> ## Documentation Index
> Fetch the complete documentation index at: https://hedra.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# Avatar Speaker Selection
> How to define which character should be speaking in a multi-character image.
## 1. Add a Start Frame
Upload or create a starting image that contains multiple characters.
You can choose from:
* **Generate**
* **Upload**
* **Take Photo**
* **Library**
***
## 2. Select the Speaker
Next to the uploaded image:
1. Click **Select Speaker** \
(the person icon on the image toolbar, next to the trash icon)
***
## 3. Position the Speaker Indicator
On the image:
1. Drag the green onto the character you want to speak
2. Resize the indicator box if needed
3. Continue creating your video
The selected character will be the one who speaks, sings, or performs based on your audio input.
***
## Available Speaker Models
Hedra supports a range of voice providers:
* **OpenAI Voices** – built-in options for natural speech
* **ElevenLabs Voices** – high-fidelity, expressive voices
* **Gemini Voices** – rich and nuanced speech output
Choose the voice that best matches your project’s style and tone.
***
## Notes
* Speaker options may vary depending on your subscription and model access.
* Some voices may require external API keys or provider configuration in your account settings.
* Voice change does not alter lip-sync timing — only the voice output.
***
## Tips for Best Results
* Choose a voice that matches the character persona you defined.
* Preview each voice before finalizing.
* Use expressive voices for narrative and emotive content.
* Use neutral voices for informational or tutorial videos.