home
The library enables decoding of MP3 and AAC compression and plays 8bit or 16bit wav files. The audio data can come from the Internet, SD card or SPIFFS. Many radio stations can be heard. Playlists are unpacked and a connection to the (first) URL is established, formats are * .pls, * .m3u and * .asx. SSL connections are possible. Google's TTS can be used.
Examples:
connecttohost("http://online.rockarsenal.ru:8000/rockarsenal_aacplus");
connecttospeech("Wenn die Hunde schlafen, kann der Wolf gut Schafe stehlen.", "de");
connecttoSD("click.mp3");
connecttoFS(SD, "/wave_test/Wav_868kb.wav");
connecttoFS(SPIFFS, "wobble.mp3");Stations can be received up to 320Kbit/s. A good connection is a prerequisite for this. Many, but not every, station that runs smoothly in the VLC player works on the ESP32 without dropouts. Shortly before the input buffer is empty, this message appears in the serial monitor
slow stream, dropouts are possible
If the connection is lost, the library tries to re-establish the connection.
Basically all 16 bit DACs that have the pins DIN, BLCK and LRC. The PCM5102A delivers good results. Most GPIOs can be used
setPinout(uint8_t BCLK, uint8_t LRC, uint8_t DOUT); Also ESP32-A1S can be used; the library https://github.com/Yveaux/AC101 can be integrated for this purpose. See the examples folder https://github.com/schreibfaul1/ESP32-audioI2S/tree/master/examples/ESP32-A1S MCLK is required in certain cases (only support GPIO0/GPIO1/GPIO3).
i2s_mclk_pin_select(const uint8_t pin)If the 8-bit sound is enough, you can do that.
...
Audio audio;
...
void setup() {
...
audio.setInternalDAC(true);
...
}The analog audio signal is then on GPIO25 and GPIO 26.
Yes, the library can be downloaded as a zip file. The installation in the Arduino IDE runs via the library manager
Tip: Use the partition scheme 'Huge App' so that there is enough memory for your own extensions
If available, PSRAM can be used. PSRAM is recognized automatically. The input buffer is then automatically relocated and enlarged.
without PSRAM, inputBufferSize is about 6.25KBytes
with PSRAM, inputBufferSize is about 29KBytes
8000 or 30000 bytes are allocated, part of which is used internally to avoid copying the audio data during operation
Internally, the volume is divided into 64 steps. With setVolume () the volume is set in 22 steps. Internally, the 22 steps are assigned to the 64 volume steps via a table. This creates a logarithmic curve. The balance attenuates the left or right channel (values between -16 ...16).
setBalance(-16); // mutes the left channel
setVolume(21); // max loudnessYes, that is possible. There are built-in IIR filters.
setTone(uint8_t l_type, uint16_t l_freq, uint8_t r_type, uint16_t r_freq){
// see https://www.earlevel.com/main/2013/10/13/biquad-calculator-v2/
// l_ left channel, r_ right channel
// type: 0- lowpass, 1- highpass, 2 - lowshelf, 3 - highshelf
// freq: frequency in Hz (between 10....samlerate/2), freq =0 - no filter
// low/highpass have Qfactor = 0.707, low/highshelf have Gain -5dBThe filter type and the nominal frequency can be selected for each channel. High and low pass filters have a quality factor of 0.707, high and lowshelf a gain of -5dB. The filter effect can be evaluated graphically here:
Example code, changes every 10s between lowpass (2KHz), highpass (2KHz) and no filter
Ticker ticker;
void setup(){
...
ticker.attach(10, tcr10s);
...
}
void tcr10s(){ // Timer all 10 seconds
static uint8_t i = 0;
if(i==0) {
log_i("lowpass");
audio.setTone(0, 2000, 0, 2000);
i++; return;
}
if(i==1) {
log_i("highpass");
audio.setTone(1, 2000, 1, 2000);
i++; return;
}
if(i==2) {
log_i("no filter");
audio.setTone(0, 0, 0, 0);
i = 0;
}
return;
}audio_info: authentification failed, wrong credentials?
The name and password can be transferred when the destination is called, use:
connecttohost("http://xxxx", "name", "password");The events are weakly integrated by the compiler. This means they can, but do not have to be used.
audio_info outputs the current status, suitable for debugging and troubleshooting
void audio_info(const char *info)audio_id3data many mp3 files contain information about artists, albums or bands. The data is read from the file and can be used further via this event
void audio_id3data(const char *info)audio_eof_mp3 is called after the end of an audio file. * info contains the file name. With this event it is possible to create playlists
void audio_eof_mp3(const char *info)Playlist example:
void audio_eof_mp3(const char *info){ //end of file
Serial.print("audio_info: "); Serial.println(info);
static int i=0;
if(i==0) audio.connecttoSD("/wave_test/If_I_Had_a_Chicken.mp3");
if(i==1) audio.connecttoSD("/wave_test/test_8bit_stereo.wav");
if(i==2) audio.connecttoSD("/wave_test/test_16bit_mono.wav");
i++;
if(i==3) i=0;
}audio_eof_stream the same as audio_eof_mp3 for podcasts or files transferred from the server
void audio_eof_stream(const char *info)audio_showstation many radio stations provide their names at the beginning of the connection. This can be used in your own applications
void audio_showstation(const char *info)audio_showstreamtitle if the radio station transmits information about the artist, music track ... in its metadata, this event is called
void audio_showstreamtitle(const char *info)audio_bitrate returns the current bit rate as text
void audio_bitrate(const char *info)audio_commercial commercials are often played at the beginning of the broadcast (and during the program). Info contains the expected duration of the advertisement. So the sound can be switched off for the time being
void audio_commercial(const char *info)audio_icyurl if the station has a homepage this is called here
void audio_icyurl(const char *info)audio_lasthost the URL that is called via connecttohost does not have to be the current URL. Sometimes it is redirected to another URL that can be read out here.
