GiliSoft VoiceLab

Give scripts a voice, transform recorded speech, and create custom voices you can use again. Bring narration, lessons, and voice projects together in one Windows app.

  • Generate speech from text with a voice you choose
  • Convert voices from audio, video, or a microphone clip
  • Create reusable voice clones and manage AI tasks
GiliSoft VoiceLab for text to speech, voice conversion, and reusable custom voices

How to Use GiliSoft VoiceLab on Windows

Turn a script into speech, change a recorded voice, or create a reusable custom voice. Choose a task below for its workspace, settings, generation steps, and output checks.

What Would You Like to Do?

Generate Speech from Text

Use a written script for narration, lessons, or a voiceover. Choose the voice and pace, then create an audio file in the output folder you select.

Enter or import a short script

Open Text to Speech and type your script in Text to Synthesize, or select Import Text File. Start with a short passage so you can check pronunciation and pacing before generating a longer project.

Text to Speech combines the script, target voice, speed, output folder, point balance, and Generate Audio controls.

Review the wording and character count after using Import Text File. Remove headings or notes that should not be read aloud. A short test containing a name, a number, and a sentence break is useful for checking pronunciation.

Choose and preview the target voice

Select Choose Voice, search or filter the Voice Library, and play a sample. Select Use for the voice you want, then confirm it under Target Voice in the generation workspace.

  1. Find the voice

    Use Search voices and the available filters, including language. Select Clear Filters when an earlier search hides the voice you need.

  2. Listen to a sample

    Select Play on a voice card or Preview Voice in the generation workspace. Compare the accent, tone, and language with your project.

  3. Use the voice and confirm the selection

    Select Use, then check the voice shown under Target Voice in Text to Speech or Voice Conversion. Use the favorite control to find it again later.

Voice Library lets you search, filter, preview, and choose a voice. The selected card is marked In Use.

A library preview plays a sample, not necessarily your own script. Generate a short passage to check names, numbers, and pauses in your content.

Set speed and output, then generate audio

Choose Speed and use Browse beside Output Folder to select the save location. Review Billing Rule and Points Left, then select Generate Audio to submit the task.

Speed
Start at 1.0x and adjust when needed. Compare the same test passage after a change.
Output Folder
Select Browse and choose a recognizable project folder for the audio.
Billing Rule
Review the rule shown for this operation, including any minimum charge or duration rounding.
Points Left
Check the current balance before submitting the AI task. A trial download is not unlimited free generation.

Select Generate Audio once and follow the submitted job in Task List. For additional credits, use the GiliSoft Points purchase page; a rate or balance pictured in a screenshot is not your current account value.

Check the task and listen to the saved audio

Open Task List and check the job's status. After it completes, locate the audio in the selected Output Folder and listen to the full result. Check the target voice, pronunciation, pace, opening, and ending before using it in your project.

If the job is still processing or has failed, use Check Tasks and Retry a Failed Job to review it. Open the completed audio from its actual output folder rather than relying only on a voice-library preview.

Change a Voice in a Recording

Start with speech already recorded in audio or video, or record a microphone clip. The source supplies the spoken content; Target Voice sets the voice used for the generated audio.

  1. Upload or record the source

    Open Voice Conversion. Select Upload Source Audio or Video, drop a media file into the workspace, or choose Record Source Audio. Listen to the clip and check that the speech is clear.

  2. Select and preview the target voice

    Use Choose Voice and Preview Voice to select a library voice or saved custom voice. Confirm the entry under Target Voice; the voice-selection tutorial shows the controls.

  3. Generate and check the converted audio

    Choose Speed and Output Folder, review Billing Rule and Points Left, then select Generate Audio. Follow the job in Task List and listen to the completed file.

Voice Conversion accepts source media or a microphone clip, then uses the selected Target Voice to generate audio.

Use a clear speaker without heavy background music or overlapping voices. If a microphone clip is silent, check the selected input and Windows microphone permission before recording again.

Video is an input for the speech source. This workflow uses Generate Audio. Adding the resulting voice to a video and synchronizing it with the picture is a separate editing step.

See how to change a voice in an audio recording for a focused example.

Create a Custom Voice Clone

Create a voice entry you can reuse for later Text to Speech or Voice Conversion tasks. Creating the voice and generating finished narration are separate operations.

  1. Provide a reference voice sample

    Open Custom Voice > Create Voice Clone. Select Upload Audio or Video, or choose Record Audio for a microphone reference. Use your own voice or a speaker who has given permission, with one clear speaker and little background noise or music.

  2. Name and identify the voice

    Enter Voice Name, choose Language, and add useful Tags. Give it a recognizable name, such as Course Narrator, so it is easy to find later.

  3. Create the entry and generate a test

    Select Create Voice Clone when the sample and details are ready. Find the completed entry in My Voices, select Use, and generate a short test in Text to Speech or Voice Conversion to check the result.

Create Voice Clone takes a reference sample plus the voice name, language, and tags. The built-in reference recorder shown here stops at 30 seconds.

The 30-second limit shown here applies to the built-in reference recorder, which stops automatically at that limit. Follow the current upload guidance for audio or video files rather than assuming every uploaded sample has the same limit.

