36;
  // config_in.fixed_mclk = sample_rate * 256
  // config_in.pin_mck = 2
  in.begin(config_in);

  // start I2S out
  auto config_out = out.defaultConfig(TX_MODE);
  config_out.copyFrom(OUT_info);
  config_out.i2s_format = I2S_STD_FORMAT;
  config_out.is_master = true;
  config_out.port_no = 0;
  config_out.pin_bck = 15;
  config_out.pin_ws = 6;
  config_out.pin_data = 7;
  config_out.pin_mck = 16;
  out.begin(config_out);

  Serial.println("I2S started...");


}

// Arduino loop - copy sound to out
void loop() {
  copier.copy();
  if (abs(analogRead(3) - Volume) > TOLENCE) {
    // 读取音量设置
    Volume = analogRead(3);
    //  设置 SD 为LOW
    HT513_WriteOneByte(0x12, 0b11110000);
    uint8_t Vol=map(Volume,0,511,0x07,0xff);
    HT513_WriteOneByte(0x16, Vol);
    //  设置 SD 为HIGH
    HT513_WriteOneByte(0x12, 0b11110100);
    Serial.print(analogRead(3), HEX);
    Serial.print("  ");
    Serial.print(Volume, HEX);
    Serial.print("  ");
    Serial.println(Vol, HEX);
  }


}

Other Steps to Reproduce

No response

What is your development environment (incl. core version info)

Arduino 2.3.6
ESP32 by Espressif system 3.2.0
arduino-audio-tools-1.1.2

I have checked existing issues, discussions and online documentation

• I confirm I have checked existing issues, discussions and online documentation

[pschatzmann]

I guess your issue is rather audio feed-back than echo.

Here is some theoretical advice:

• Make sure that you use a directional microphone that is not in the directly in the speaker path: maybe you can use some wave guides or use a headphone or earphones.
• Reduce the sensitivitiy of the microphone (evenutally switch off or reduce the volume the microphone dynamically if it is not used or if you get feed back). You can use the VolumeStream and the VolumeMeter if your ADC does not support this.
• You can try to use a Notch Filter to filter out the frequencies that cause problems. Eventually you can try to combine it with FFT to determine the frequencies with issues.

[ziv2013]

Thanks for your suggestion. I will try and feedback.

[ziv2013]

I use VolumeStream to reduce the sensitivity of the microphone. If the output sound after AMP(HT513), there still would be echo.