什么是配音（Voice Over）：定义、示例与工作原理

Voice over is a video production technique where you hear a speaker’s voice, but don’t see them on-screen. It’s often used in film, TV, advertising, and social media to provide more context for what’s happening in the video.

You don’t necessarily need to be a professional voice actor to create a high-quality voice over for your videos. With the right technology, you can record your own voice at home, or use AI to create a custom voice that suits your needs.

In this article, we’ll explore how voice overs work, how to create your own, and how modern AI models have made professional voice overs more accessible than ever.

Summary

A voice over is when an audio track of someone speaking is played over a video.
Voice overs are most often used to narrate or announce what’s on screen. In fiction, they’re also used to share a character’s inner monologue.
Voice overs are found in many different types of media, including film, television, video games, advertising, social media, and educational content.
With AI, you can create and customize voice overs for your creative projects.

What is voice over? Definition, purpose, and key traits

A voice over is an audio track of an unseen person speaking that plays over a video. Voice overs can consist of pre-recorded human performances or AI-generated speech.

You’ll likely see the term written in three ways: voice over, voiceover, and voice-over. All three are commonly used, with “voiceover” being typical in the production industry, while dictionaries tend to prefer the hyphenated “voice-over.”

Voice overs are used for a variety of purposes, including:

Narration: The voice over tells the story of what’s happening on screen. It’s used in both fictional and non-fictional content, including movies, TV shows, and documentaries.
Character inner monologues: In fictional content, voice overs are often used to share a character’s thoughts while they’re on screen.
Announcements: Commercials and marketing videos often use voice overs to announce new products and deals.
Education: Training videos often combine visual demonstrations with a voice over that provides step-by-step instructions.

During a voice over, the tone, pacing, and delivery of the speaker’s voice match the visual content on screen so the entire video flows smoothly. These key traits are present whether a human or AI model creates the voice over.

Voice over vs. voice acting, narration, and dubbing

The term ‘voice over’ is often confused with other audio and video production techniques, including voice acting, narration, and dubbing. While there’s some overlap between these concepts, they all have different definitions.

Voice acting is a specific type of voice over where the speaker is playing a distinct character or using an extensive range of emotions. Voice acting is most common in animated movies, TV shows, and video games, where the voice over is a key component of the story.

Narration is another type of voice over where the off-screen voice tells the story of what’s happening on screen, or provides more context for the plot. However, the speaker isn’t usually a specific character. Narration is a common storytelling technique in documentary-style videos.

Dubbing is when a new audio track replaces the existing audio in a video. For example, you might dub a TV show in a different language to make it accessible to international audiences, or dub in re-recorded dialogue if the existing audio quality is poor.

Many video production teams use a combination of different voice over techniques and dubbing to achieve the final product. The table below puts the four techniques side by side:

Technique

What it is

Does the speaker play a character?

Does it replace existing audio?

Common uses

Voice over

An unseen speaker’s voice plays over a video

Sometimes

No, it is added to the video

Ads, explainers, documentaries, social media, training

Voice acting

A performance in a character’s voice, often with a wide emotional range

Yes

Animation, video games, audio dramas

Narration

An off-screen voice that tells the story or adds context

Usually not

Documentaries, audiobooks, explainer videos

Dubbing

A new audio track replaces the original audio

Often, to match the original speakers

Yes

Translating films and TV, replacing poor-quality audio

Voice over examples: Film, ads, games, and explainers

Voice overs exist in many different types of media, including:

Film and television: Voice overs are necessary to create animated movies and TV shows. They’re also used in live-action film for narration and storytelling that doesn’t fit into on-screen dialogue.
Non-fictional entertainment: Voice over narration is common in reality TV, documentaries, and online explainer videos. The voice over guides the plot and provides additional context for what’s happening on screen.
Video games: Video games use voice overs for both interactive in-game dialogue and cutscenes. High-quality voice overs make video games feel more immersive and engaging.
Advertising: Voice overs are used in both online and TV ads to grab the viewer’s attention, reinforce brand messaging, and announce specific products, services, and promotions.
Instructional videos: Voice overs are used in corporate training videos, online courses, and how-to videos to provide step-by-step instruction and guidance.

Today’s AI models can create voice overs for all of these media categories. AI voice overs are a helpful solution when budgets or production timelines are tight. They also work well for testing different voices and styles before making a final commitment.

What does a human voice over artist do?

Human voice over artists provide audio for many different types of video projects, including film, television, video games, ads, and educational content. They can act out dialogue, provide narration, and read scripts.

Recording the audio is just one part of a voice over artist’s job. Many of these artists have acting backgrounds, which they use to create a character or persona that’s appropriate for the project and aligns with the director’s vision. They adjust their pacing, tone, volume, and even their accent as they record to bring this character to life.

Some voice over artists also have technical audio skills, which they use to configure their microphones or even edit their own recordings. However, not all voice over artists offer these services.

