ESP32 - Playing around with the I2S peripheral and DAC audio output

 ESP32 Arduino - Playing around with the I2S peripheral, ADC and DAC audio output

 In some of my last posts, I mentioned the ESP32 based boards I recently received from DigitSpace, and I finally found some time to play around with some of the audio based peripherals. In this case I'm starting small, attempting to create a simple sine wave output using the internal DAC.

The documentation for the I2S peripheral is a bit limited it seems, but there is enough information in the Arduino Github repo to make sense of things and get some basic audio output going.

The general theory seems to be pretty simple compared to working with previous libraries, TMRpcm and AutoAnalogAudio, which utilize timer driven interrupts and nested interrupts (AVR devices) to handle asynchronous audio processing. The ESP32 provides a number of methods to handle audio processing, with peripherals designed to offload this work from the central processor, making the user interface mostly about buffering and handling audio data.

I've put together a simple outline of the code on GitHub.

The above code will produce a simple sine wave output on the DAC pins (see previous ESP32 related posts for pinout) and can be toggled off/on by entering a '1' on the Arduino Serial monitor. An audio file can also be defined and played by entering a '2' in the Serial monitor. In this case the sketch is setup for a 32Khz, 16-bit, Mono audio file.

In this case the hardest part is following the documentation found above and figuring out just how simple it actually is to handle the I2S peripheral. Really, you just configure the appropriate sample rate and output pins, mode etc, and then feed data into the peripheral. If doing any digital signal processing, its really about providing enough CPU time to handle the data and feed it into I2S. If handling audio files, the hardest part is decoding.

This should prove to be a good basis for the AutoAnalogAudio library since the only thing left to do is add ADC support and either use timers or RTOS tasks to handle asynchronous processing of the audio.