Language and Voice Settings

Digital Person Language and Voice

Digital DNA Studio allows you to customise voice settings for your Digital Person, enabling them to speak in a style appropriate for their role and personality.

Open

Choose a language and voice for your Digital Person.

Choose a Language

The Choose a Language dropdown allows you to select a language for your Digital Person to speak. Regional variations are also provided where there are voices available for that region. For example, English is available in many variations including English (US), English (UK), and English (Australia).

A Digital Person may speak only one language.

Choose a Voice

You may Choose a Voice for your Digital Person from the list of available voices. The list of suggested voices is filtered based on your selected language in the Choose a Language dropdown. Changing the Language selection will change the list of suggested voices.

Tune Your Voice

When using a Google voice, you may optionally modify the way the voice is expressed by changing the Speed and Pitch.

The default values for an unmodified voice are Speed: 1 and Pitch: 0.

Open

Google Voice Tune Controls

Test Your Voice

You may test your voice in two ways; either with or without the Digital Person. Testing with the Digital Person requires a refresh of the preview, while testing without the Digital person does not require a refresh.

Testing with the Digital Person

To test a voice with the Digital Person, make your voice selection and then click Update Preview to refresh the Digital Person. When the Digital Person has loaded, enter some text in the Test your Digital Person text box and then click Play Script.

Voice Service Providers

Voices are available from the following speech providers:

• Google Speech Services

• Microsoft Cognitive Services

• Amazon Polly

Audience Language

Configuration of the audience language settings is important for the quality of the overall experience, as Speech Recognition accuracy will be higher if the end user speaks using the language and accent configured in Digital DNA Studio.

Choose a Language

Select a language and region (accent) which most closely approximates the expected language and accent of your audience. The accuracy of speech recognition will be higher when the end user speaks to the Digital Person with a language and accent matching that which is expected by the speech recognition service.

Choose a Provider

Language support varies per provider. If the selected language is not supported by a particular provider, then that provider will not be available as an option.

Speech recognition is available from the following providers:

• Google Speech Services

• Microsoft Cognitive Services (HumanOS 2.6+)

Add Speech Hints

The Speech Hints feature helps improve the accuracy of your Digital Person’s speech recognition capability. Training your Digital Person to understand utterances can help improve the conversation flow and may reduce confusion when your Digital Person struggles to comprehend or keep up with the conversation topic.

Digital DNA Studio provides a facility that enables you to build a list of words or phrases that you want to train your Digital Person to recognize and understand. These include hard to pronounce and uncommon words/phrases which can become difficult to understand, such as unique or special words that are used within your business, e.g., product/brand names, part numbers, jargon, geekspeak, etc.

Projects in a Foreign Language

Creating a project in a foreign language is supported through the Studio. We currently support these languages including Arabic, French, Japanese, and Mandarin (Chinese).

Here is how you create an experience in a foreign language:

1. We support foreign language corpora, using Dialogflow and Watson NLP.

2. Create your corpus as you normally would.

3. Set your Language in the Studio (see detailed steps in Creating a New Project section). 

4. Deploy and give it a spin.

Contents