Creating voiceovers using synthetic voices

Creating voiceovers using synthetic voices

Turn a written script into a finished voiceover without a recording studio. In Feltiv's Voice Over editor you write or paste your script, give each line a synthetic voice from the voice library, generate the audio, check it against a video if you wish, and export it as a single MP3 or WAV file.

This article walks you through the whole journey, from starting the project to exporting the audio.

Before you start

  • Have your script ready. You can type it straight into the editor or paste it in.
  • Decide which voice, or voices, you want. Each voice you add becomes a speaker in the project, so a script with two narrators needs two voices.
  • If you want to check your audio against a video, keep the file handy. MP4, MOV and WebM are supported.

Step 1: Start a Voice Automation project

  1. Click New in the left-hand menu to open the Start a project page.
  2. Select the Voice Automation card. Its description lists the languages you can create voiceovers in, and the audio formats you can export (MP3 and WAV).
  3. Click Start Project.
  4. In the Start a Voice Automation Project window, the workflow is already set to Voiceover with AI. Enter a unique Project Name. The Description is optional.
  5. Click Start. The Voice Over editor opens straight away.

[[IMAGE 01 - kb-screenshots/voiceover/01-voice-automation-card.jpg - Voice Automation card on the Start a project page]]

[[IMAGE 02 - kb-screenshots/voiceover/02-start-voiceover-window.jpg - Start a Voice Automation Project window with a project name entered]]


Step 2: Write your script

The editor has three areas: the Script panel on the left, the Line / Video panel on the right, and the timeline along the bottom. A new project starts with one empty line.
  1. Click into the line and type or paste your text. Each line is one block of speech, and each line gets its own audio.
  2. Click Add line to add another line. Keep lines to a sentence or a short paragraph so that they are easy to review and regenerate one at a time.
  3. To add a line directly under an existing one, open the line's three-dot menu and choose Add line below, or press Alt + Shift + N.
  4. To remove a line, open its three-dot menu and choose Delete.
Once a line has a voice, the Add line arrow lets you choose Add line as… so that the new line starts with a speaker you have already set up.

[[IMAGE 03 - kb-screenshots/voiceover/03-voiceover-studio-empty.jpg - The empty Voice Over editor: Script panel, Line panel and timeline]]

[[IMAGE 04 - kb-screenshots/voiceover/04-script-line-typed.jpg - A script line typed in]]


Step 3: Choose a voice for each line

  1. On the line, click Assign speaker and choose New speaker from voice library….
  2. The Browse Voice Library window opens. Narrow the list with the filters on the left:
    • Language: pick the language of your script.
    • Gender: All, Male or Female.
    • Engine: the technology that produces the voice, such as Microsoft Text-to-Speech, Google Gemini or Sarvam Bulbul.
    • Style: for example Cheerful, Friendly, Narration, Newscast or Professional. Pick the tone that suits your content.
  3. Use the search box at the top to find a voice by name. The All, Favorites and Recent tabs help you get back to voices you like. Click the star on a voice card to save it as a favourite.
  4. Click the yellow play button on a card to hear a short preview before you decide.
  5. Select the voice you want and click Use this voice.
Good to know: some voice cards carry a small warning marker, for example a Google Gemini voice when you have filtered to English (United States). Hovering over it explains that the voice is not specifically tuned for that language or accent and uses a generic English voice instead. Preview the voice first and choose another if the accent matters to you.

The voice becomes a speaker in the project. It appears on the line, in the Speaker section of the Line panel, and as its own track on the timeline. To use a different voice for another line, assign a new speaker from the voice library, or choose an existing speaker from the list.

[[IMAGE 05 - kb-screenshots/voiceover/05-voice-library.jpg - Browse Voice Library with filters, voice cards and the Use this voice button]]

[[IMAGE 06 - kb-screenshots/voiceover/06-speaker-assigned.jpg - A line with a speaker assigned and the Regenerate button]]


Step 4: Generate the voiceover

  1. Click Regenerate on a line to create the audio for that line, or Regenerate all at the top of the Script panel to create audio for every line that needs it.
  2. When a line is ready, its duration appears next to its number and a waveform appears on the timeline.
  3. Click the play button on the line, or in the Line panel, to listen to it.
If you change the text of a line, or change its voice, the line is marked Needs regenerate. Click Regenerate again to create fresh audio that matches your change.

Step 5: Review each line

Click a line to open it in the Line panel on the right. There you can:
  • Play the line and see its start time, end time and duration.
  • See which voice it uses, and click Change voice to pick another from the voice library.
  • Check the Version history. Each time you generate a line, Feltiv keeps that take. The take currently being used is marked In use, and you can play earlier takes to compare them.
The Pronunciation section, for teaching a voice how to say names, brands and acronyms, is marked Coming soon.

You can also drag, trim and zoom the blocks on the timeline. See Editing segments on the timeline for the full set of controls.

[[IMAGE 09 - kb-screenshots/voiceover/09-line-generated-version-history.jpg - Line panel showing the generated line, speaker and Version history with Take 1 In use]]

[[IMAGE 10 - kb-screenshots/voiceover/10-needs-regenerate-after-voice-change.jpg - A line marked Needs regenerate after the voice was changed]]


Optional: check your voiceover against a video

  1. In the right-hand panel, switch from Line to Video.
  2. Drop a file onto the panel, or click to browse. MP4, MOV and WebM files are supported.
The video plays in sync with the timeline while you work on the voiceover, so you can see whether each line lands at the right moment.

[[IMAGE 07 - kb-screenshots/voiceover/07-video-tab.jpg - The Video tab with Attach a video]]


Step 6: Export your voiceover

  1. Click Export at the top right of the editor. The Export Voiceover window opens. Earlier exports of this project are listed under Previous exports.
  2. Under New export, Full Mix gives you a single mixed voiceover track.
  3. Choose the Format: MP3 or WAV.
  4. Click Generate.
The Generate button stays greyed out until the project has generated audio, so generate your lines first.

[[IMAGE 08 - kb-screenshots/voiceover/08-export-dialog.jpg - Export Voiceover window with Full Mix, MP3 and WAV]]


Your work is saved as you go

The editor saves automatically. The status at the top of the page shows Saving… and then All changes saved. Use the undo and redo arrows next to it if you change your mind. The three-dot menu at the top lets you adjust the on-screen font size and check the Auto Save setting.

Tips

  • Generate and review a couple of lines first, then use Regenerate all once you are happy with the voice.
  • Use one speaker per voice. Keeping a narrator on a single speaker keeps the timeline tidy and the voice consistent.
  • Preview voices in the library with a sentence from your own script in mind, since tone and pace can differ between voices.

Related articles

  • Editing segments on the timeline
  • Downloading your output
  • Understanding the project details page