audio_capture

This repositiory provides a set of ROS 2 packages for audio. It provides a Python version to capture and play audio data using pyaudio.

Table of Contents

1. Installation
2. Docker
3. Nodes
4. Demos

Installation

$ cd ~/ros2_ws/src
$ git clone https://github.com/mgonzs13/audio_common.git
$ cd ~/ros2_ws
$ rosdep install --from-paths src --ignore-src -r -y
$ pip3 install -r audio_common/requirements.txt
$ colcon build

Docker

You can create a docker image to test audio_common. Use the following common inside the directory of audio_common.

$ docker build -t audio_common .

After the image is created, run a docker container with the following command.

$ docker run -it --device /dev/snd audio_common

To use a shortcut, you may use following command:

$ make docker_run

Nodes

audio_capturer_node

Node to obtain audio data from a microphone and publish it into the audio topic.

Parameters

• format: Specifies the audio format to be used for capturing. Common values are pyaudio.paInt16 (16-bit format) or other formats supported by PyAudio. Default: pyaudio.paInt16

• channels: The number of audio channels to capture. Typically, 1 for mono and 2 for stereo. Default: 1

• rate: The sample rate that is is how many samples per second should be captured. Default: 16000

• chunk: The size of each audio frames. Default: 4096

• device: The ID of the audio input device. A value of -1 indicates that the default audio input device should be used. Default: -1

• frame_id: An identifier for the audio frame. This can be useful for synchronizing audio data with other data streams. Default: ""

ROS 2 Interfaces

• audio: Topic to publish the audio data captured from the microphone. Type: audio_common_msgs/msg/AudioStamped

audio_player_node

Node to play the audio data obtained from the audio topic.

Parameters

• channels: The number of audio channels to capture. Typically, 1 for mono and 2 for stereo. Default: 1

• device: The ID of the audio input device. A value of -1 indicates that the default audio input device should be used. Default: -1

ROS 2 Interfaces

• audio: Topic subscriber to get the audio data captured to be played. Type: audio_common_msgs/msg/AudioStamped

music_node

Node to play the music from a audio file in wav format.

Parameters

• chunk_time: Time, in milliseconds, that last each audio chunk. Default: 50

• frame_id: An identifier for the audio frame. This can be useful for synchronizing audio data with other data streams. Default: ""

ROS 2 Interfaces

• audio: Topic subscriber to get the audio data captured to be played. Type: audio_common_msgs/msg/AudioStamped

tts_node

Node to generate audio from a text (TTS).

Parameters

• chunk: The size of each audio frames. Default: 4096

• frame_id: An identifier for the audio frame. This can be useful for synchronizing audio data with other data streams. Default: ""

ROS 2 Interfaces

• audio: Topic publisher to send the audio data generated by the TTS. Type: audio_common_msgs/msg/AudioStamped

• say: Action to generate audio data from a text. Type: audio_common_msgs/action/TTS

Demos

Audio Capturer/Player

$ ros2 run audio_common audio_capturer_node

$ ros2 run audio_common audio_player_node

TTS

$ ros2 run audio_common tts_node

$ ros2 run audio_common audio_player_node

$ ros2 action send_goal /say audio_common_msgs/action/TTS "{'text': 'Hello World'}"

Music Player

$ ros2 run audio_common music_node

$ ros2 run audio_common audio_player_node

$ ros2 service call /music_play audio_common_msgs/srv/MusicPlay "{audio: 'elevator'}"