How to create a voice over with AI, step by step

With modern AI models, you don’t need a human voice over artist for every project. AI helps you create realistic, high-quality voice overs without a lengthy casting or production process.

Here’s a step by step breakdown of how to create a voice over with AI using ElevenCreative Studio:

Describe your project or upload a script: Describe the type of project you want to create to the AI model. Specify the medium, such as an audiobook, podcast, or YouTube video. Then, provide details about the type of voice you want, such as the speaker’s accent, gender, and tone. If you have a script already written, you can upload it here, or AI can help you write one.
Let Studio Agent draft a first cut: The Studio Agent will create a first draft based on your prompts. From there, provide feedback to refine the voice over so it fits your needs.
Pick or adjust a voice from the Voice Library: The Voice Library includes Iconic Voices from renowned human voice actors, alongside a huge range of other styles, accents, and languages. Test out different voices to see which one works best for your project, or customize a voice by adjusting the emotion and delivery.
Refine the script with the text-based editor: After you’ve generated your first draft, make adjustments to the script until you’re happy with the result. The text-based editor makes it easy to edit the script, no technical skills required.
Export the final product: Once you’re happy with the voice over, export the audio file so you can use it in your video editing software of choice. You can also make basic video edits directly in ElevenCreative Studio. Upload MP4 or MOV files and add your voice over, background music, sound effects, and captions directly to the video.

How to write a voice over script

The best voice overs are those supported by a quality script that’s written for the ear.

These tips will help you write a script that sounds natural when read aloud:

Write the way you speak: Use short sentences, everyday words, and contractions. Read each line aloud, and rewrite anything that makes you stumble. Watch out for cacophony that muddles a line.
Match the script to the visuals: Plan each line around what is on screen so the words and images reinforce each other. Leave short gaps where the visuals can speak for themselves.
Keep one idea per sentence: Listeners can’t reread a line, so avoid long, nested sentences.
Mark pauses and emphasis: Add notes for where the speaker should slow down, pause, or stress a word. For hard-to-pronounce names or terms, add a phonetic spelling.
Check the length: Most narration runs at a comfortable pace of roughly 150 words per minute, so use that as a starting point to estimate how long your script will run.

With a finished script in hand, you’re ready to record or generate it in a few minutes with AI.

Choosing a voice: Voice Library, Voice Design, or creating your own

To create your voice over, choose from pre-made AI voices or create your own.

The ElevenCreative Voice Library features over 17,500 distinct AI voices. This includes a variety of languages and accents from all over the world, as well as a wide range of styles and tones.

You can use the Eleven v4 emotion wheel to see a voice come to life.

See Eleven v4 in action

You’ll also find licensed voices of famous actors and characters. Human voice actors can clone their voices and earn cash rewards when paid subscribers use their voice.

You can also create your own voice using the Voice Design feature. With this feature, you’ll describe the type of voice you’re looking for in a text prompt. The Eleven v4 model uses that prompt to create a unique voice. From there, you can adjust features like the speaker’s accent, pacing, tone, pitch, and emotion. This is an easy way to test out different voices and find the best fit for your project.

Another option is Voice Cloning, which creates a replica of your voice based on a short audio sample. Once you’ve cloned your voice, use text to speech to create a script and then adjust the emotion and pacing of the delivery. This is a great option if you want to use your own voice, but don’t have professional recording equipment.

AI voice over vs human voice over

There’s absolutely a place for both human and AI voice overs in creative industries. Voice actors are the backbone of video production, with studies suggesting the industry is valued at around $5.8 billion.

Many voice actors have even combined their skills with AI, creating replicas of their voices to use on platforms like ElevenLabs. Doing so allows them to generate income when people use their voice, without the need to individually record each of these projects on their own schedule.

The right choice between hiring an actor for a voice over and using AI comes down to your budget, timeline, performance needs, and the scope of your project. Here’s a reference to get started with:

Factor

AI voice over

Human voice over

Cost

Usually priced per project or per hour, and costs rise with studio time, usage rights, and revisions

Speed

Minutes to generate, with instant regeneration after script changes

Requires casting, scheduling, recording, and often a rebooking for script changes

Languages

One voice can speak dozens of languages, so you can localize quickly

Each language usually needs a separate native-speaking artist

Control

Edit the text, then adjust tone, pacing, and emotion yourself

Direction happens through takes and feedback, which gives nuanced results but takes more time

Performance

Highly realistic and expressive for most narration, explainer, and social content

Best for projects that need deep acting range or a very distinctive character performance

Best for

Tight timelines, frequent updates, multilingual content, and testing voices

Flagship brand campaigns, feature film, and character-driven performances

Create professional voice overs with ElevenCreative

Create a custom voice over for your next project with ElevenCreative’s Studio 4.0. Whether you choose from our Voice Library, use Voice Design, or even clone your own voice, you can adjust every detail of the voice over until it’s the perfect fit.

The Voice Library has over 17,500 voices to choose from for your next project. Sign up for ElevenCreative or learn more about Eleven v4 to see what’s possible with AI voice overs.