SCRIPT
TTS script preparation
We prepare approved text, expand abbreviations and agree how numbers, dates and units should be read. We divide scripts into segments linked to slides, scenes or files so that later corrections are easy to identify.
Scope of support / Audio / TTS
A good voice starts with a well-prepared script.
Text-to-speech speeds up the production of additional language versions, but requires careful work on scripts, intonation and editing. We combine AI voice generation with audio editing and recording checks.
VOICE
Voice selection and samples
We prepare AI voice samples suited to the language, tone and audience. We agree on pace, narration style and expression. The approved sample serves as the reference for production and subsequent recording checks.
PRONUNCIATION
Names and specialist terminology
We create a pronunciation guide: names, abbreviations and difficult terms are written in a way that helps the AI TTS voice pronounce them correctly. For example, we can use “kay-pee-eye” for “KPI” if needed. We check pronunciation in a sample with your SME.
TTS
AI voiceover production
We generate audio in ElevenLabs or another agreed TTS environment. We adapt settings and script preparation to the capabilities of the selected model. Production is organized in controlled batches to maintain a consistent narration style.
EDITING
Audio editing and adaptation
We edit recordings, adjust pauses and loudness, remove unwanted sections and prepare versions that meet the specification. For videos and courses, we check segment lengths against the visuals and the time available to present the content.
QA
TTS recording checks and delivery
We listen for omissions, repetitions, artifacts and departures from the script. Language queries are submitted for approval. We deliver organized WAV/MP3 files or other agreed formats with version information.
How we work together
Clear responsibilities. Agreed approval points.
We adapt the scope to your project. Your team makes linguistic and subject-matter decisions, while KeboDesign handles the agreed production and technical checks.
Explore who does what:
Hover, focus or click to highlight the selected team’s responsibilities.
1Brief, script and quote
Your team
You provide the script and source video or audio for a quote, together with languages, duration, preferred style and technical requirements.
KeboDesign
We analyze the production scope, number of files and synchronization needs. We agree the cost, schedule and approval stages.
2Sample and pronunciation
Your team
You approve the voice, pace and pronunciation of names and terminology. You supply the target translations.
KeboDesign
We prepare representative samples and flag pronunciation queries. We adapt the script for TTS generation.
3Production and editing
Your team
You confirm the final wording before production and answer language queries.
KeboDesign
We generate recordings, edit audio tracks and fit segments to the video or course. We retain agreed file names and identifiers.
4Listening review and corrections
Your team
Your language reviewer or SME checks the wording, pronunciation and pace in Frame.io. The video preview provides visuals and narration, with comments linked to specific moments.
KeboDesign
We share a review version in Frame.io, collect approved feedback and correct the relevant TTS segments. We check consistency with the remaining audio and visuals.
5QA and delivery
Your team
You approve the recordings and final delivery specification.
KeboDesign
We check file completeness, script accuracy, formats and audio levels. We deliver the approved language versions.
Pronunciation approval
First the sample. Then full production.
In specialist videos and courses, a product name or abbreviation can matter more than an impressive voice. We therefore check selected terms in a sample before generating the full audio set.
The reviewer approves the content and pronunciation guide. TTS-specific spelling is kept separate from the text published in subtitles. We share audio or narrated video in Frame.io so that your SME can assess pronunciation, context and synchronization and easily add comments.
A translation longer than the source does not have to mean unnaturally fast narration. First, we ask for the wording to be shortened without losing meaning. If that is not possible, we assess whether a section of the source video, a title card or a pause can be extended to allow a natural speaking pace. Feasibility depends on the actual material.
Deliverables / Audio / TTS
Materials ready for the next stage.
We agree formats and the final delivery scope in the brief. Your team receives approved deliverables with a clear handover.
Multilingual audio files
Audio tracks after editing, checks and linguistic approval. Parameters and file structure are adapted to their intended use.
WAV / MP3 · agreed specificationsSegment and script package
Recordings linked to scenes or slides, together with approved text. Changes can be identified by segment ID.
Scripts · identifiers · versionsAudio adapted to your content
Duration, pauses and editing tailored to the video or course. Integration with the source is included where agreed.
Audio tracks · synchronization · QAMaterials for a quote
What helps us plan the work?
- A script or sample material, target languages and approved translations.
- Preferred tone, a reference sample and pronunciation guidelines.
- The video, course or details of the required segment structure and duration.
- Format requirements, number of files, deadline and designated approver.
Questions / Audio / TTS
Let’s define the right scope of support.
Can TTS voiceover replace every voice talent session?
The recording approach depends on the content, desired expression and client requirements. We start with a sample so that you can evaluate the AI voice in the context of your project.
Should we supply translated scripts ready for recording?
Yes. An approved translation adapted for recording is essential: it should sound natural and remain as close as practical to the source length. This helps narration fit the scenes without excessive acceleration. We handle TTS script preparation, samples, editing and technical checks; any necessary shortening is agreed with your team.
How do you check the pronunciation of specialist terms?
We identify difficult names and abbreviations, prepare samples and collect feedback from your reviewer or SME. We apply the agreed rules throughout production and verify them by listening.
Can just one section of a recording be corrected?
Yes, provided we have approved wording and a segment that can be isolated. After editing, we check joins, pace and consistency with the rest of the recording.
Which audio files can we receive?
Usually WAV or MP3 with agreed specifications. We adapt file structure, naming and delivery to the video, Storyline project, LMS or other environment.
What determines the cost of TTS production?
Script length, number of languages and files, pronunciation requirements, editing and revision rounds. We agree any voice usage and platform costs before production.
Contact / Audio / TTS
Let’s discuss your content, languages and project scope.
Send a sample or a link to your materials. We will define requirements, delivery formats and the information needed for a quote.
Go to the contact page ↗AI Voice Generation for E-learning
AI TTS narration can support standalone audio and videos embedded in a course. Quality depends on more than the voice and script: sample approval, consistent pronunciation and review in the training context also matter.
Read the guide ↗