Every voice project begins with the same question: who should speak?
In BreezeBlue, there are three practical ways to answer it. You can choose a ready-made voice from the Voice Library, design an original voice from a description, or clone a specific voice from an authorized recording. After choosing one, you can use Voice Direction to shape its performance.
The best option depends less on the final format and more on what you already have at the beginning.
The short answer
| Your starting point | Best choice |
|---|---|
| You need a strong voice quickly and are flexible about identity | Voice Library |
| You have a character brief or brand personality but no recording | Voice Design |
| You need a specific real voice and have that speaker's explicit permission | Voice Clone |
| You already like the voice but need a different emotion or delivery | Keep the voice and use Voice Direction |
That final row prevents one of the most common mistakes: redesigning or recloning a voice when the real problem is the performance.
Choose the Voice Library when speed matters
The Voice Library is the fastest route from a script to a polished read. Browse voices, preview the ones that fit your language and style, and select a starting point.
It works especially well when:
You need a narrator, host, or character today.
You are open to discovering a voice rather than reproducing a precise idea.
You want to test a script before investing in a custom voice.
A curated voice already fits the audience and tone.
Imagine you are making a product walkthrough. Your brief calls for a clear, approachable speaker with calm confidence, but the voice is not part of the brand identity. A library voice may solve the problem immediately.
Choose Voice Design when the voice does not exist yet
Voice Design is for original voices. Instead of uploading a recording, describe the audible identity you want: age, range, texture, accent, pacing, role, and attitude.
It is the right choice when:
You are casting an original game, animation, or story character.
Existing library voices are close but miss a defining quality.
You want a distinctive brand or virtual-host voice that is not tied to a real speaker.
You need many related but clearly differentiated character voices.
For example:
A tired night-shift detective in his early fifties. Low midrange, dry and lightly gravelly, with clipped phrasing and a guarded, observant attitude. Understated rather than theatrical.
The strongest descriptions focus on sound. A long paragraph about the detective's childhood will not help unless it is translated into something audible.
Voice Design produces a new voice, not a guaranteed one-of-a-kind identity. Similar descriptions can lead to similar results, so audition the voice in the context of your cast or brand.
Choose Voice Clone when identity already exists
Voice Clone starts with a real speaker. Use it only when you have the necessary permission or rights to use the recording and reproduce the voice.
It is the right choice when:
A creator wants to narrate future videos in their own voice.
A host has authorized production of scripted episodes in their voice.
An actor has approved a voice workflow for a character.
A company has a consenting spokesperson whose vocal identity must remain recognizable.
The quality of the source matters. A clean recording of one person speaking naturally in a quiet space is more useful than a longer clip with music, echo, effects, or overlapping voices.
In BreezeBlue Creator, you can upload an MP3 or WAV sample or record in the browser, confirm consent and rights, preview the clone beside the original, and decide whether to save it.
What Voice Clone is not for
Do not use cloning to imitate someone who has not clearly consented. Do not use a clone to mislead listeners about who said something. If you need a particular archetype rather than a particular person, Voice Design is the better tool.
Voice Direction is the layer after voice choice
A voice and a performance are related, but they are not the same thing.
Suppose your narrator sounds exactly right in a calm introduction but too neutral during the conclusion. You do not need a new narrator. Keep the voice and direct the final paragraph:
Build from calm confidence to genuine excitement. Keep the pace conversational, then slow slightly before the final promise so it lands clearly.
The same principle applies to designed and cloned voices. Once saved, they remain reusable. Voice Direction changes the emotion, intent, pacing, and delivery of a particular passage while preserving the speaker's identity.
A reusable selection workflow
Write a one-sentence role brief.
Decide whether identity must be exact, original, or simply appropriate.
Choose Library, Design, or Clone.
Audition candidates with one shared script.
Test at least one neutral and one expressive direction.
Save the winner with a clear, searchable name.
Document the final voice and direction for future production.
Choose your starting point
Use the Voice Library for speed, Voice Design for an original identity, and Voice Clone for a specific consenting speaker. Then use Voice Direction to make that voice perform each moment.
The goal is not to choose the most advanced workflow. It is to choose the simplest workflow that gives your project the right speaker.