How to Use SSML Tags with Ondoku’s New AI Voice | Explained with Audio Samples

Thank you for using Ondoku.

We’re excited to introduce “OndokuBeta”, a new AI voice that can read text aloud “more naturally” and “more humanly”!

AI Voice Synthesis [Beta] - Next-Generation TTS with Rich Emotional Expression | Ondoku

Generate expressive audio with the latest AI technology. Next-generation text-to-speech with 25 voices to choose from. Create natural-sounding audio with speed and pitch adjustments and customizable reading styles.

With Ondoku’s new AI voices, you can now describe reading styles freely and express a wide range of things, including emotions, acting, age, and dialects!

And there’s more!

Depending on the voice, the new AI voices can also use “SSML tags.”

This article introduces examples of how to use SSML tags.

Why not try expressing things in different ways with the new AI voices?

Examples of Using SSML Tags with the New AI Voice “OndokuBeta”

With “OndokuBeta”, you can use SSML tags such as these.

All sample audio uses “Ellis.”

<prosody rate="〇〇"> Tag | SSML Tag for Adjusting Reading Speed

Use the <prosody rate="〇〇"> tag to adjust the reading speed with an SSML tag.

• <prosody rate="slow">: Speaks slowly.
• <prosody rate="fast">: Speaks quickly.

Reading style: “Bright and energetic”

<prosody rate="slow">This is important, so I’ll explain it slowly.</prosody><prosody rate="fast">We’re short on time, so I’ll explain it quickly.</prosody>For comparison, I’ll explain it at a normal pace.

<prosody pitch="〇〇"> | SSML Tag for Adjusting Voice Pitch

Use the <prosody pitch="〇〇"> tag to adjust the pitch of the voice reading the text aloud.

• <prosody pitch="+2st">: Reads aloud in a higher voice (you can change the number).
• <prosody pitch="-3st">: Reads aloud in a lower voice (you can change the number).

Reading style: “Bright and energetic”

<prosody pitch="+2st">I’ll speak brightly in a slightly higher voice.</prosody><prosody pitch="-3st">Conversely, I’ll speak calmly in a slightly lower voice.</prosody>Lastly, I’ll speak at my normal pitch.

<emphasis level="〇〇"> | SSML Tag for Emphasizing a Specified Range

Use the <emphasis level="〇〇"> tag to emphasize specified text with an SSML tag.

• <emphasis level="strong">: Strongly emphasizes the specified text.
• <emphasis level="moderate">: Emphasizes the specified text less strongly than <emphasis level="strong">.

Reading style: “Business”

This item is <emphasis level="moderate">important</emphasis>, but the next item is <emphasis level="strong">even more important</emphasis>.

<say-as interpret-as="digits"> | SSML Tag for Reading Numbers Digit by Digit

Use the <say-as interpret-as="digits"> tag to read 1234 digit by digit as “one two three four,” instead of “one thousand two hundred thirty-four.”

Reading style: “Bright and energetic”

Our inquiry number is <say-as interpret-as="digits">03 1234 5678</say-as>.

The <break time="○○ms"/> Tag Cannot Be Used

The <break time="○○ms"/> tag, which creates pauses (periods of silence) in audio, is not supported by the new AI voices.

What’s more, it can cause the text after the tag to be omitted and not read aloud, so please do not enter <break time="○○ms"/>.

You also cannot create pauses by entering lots of punctuation, such as “....,” adding multiple line breaks, or writing something like “(3-second pause).”

Please use video or audio editing software instead.

*With the previous voices, you can still add pauses using the SSML <break> tag, as before.

Related Articles on SSML Tags and Reading Styles

For information on SSML tags supported by the original “Ondoku,” see this article.

Be sure to check it out as well.

What Is Speech Synthesis Markup Language (SSML)? How to Use It with Text-to-Speech Software and a List of Key Codes | Ondoku Text-to-Speech Software

SSML stands for Speech Synthesis Markup Language. By writing SSML code, you can further control how Ondoku speaks. This article explains in detail how to use SSML in Ondoku and the codes available.

This article also explains in detail examples of using the “reading styles” for the new AI voices.

You can create more nuanced expressions than with SSML tags, including emotions, personality, age, and dialects. Why not give it a listen and try it for yourself?

[Voice Tuning] A Complete Guide to Reading Styles for the New AI Voices. Create More Human-Like Emotional Expressions and Performances for Free!

A complete guide to voice-tuning techniques for the next-generation AI voice “Ondoku Beta.” Easily create emotional expressions such as joy, anger, sorrow, and happiness, specify age, create character voices, use dialects, and even speak foreign languages such as English. This article introduces tips and techniques for tuning AI text-to-speech voices that are free for commercial use, with samples.

Why Not Try Using SSML Tags with the New AI Text-to-Speech Voices?

With the AI voice “OndokuBeta”, which now reads aloud more humanly and naturally, you can create an even wider range of expressions using SSML tags.

Why not start by trying the new AI text-to-speech voices for free?

AI Voice Synthesis [Beta] - Next-Generation TTS with Rich Emotional Expression | Ondoku

Generate expressive audio with the latest AI technology. Next-generation text-to-speech with 25 voices to choose from. Create natural-sounding audio with speed and pitch adjustments and customizable reading styles.