How To Change Your Assistant's Voice

Using TTS server plugin

General

TTS server plugin is using services hosted by OVOS community: https://openvoiceos.github.io/community-docs/325-members/#members-hosting-services

That reduces local resource consumption. Its recommended to use TTS server plugin for resource constraint hardware (eg. RPi 3)

[!NOTE] You can host TTS server by yourself - check this blog post for details
The post is specifically for running Piper TTS, but could also be adapted to run other TTS plugins.

Changes

Open the file for editing. nano ~/.config/mycroft/mycroft.conf.

If your file is empty, or does not have a "tts" section, you need to create it. Add this to your config

{
    "tts": {
        "ovos-tts-plugin-server": {
            "voice": "amy-low"
        }
    }
}

For available voices please refer to piper-voices section

Using local piper plugin

General

Local piper plugin does all Text-To-Speech translation on your local device.

Changes

Open the file for editing. nano ~/.config/mycroft/mycroft.conf.

If your file is empty, or does not have a "tts" section, you need to create it. Add this to your config

{
    "tts": {
        "module": "ovos-tts-plugin-piper",
        "ovos-tts-plugin-piper": {
            "voice": "alba-medium"
        }
    }
}

For available voices please refer to piper-voices section

[!WARNING]
After a configuration change it's important to reload the audio service. (an example used with systemd services systemctl --user restart ovos)

If your system is not using systemd - please refer to relevant documentation

Voices available for piper-plugin

List of available voices can be found here.

You can listen to voice samples on the official piper site - https://rhasspy.github.io/piper-samples/