When generating audio for your virtual tour, the AI text-to-speech tool can mispronounce specific proper names, words, short-hand phrases, or acronyms, that are used throughout your content and can be spoken differently than the AI’s interpretation. To improve AI audio pronunciation of these terms in your Circuit virtual tour, the Audio Pronunciation section of your CMS allows you to create a lexicon of terms and their correct pronunciation that will then be referenced by the AI when generating audio in your virtual tour.
⚠️ To improve the consistency of your desired result in your AI-generated audio, always preview the target word in a sentence. This helps the AI understand important context clues like the language of the content.
To create a pronunciation record to improve AI pronunciation in your tour:
As a recommendation, before generating audio across your tour, think of tricky words or phrases that the AI might struggle with and add them to the Audio Pronunciation section of your CMS in advance.
Navigate to the Audio and Pronunciation section of your Circuit CMS.
Click Add Word.
In the Word field, enter the term as it is written in your virtual tour copy.
An example sentence will be pulled from your content, if one is available. You can use this sentence, or enter your own. Press the play button beside the sentence to preview how AI audio generates for this word in your selected default voice profile.
If the term sounds correct as-is, then no Word entry is needed. You can navigate away from the page without saving a new Word. If the term sounds incorrect, continue to step #6.
In the Pronunciation field, write out the phonetic pronunciation of this term as you would like the AI to pronounce it. Click the play button alongside the sentence again to preview how the AI audio generates this term. Continue making adjustments until you are happy with the result.
Click Create to save your record.
Frequently asked questions
Can I improve pronunciation for terms in non-English languages?
If you have Circuit Multi-language in your plan, use the Language Menu to see additional language fields and define the desired pronunciation in the Pronunciation field for your selected language using the same steps above.
How do I create pronunciation record when reviewing audio in a scene or hotspot?
If you identify a term that you’d like to create a pronunciation record for once your audio has already been generated, click the Help Text to navigate to the Audio and Pronunciation section of your CMS to create the record. If you do this in a new tab, the correct pronunciation will be included when regenerating audio in the original tab. As always, we recommend previewing the generated audio for correct pronunciation accuracy before saving.
Ready to generate audio in your scenes?