void audio_lasthost(const char *info)audio_end_of_speech is called after the voice transmission ends [using connecttospeech(...)]
void audio_eof_speech(const char *info)audio_id3image mp3 files can contain pictures. Here the pointer to the current mp3 file, the position of the beginning of the picture and the size is transmitted
void audio_id3image(File& file, const size_t pos, const size_t size)There are other useful functions for building MP3 players, for example
getAudioFileDuration() Indicates the expected length of an audio file in seconds. With a constant bit rate, CBR, the value is exact, with a variable bit rate, VBR, the duration is estimated based on the first 100 mp3 frames and can therefore deviate slightly from the actual playback time
uint32_t getAudioFileDuration()getAudioCurrentTime() returns the current playing time in seconds
uint32_t getAudioCurrentTime()An example program could look like this:
Ticker ticker;
void setup() {
...
ticker.attach(1, tcr1s);
...
}
void tcr1s(){
uint32_t act=audio.getAudioCurrentTime();
uint32_t afd=audio.getAudioFileDuration();
uint32_t pos =audio.getFilePos();
log_i("pos =%i", pos);
log_i("audioTime: %i:%02d - duration: %i:%02d", (act/60), (act%60) , (afd/60), (afd%60));
}The output in the serial monitor
Sometimes you want to play an audio file in a loop.
setFileLoop() the position is determined internally after the audio header. At the end of the file there is a jump to the audio start position
bool setFileLoop(true);In some projects there is only one audio amplifier or speaker. Then it makes sense to convert the stereo signal into a mono signal. With forceMono(true); the mean value is calculated from the signal of both channels and placed on the left and right channel.
void forceMono(true); // change stereo to mono
void forceMono(false); // default stereo will be playedHere is a simple program example, you need an ESP32 developer board and an external DAC (e.g. PCM5102A)
#include "Arduino.h"
#include "WiFi.h"
#include "Audio.h"
#define I2S_DOUT 26 // connect to DAC pin DIN
#define I2S_BCLK 27 // connect to DAC pin BCK
#define I2S_LRC 25 // connect to DAC pin LCK
Audio audio;
const char* ssid = "SSID";
const char* password = "password";
void setup() {
Serial.begin(115200);
WiFi.begin(ssid, password);
while (WiFi.status() != WL_CONNECTED) delay(1500);
audio.setPinout(I2S_BCLK, I2S_LRC, I2S_DOUT);
audio.connecttohost("http://s1.knixx.fm/dein_webradio_64.aac"); // 64 kbp/s aac+
}
void loop() {
audio.loop();
}
void audio_info(const char *info){
Serial.print("info "); Serial.println(info);
}The output in the serial monitor:
rst:0x1 (POWERON_RESET),boot:0x13 (SPI_FAST_FLASH_BOOT) configsip: 0, SPIWP:0xee clk_drv:0x00,q_drv:0x00,d_drv:0x00,cs0_drv:0x00,hd_drv:0x00,wp_drv:0x00 mode:DIO, clock div:1 load:0x3fff0018,len:4 load:0x3fff001c,len:1216 ho 0 tail 12 room 4 load:0x40078000,len:10944 load:0x40080400,len:6388 entry 0x400806b4 info PSRAM not found, inputBufferSize = 6399 bytes info buffers freed, free Heap: 228148 bytes info Connect to new host: "http://s1.knixx.fm/dein_webradio_64.aac" info Connect to "s1.knixx.fm" on port 80, extension "/dein_webradio_64.aac" info Connected to server info Server: nginx/1.14.2 info audio/aac seen. info format is aac info AACDecoder has been initialized, free Heap: 199916 bytes info chunked data transfer info Connection: close info ice-audio-info: channels=2;samplerate=44100;bitrate=64 info icy-description: Wir spielen Musik von den 60ern bis Heute! Und immer um halb aktuelle Country-Music. info icy-genre: variety,pop,oldies,country info icy-name: knixx.fm - Dein Webradio. / 64 kbp/s aac+ info icy-pub: 1 info icy-url: https://knixx.fm info Cache-Control: no-cache, no-store info Access-Control-Allow-Origin: * info Access-Control-Allow-Headers: Origin, Accept, X-Requested-With, Content-Type info Access-Control-Allow-Methods: GET, OPTIONS, HEAD info Expires: Mon, 26 Jul 1997 05:00:00 GMT info X-Frame-Options: SAMEORIGIN info X-Content-Type-Options: nosniff info Switch to DATA, bitrate is 64000, metaint is 4096 info inputbuffer is being filled info StreamTitle="Michael Bolton - Soul Provider -- 1989" info stream ready info buffer filled in 7 ms info syncword found at pos 0 info AAC Channels=1 info AAC SampleRate=22050 info AAC BitsPerSample=16 info AAC Bitrate=64000 info StreamTitle="Symbol - The Most Beautiful Girl In The World -- 1994"
building it on a breadboard:
the schematic:
There are displays for the Raspberry Pi with a resolution of 480x320 pixels and an SPI bus. These are particularly suitable, see the radio folder