Manage and Reuse Custom Voices

  1. Find the saved voice

    Open Custom Voice > My Voices. Search for the voice by name, clear filters when needed, or use Recent and Favorites to return to an entry.

  2. Preview and use it for a project

    Select Play to hear a sample and Use to select the voice. Return to Text to Speech or Voice Conversion and confirm it under Target Voice before generating audio.

  3. Keep the list organized

    Use Rename for a clear speaker or project name and the favorite control for frequently used voices. Use Delete only for an entry you no longer need. Renaming or selecting a voice does not itself generate a new audio file.

My Voices keeps custom entries available for reuse. Recent, Favorites, and search help you return to the voice you need.

Check Tasks and Retry a Failed Job

  1. Locate the submitted job

    Open Task List. Use Search tasks, All Status, All Types, or Refresh to find the job by its time, type, and selected voice.

  2. Review status, details, and the point record

    For a job still processing, refresh its status instead of submitting duplicates. For Failed, open Detail and check the source, chosen voice, output folder, and available points against the reported cause.

  3. Retry after correcting the cause, or open the result

    Correct the issue before using Retry. For a completed task, open the audio in the selected Output Folder and listen to the full file. Check your own point record; the example's Released status does not establish the charging outcome for every failure or retry.

This Tasks example shows a failed job with Detail and Retry. It illustrates troubleshooting, not a completed export.

A voice-library sample and a completed project output are different files. Confirm the completed job, output path, and saved audio before placing it in a lesson, video, or other project.

Check the finished voice and troubleshoot problems

Words and pronunciation
Compare the result with the script or source. Listen carefully to names, numbers, abbreviations, and sentence endings.
Voice and pace
Confirm the intended target voice and check that Speed suits the lesson, narration, or recording.
Sound quality
Listen for noise, missing speech, distorted sections, and abrupt joins. Check more than the opening sample.
File and task
Confirm the completed status, output location, and point record before treating the job as finished.
ProblemNext check
The wrong voice is usedConfirm the selected card under Target Voice. Select Use on the intended library or custom entry before generating again.
Speech is too fast or sounds unnaturalTest a short passage, adjust Speed, and revise punctuation or difficult wording. Compare the same passage after each change.
Converted speech is unclearListen to the source for background music, noise, or overlapping speakers. Try a clearer clip before repeating the full task.
A custom voice is difficult to findOpen My Voices, check the name, and clear the search or filters. A library voice and a custom entry are managed in different tabs.
A job remains in progressUse Task List and Refresh to review the same job. Avoid submitting duplicates while waiting.
A job fails or no output appearsCheck status and Detail, then verify the source, selected voice, point balance, and Output Folder. Retry after correcting the cause.

GiliSoft VoiceLab FAQ

What is the difference between Text to Speech and Voice Conversion?

Text to Speech starts with a written script and generates spoken audio in the selected voice. Voice Conversion starts with an existing audio or video recording, or a microphone clip, and creates audio using the selected target voice. Choose by the source material you already have.

How do I create a custom voice clone in VoiceLab?

Open Custom Voice, select Create Voice Clone, and upload a reference audio or video file or choose Record Audio. Enter Voice Name, select Language, and add useful Tags. Create the voice, then find it under My Voices and select Use for a later generation task.

How long should the custom voice reference be?

Use a clear, single-speaker sample and follow the upload guidance shown by your current app. The built-in reference recorder illustrated in this guide stops automatically at 30 seconds. That recording limit should not be treated as a rule for every uploaded file.

Can I reuse, rename, and organize a custom voice?

Yes. Open Custom Voice and choose My Voices. Voice cards provide Play, Use, Rename, Delete, and a favorite control. Give each voice a recognizable name and use Recent, Favorites, or search to find it again.

Where is the generated audio saved?

Before generating, choose Output Folder with Browse in the Text to Speech or Voice Conversion workspace. After the task completes, check that folder and play the audio file. Task List provides status and Detail for checking the job; a failed task should not be treated as a saved result.

What should I do when a generation task fails?

Open Task List, select Detail, and review the error together with the task's source, voice, and point status. Correct the source or settings before selecting Retry. Do not submit duplicate jobs just because one is still processing, and check the displayed point record rather than assuming how a retry will be charged.

Is VoiceLab free, and how are AI points used?

VoiceLab has a trial download, while AI generation and custom voice creation use GiliSoft Points. Check Points Left and the displayed Billing Rule for the operation before processing. A free download does not mean unlimited free generation; the purchase page lists the available point packages.

Does uploading a video automatically produce a dubbed video?

Voice Conversion accepts video as a source for the spoken audio. The workflow shown here uses Generate Audio, so review the resulting voice file as audio. Adding that voice to a video and synchronizing it with the picture is a separate editing step.

How can I improve pronunciation and speaking speed?

Test the same short passage with your selected voice, adjust punctuation or difficult wording, and change Speed when needed. A library preview plays a sample, not necessarily your own script. Generate and listen to your test passage before processing the full text.

Create the voice your next project needs

Start with a script or recording, choose a voice, and bring narration, lessons, and voice projects to life with GiliSoft VoiceLab.