Running the Service

This document provides instructions on how to run the tts service node and interact with it.

How to Run the Service

To run the text-to-speech service, you first need to source your ROS2 environment and then use the ros2 run command.

You can run the service node with the following command. This will start the TTS client and make it available for receiving text-to-speech requests.

# Source your ROS2 workspace
source install/setup.bash

ros2 run tts service

When running the service in docker, you can enter the container with the following command, where the above steps are already done:

docker exec -it tts-node bash

With that, you should be able to follow the next instructions.

Parameters

You can customize the behavior of the node by passing the following ROS parameters.

ros2 run tts service --ros-args -p <argument>:=<value>

Argument	Description	Default
`server_url`	The URL of the llama.cpp server's completions endpoint for TTS inference.	`http://localhost:8080/v1/completions`
`en_model`	Model identifier for English TTS.	`en`
`en_voice`	Voice profile to use for English text-to-speech.	`leah`
`en_max_tokens`	Maximum number of tokens to generate for English TTS.	`10240`
`en_temperature`	Controls randomness in English TTS generation. Higher values increase creativity.	`0.6`
`en_top_p`	Nucleus sampling parameter for English TTS. Controls diversity of token selection.	`0.9`
`en_repeat_penalty`	Penalty for token repetition in English TTS to encourage more varied output.	`1.1`
`de_model`	Model identifier for German TTS.	`de`
`de_voice`	Voice profile to use for German text-to-speech.	`max`
`de_max_tokens`	Maximum number of tokens to generate for German TTS.	`10240`
`de_temperature`	Controls randomness in German TTS generation. Higher values increase creativity.	`0.6`
`de_top_p`	Nucleus sampling parameter for German TTS. Controls diversity of token selection.	`0.9`
`de_repeat_penalty`	Penalty for token repetition in German TTS to encourage more varied output.	`1.1`

Service Requests

`/tts`

To convert text to speech, you can call the /tts service.

It uses the ric_messages/srv/TextToAudioBytes service type. Replace ${text} with your desired text and ${language} with the target language.

ros2 service call /tts ric_messages/srv/TextToAudioBytes "{'text': '${text}', 'language': '${language}'}"

Supported Languages

The service currently supports the following languages:

English: Use "english" or "en"
German: Use "german" or "de"

Example Usage

Convert English text to speech:

ros2 service call /tts ric_messages/srv/TextToAudioBytes "{'text': 'Hello, how are you today?', 'language': 'english'}"

Convert German text to speech:

ros2 service call /tts ric_messages/srv/TextToAudioBytes "{'text': 'Hallo, wie geht es dir heute?', 'language': 'german'}"

Response

The service returns audio data in WAV format as a byte array in the audio field of the response. This audio can be saved to a file or played directly by audio processing applications. You can use the convert.py script as an example to save the audio response to a WAV file. First, save the service response to a JSON file, then convert it:

# Call the service and save response to JSON
ros2 service call /tts ric_messages/srv/TextToAudioBytes "{'text': 'Hello world', 'language': 'english'}" > input.json

# Convert the JSON response to a WAV file
python3 convert.py

The convert.py script reads the JSON response, extracts the audio byte array, and saves it as output.wav:

After running the script, you can play the generated output.wav file with any audio player.

Troubleshooting

Common Issues

Service not available: Ensure the TTS server is running and accessible at the configured URL
Unsupported language: Check that you're using a supported language identifier ("en"/"english" or "de"/"german")
Empty audio response: Verify that the input text is not empty and the model parameters are correctly configured
GPU out of memory: If using GPU deployment, consider reducing LLAMA_ARG_N_GPU_LAYERS or switch to CPU deployment

Checking Service Status

To verify the service is running:

ros2 service list | grep tts

To check service type:

ros2 service type /tts

Logs

View service logs when running in Docker:

docker logs tts-node