SoftBank Robotics documentation What's new in NAOqi 2.5?

ALAudioDevice

NAOqi Audio - Overview | API


What it does

ALAudioDevice provides other NAOqi modules with an access to NAO’s audio inputs (microphones) and outputs (loudspeakers).

Every NAOqi module intended to process signals coming from NAO’s microphones or that needs to send signals to NAO’s loudspeakers should use the API provided by ALAudioDevice.

How it works

ALAudioDevice is based on the standard Linux ALSA (Advanced Linux Sound Library) library to communicate with NAO’s sound driver and subsequently to microphones and speakers.

Any NAOqi module can send data to NAO’s loudspeakers by calling, when needed, one of the two appropriate ALAudioDevice methods. Provided that the data is correctly formated, such a call can be made without having to configure anything else.

To process data coming from the microphones, the procedure is different. Indeed, a NAOqi module willing to process such data will first “subscribe” to ALAudioDevice and specify the format of the data that it requires (number of channels, sample rate, etc...). The data correctly formated will then be automatically and regularly sent to the requesting module by using one of its methods as a callback.

Performances and Limitations

ALAudioDevice can provide the sound coming from the microphones formated in one of these ways:

  • four channels interleaved, 48000Hz, 170ms buffer (this is the default setting)
  • four channels deinterleaved, 48000Hz, 170ms buffer
  • one channels (front, rear, left or right), 16000Hz, 170ms buffer

ALAudioDevice can send data to loudspeakers in one of these following framerate:

  • two channels interleaved, 16000Hz (default when an asian language is set)
  • two channels interleaved, 22050Hz (default when a non asian language is set)
  • two channels interleaved, 44100Hz
  • two channels interleaved, 48000Hz

Getting started

To get started about writing a NAOqi module that can process NAO’s microphones data, please refer to the following example that can be used as a nutshell:

alsoundprocessing.h

alsoundprocessing.cpp

To get started about writing a NAOqi module that can send data to NAO’s loudspeakers, please refer to the following example:

alsoundplayback.h

alsoundplayback.cpp

The projects (that can be easily compiled) corresponding to these files can be found in the audio examples directory.