AI Media
A guide to voice profiles, timbre, speaking style, emotion, pace, pitch, emphasis, pauses, pronunciation, SSML, speaker consistency, and responsible voice cloning.
12 matching terms
Voice character
Voice profile
A compact description of a synthetic speaker's age range, vocal weight, character, and delivery.
Use it as the stable foundation before adding emotion or scene-specific direction.
warm adult narrator, clear mid-range voice, calm and trustworthy deliveryVoice character
Timbre
The tonal color that makes one voice sound distinct from another at a similar pitch and volume.
Use it to request qualities such as bright, breathy, resonant, smooth, or textured.
soft resonant timbre with a light airy texture, no raspPerformance
Speaking style
The communicative manner of delivery, such as conversational, instructional, documentary, or promotional.
Use it to align the voice with audience, format, and purpose.
conversational explainer style, friendly but concise, avoid announcer-like deliveryPerformance
Emotional delivery
Controls the emotional attitude expressed through tone, energy, rhythm, and emphasis.
Use one clear emotional direction and describe how strongly it should be expressed.
quiet relief with restrained warmth, emotionally sincere, not theatricalProsody
Speaking rate
The overall speed of spoken delivery, usually perceived through words, pauses, and phrase length.
Use it to fit a time limit or improve comprehension for complex material.
moderately slow pace, approximately 135 words per minute, clear phrase boundariesProsody
Pitch range
The span and movement of vocal pitch used during speech.
Use a narrower range for calm authority or a wider range for expressive storytelling.
natural mid pitch with a narrow range, gentle downward ending on statementsProsody
Emphasis
Makes selected words more prominent through stress, duration, pitch, or volume.
Use it to clarify contrast, key benefits, warnings, or calls to action.
place subtle emphasis on 'secure' and 'before Friday', keep surrounding words neutralProsody
Pause control
Specifies where and how long the voice should pause between ideas.
Use it for natural phrasing, dramatic timing, list separation, or edit points.
short pause after the title, medium pause before the final sentencePronunciation and markup
Pronunciation guide
Provides an intended pronunciation for names, acronyms, brands, numbers, or unfamiliar words.
Use it whenever default pronunciation could be ambiguous or domain-specific.
pronounce 'SQL' as 'sequel' and 'Nguyen' as 'win'Pronunciation and markup
SSML
Speech Synthesis Markup Language adds structured controls for pauses, emphasis, pronunciation, and prosody.
Use it when a synthesis engine supports markup and precise repeatable control is required.
Supported SSML elements and behavior vary by service.
<speak>Welcome.<break time="500ms"/>Let's begin.</speak>Consistency
Speaker consistency
Keeps the same recognizable voice qualities across separate sentences, scenes, or generation sessions.
Use it for recurring narrators, characters, courses, and episodic content.
maintain the same speaker identity, timbre, accent, and recording character in every segmentConsent and safety
Voice cloning consent
Documented permission from a person to create and use a synthetic version of their voice.
Use only authorized source recordings and define the allowed purpose, audience, and duration.
Do not imitate or clone an identifiable person's voice without informed permission.
use only the approved speaker voice model for the authorized training narration