Generate narration with Authoring AI 2.0 voices (beta)

The quick way

1. In the Admin App, go to Libraries>Brainsharks, click the Brainshark's name, and select Edit Brainshark.
2. Select a slide, open the Notes tab, enter your script, and click Save.
3. Click Generate narration, then set Voice library to Authoring AI 2.0.
4. Choose a language and voice. Click Play to hear a sample.
5. Select Pronunciation dictionary to use your organization’s preferred pronunciations.
6. Click Generate and save, then click Play to listen.

Do this step by step

Open the presentation editor

1. In the Admin App, go to Libraries.
2. Under Content, select Brainsharks.
3. Click the Brainshark's name.

4. Select Edit Brainshark.

   Tip: The list shows every Brainshark in the selected Library/Division. Use the search bar to find a Brainshark by name, and the items-per-page control at the bottom to page through the list.

Generate audio with an Authoring AI 2.0 voice

To use Authoring AI 2.0, select it as the voice library when you generate narration.

1. Select the slide you want to generate audio for.
2. In the Slide properties panel on the right side, click the Notes tab.
3. Enter your script into the text box and click Save.
4. Click Generate narration above the notes field.
5. In the Voice library selector, select Authoring AI 2.0.
6. Select a language and choose your preferred AI voice from the library. To hear a sample, click the play icon.

7. If you'd like to ensure specific words and terms are pronounced as your organization prefers, check the Pronunciation dictionary box.

   Note: An administrator must enable the pronunciation dictionary first. See Set up AI pronunciation rules.

8. Click Generate and save.

9. Click the play icon to listen to the narration.

   Note: Authoring AI 2.0 voices are multilingual, so they can narrate content in supported languages, but they may keep their native accent.

   Tip: To generate audio for all slides with notes at once, click Narration at the top of the page, select Generate narration, and choose an Authoring AI 2.0 voice.

Add emotional context tags to your script

Emotional context tags let you direct how a line is delivered, like giving a note to a voice actor. The narrator performs the emotion, not just the words.

1. In the Notes tab, open the script for your slide.
2. Add a tag in square brackets where you want the effect. Tags can go anywhere in a sentence, not only at the start. Tag categories include:

• Emotions: Shape the feeling of the delivery, for example [curious], [crying], [mischievously].
• Delivery direction: Control how something is said, for example [whispers], [shouts].
• Human reactions: Add natural texture, for example [laughs], [clears throat], [sighs].

1. Click Save and generate to add narration with an Authoring AI 2.0 voice.

Example: [clears throat] Let me walk you through what happened. And I have to say, [laughs] it wasn't what we expected.

Control pauses and pacing

• Use [pause], [short pause], or [long pause] to add pauses.

• Commas, semicolons, colons, periods, and hard returns control pauses and pacing more consistently.

  Note: The [break] tag doesn't work with Authoring AI 2.0 voices. It still works with Authoring AI 1.0 voices.

Control pronunciation with phonemes

With Authoring AI 2.0, phonemes work in every language, not only English. You can add them in two ways:

• Pronunciation dictionary: Add phonemes to the pronunciation dictionary for any language. Aliases continue to work in all languages.
• Inline International Phonetic Alphabet (IPA) in your script: Replace a word in your script with its IPA notation between slashes, for example /ˈdɛfɪbrɪleɪtər/. Inline IPA works on single words only. You don't need a dictionary entry.

Read about the International Phonetic Alphabet for details.

Use custom voices with Authoring AI 2.0

Custom voices (voices you've uploaded and cloned) automatically get the same capabilities as Authoring AI 2.0 voices: better audio quality, emotional context tags, and expanded phoneme support. You don't need to do anything.

Authoring AI 1.0 voices remain available if you prefer them. You can switch libraries at any time in the voice library selector.