Voiceover glossary
Short, clear definitions of 27 terms related to AI voiceover, call centers and voice technology.
- After-hours message
- A prompt played to callers outside call center opening hours, stating the opening hours and alternative channels.
- Ready-made IVR templates
- AI voiceover
- Turning written text into speech that sounds very close to a human voice using deep-learning models. Used for call center prompts, e-learning, advertising and corporate communications.
- The complete guide
- API (Application programming interface)
- An interface that lets software systems talk to each other. A voiceover API lets you send text, receive an audio file and add voiceover to your own systems automatically.
- Automated voiceover via API
- Biometric data
- Data based on a person's physical or behavioral characteristics that can identify them. A voice may count as biometric data when used to identify or verify a person; under Türkiye's KVKK and the EU's GDPR it is a special category of personal data.
- Bitrate
- The amount of data used per second in an audio file (e.g. 128 kbps). Higher bitrates mean better quality and larger files. For speech, a 128 kbps MP3 is enough for most uses.
- Cross-border data transfer
- Transferring personal data to a recipient in another country. In Türkiye it is governed by Article 9 of KVKK; since 1 June 2024, adequacy decisions, appropriate safeguards and incidental transfer mechanisms apply.
- AI voiceover and KVKK
- Diction
- Pronouncing words clearly, intelligibly and correctly. One of the most important criteria when choosing a voice for announcements and training.
- Dubbing
- Translating the speech in a video into another language and re-voicing it. Unlike voiceover, it must match the timing and lip movements of the original speech.
- Explicit consent
- Consent that is specific, informed and freely given. Under data protection laws such as KVKK, it is usually required for processing such as cloning a person's voice.
- Expression (stability)
- A setting that controls how emotional and varied, or how calm and consistent, an AI voice's read is. More expressive settings suit ads; more consistent ones suit announcements and training.
- IVR (Interactive Voice Response)
- A call center system that routes callers to the right team using pre-recorded prompts and key-press menus.
- Voiceover for IVR prompts
- MP3
- A compressed, widely supported audio file format. Suitable for video, presentations and the web; some phone systems may require other formats.
- Multilingual voice
- An AI voice that can speak several languages with the same voice identity. It lets you produce different language versions of an announcement with the same timbre.
- Narrator
- The voice that reads the script in a video, course or audiobook. Using the same narrator across all modules gives corporate content consistency.
- Personal data
- Any information relating to an identified or identifiable natural person. A person's name, phone number and voice are personal data.
- Pitch
- The frequency characteristic that makes a voice sound high or low.
- Pronunciation dictionary
- A list of rules mapping how a word is written to how it should be read (e.g. "GDPR" → "gee dee pee ar"). Ensures brand names and abbreviations are read correctly in every voiceover.
- Pronunciation dictionary guide
- Prosody
- The combination of stress, intonation, rhythm and pauses in speech. The most important sign of a natural-sounding AI voice.
- Sample rate
- The number of samples taken per second of audio (e.g. 44.1 kHz or 48 kHz). Higher sample rates give broadcast-quality audio; phone lines usually use 8 kHz.
- Sonic branding (Voice identity)
- The set of voice, tone and pronunciation rules a brand uses consistently across all of its communication channels.
- SSML (Speech Synthesis Markup Language)
- A markup language used to give speech systems instructions such as pauses, emphasis and pronunciation.
- Text normalization
- Converting numbers, dates, abbreviations and symbols into a readable form before voicing (e.g. "2026" → "twenty twenty-six").
- Text-to-speech (TTS)
- Technology that converts written text into synthetic speech. Modern AI models produce results very close to a human voice.
- What is AI voiceover?
- Usage rights (license)
- Permission defining in which media, for how long and for what purposes generated audio can be used. In AI voiceover, commercial usage rights are usually included in paid plans.
- Voice cloning
- Creating an AI voice that sounds like a specific person from recordings of their voice. It should only be done with the voice owner's explicit written consent.
- Voiceover
- Voicing a script for a video, ad, course or announcement. It can be done by a human voice actor or with an AI voice.
- AI voice or voice actor?
- WAV
- An uncompressed audio file format. It offers the highest quality but large file sizes; preferred for professional editing and by some phone systems.
Turn your script into voice in seconds
Let's schedule a 30-minute demo for your team, try Voice Go with your own scripts and prepare a quote that fits your volume.