Using the Arduino Audio Tools in Xeus Jupyterlab¶

In my AudioTools I have quite a few sound effects and it is quite a challange to test them all. In order to make my life a little bit easier I decided to make my framework usable in Jupyterlab.

xeus-cling is a Jupyter kernel for C++ based on the C++ interpreter cling and the native implementation of the Jupyter protocol xeus. So we can use the AudioTools directly in Jupyterlab with Xeus/Cling!

As a precondition I expect that you have Xeus/Cling already installed!

So, first we need to provide the path to the source code. We can use the one that we have installed for Arduino

In [1]:

#pragma cling add_include_path("/home/pschatzmann/Arduino/libraries/arduino-audio-tools/src")

Next we can include the application. We also add AudioLibs/Jupyter.h which provides the API for Jupyter

In [2]:

#include "AudioTools.h"
#include "AudioTools/AudioLibs/Jupyter.h"

Audio API¶

Now we are ready to define the audio. Nothing special here:

In [3]:

int frequency = 800;
AudioInfo info(44100, 2, 16);
SineWaveGenerator<int16_t> sineWave(32000);   
sineWave.begin(info, frequency);
GeneratedSoundStream<int16_t> sound(sineWave);             // Stream generated from sine wave

In order to output sound in Jupyterlab we create a JupyterAudio object which defines the generated wav file name, the audio source and the number of buffers and buffer size to limit/specify then length of the generated audio.

In [4]:

JupyterAudio audio("test1.wav", sound, 600, 1024);

Outputting the audio object is generating a Web Audio Player

In [5]:

audio

Out[5]:

We can also display the audio as a chart

In [7]:

audio.chart(0)

Out[7]:

Files¶

I am also supporting the Arduino SD File API

In [8]:

auto file = SD.open("test1.wav", FILE_READ);
file.size()

Out[8]:

In [9]:

file.close();
SD.remove("test1.wav");

Using C++¶

We can use standard C++ to process or output data:

In [10]:

for (int j=0;j<10;j++){
    std::cout << sineWave.readSample() << endl;
}

In [ ]: