Monotonic, flat narration can quickly ruin an otherwise great video or audio presentation. While standard neural voices sound smooth, truly engaging voiceovers require natural emotional inflections, such as enthusiastic announcements, somber storytelling, or dramatic whispers, along with well-timed pauses between sentences.

TTS Engine includes built-in support for expressive neural styles and SSML (Speech Synthesis Markup Language). You can either select emotional tones visually from the interface or write inline tags directly into your script to blend multiple emotions across a single paragraph.

Here is how to unlock and control expressive voice styles in TTS Engine.

Note: Advanced emotion styles and millisecond-accurate pause sliders are powered by the Azure Speech API. Make sure you have entered your free Azure key in the ⚙️ API menu (as shown in our Azure API setup tutorial) to reveal these controls.

Step 1: Select a Voice Marked with a Star (⭐)

Not every neural voice supports emotional modulation. In TTS Engine, voices capable of expressive styling are clearly indicated:

  1. Under VOICE SETTINGS, click the voice selector button to open the Voice Catalogue.
  2. Look for voices tagged with a gold star (⭐) next to their name:
    • English: Jenny ⭐ (cheerful, sad, whispering, newscast, friendly, chat, assistant), Guy ⭐ (newscast, cheerful, sad), Aria ⭐ (cheerful, empathetic, newscast, sad), Sonia ⭐, Ryan ⭐.
    • Spanish: Jorge ⭐ (cheerful, sad, whispering).
    • French, German, Italian, Portuguese: Denise ⭐, Henri ⭐, Katja ⭐, Conrad ⭐, Elsa ⭐, Francisca ⭐.
  3. Click Select on any starred voice, then click Accept.

Step 2: Choose an Emotional Tone from the Dropdown

Once an eligible voice is selected, the Emotion dropdown appears automatically on the right side of the voice settings card:

  1. Click the Emotion: menu (by default set to Default).
  2. Choose the speaking style that fits your scene:
    • Cheerful: Great for marketing intros, energetic tutorials, and celebratory news.
    • Whispering: Ideal for suspenseful storytelling, bedtime audio, or dramatic secrets.
    • Sad: Suitable for reflective commentary, audiobooks, and emotional narratives.
    • Newscast / Friendly / Assistant: Best for professional presentations, guides, and corporate videos.
  3. Once selected, this style applies to your entire script during generation.

Step 3: Calibrate Inter-Line Pauses

In natural human speech, speakers pause between thoughts and paragraphs. Without pauses, lines separated by line breaks can sound rushed.

  1. Locate the Line pause: X ms slider directly beneath the voice selectors.
  2. Drag the slider to define how much silence to insert whenever you press Enter in your script:
    • 300 ms (Default): A subtle, natural sentence break.
    • 600 ms – 1000 ms: A clear conversational pause between distinct paragraphs.
    • 1500 ms – 3000 ms: Dramatic silence before transitioning to a new topic or question.

(TTS Engine automatically converts your line breaks into prosodic <break time="..."/> tags behind the scenes).

Step 4: Advanced Scripting with Inline SSML Tags

If you need a single paragraph to shift emotions mid-sentence (for instance, starting cheerful, inserting an exact silence, and ending with a whisper), you can write SSML tags directly into the text editor.

TTS Engine automatically detects < and > markup and processes your raw SSML payload.

Example: Combining Multiple Emotions and Pauses

Paste the following example into the editor using en-US-JennyNeural:

<mstts:express-as style="cheerful">
  Welcome to our channel! We have a huge announcement today.
</mstts:express-as>

<break time="800ms"/>

<mstts:express-as style="whispering">
  Keep this between us for now, because it is still top secret.
</mstts:express-as>

Common SSML Tags You Can Use:

Step 5: Listen and Export to MP3

  1. Click Listen to verify how the emotional shifts and pauses sound together.
  2. If the tone needs tweaking, change the emotion keyword or adjust your millisecond break values.
  3. Once the performance sounds natural, click Save MP3 to store your polished audio file on your PC.

Using emotional styles and SSML gives your audio creations a level of nuance and production value that standard text-to-speech tools simply cannot replicate.

<< Back