Loading the exercise
Loading the exercise
Turn a written notice into natural-sounding, verified Hungarian narration — without voice cloning.
ElevenLabs uses text-to-speech (TTS) technology to create speech from written text: voice-overs for video, learning material, presentation narration or accessible content. In this exercise you turn a short (130-170 word) Hungarian text into natural-sounding narration: you learn how the voice, the model, the speed and punctuation shape the result, how to check Hungarian pronunciation, abbreviations and numbers, and how to label AI use transparently. Use only built-in or freely available voices — voice cloning is not part of the exercise.
In this exercise you will learn: how to turn on-screen text into listenable narration; how voice, model, speed and punctuation shape the result; how to check Hungarian pronunciation, abbreviations and numbers; how to make easy-to-follow audio; how to label AI voice use transparently; and what legal/ethical limits apply to AI voices.
Suggested schedule: preparing the narration text: 15-20 minutes; first generation and error hunting: 15-20 minutes; improving pronunciation, pace and phrasing: 20-30 minutes; comparison and final export: 15-20 minutes; documentation and reflection: 10-15 minutes.
Tip: The motto of this exercise: REWRITE. GENERATE. LISTEN BACK. FIX. — the four stages of the flow: 01 Text (turning it into spoken language) → 02 Voice (model and built-in voice) → 03 Test (pronunciation and pace) → 04 Export (audio file and labeling).
By the end of the exercise you will have: a 130-170 word Hungarian narration text; a first, unfixed voice generation; at least two variants compared with different settings; a pronunciation and quality-check log; a final 60-90 second MP3 or WAV file; an accompanying text with source labeling; and a short reflection on how the text changed after listening.
Free preview
The first step is open to everyone. The remaining 21 steps, progress tracking and feedback open up with a free account.